3 Ambient-Power-enabled-IoT-White-Paper

Ambient Power-enabled IoT
Second Edition of
“Zero Power Communication”
01.
CONTENT
01
Foreword
02. 05
Typical use cases
of Ambient IoT
03. 21
Technical principles of Ambient IoT
communication
04. 37
Overall design of Ambient IoT
communication system
05. 61
Key techniques and challenges
of Ambient IoT
06. 86
Standardization and future trends
of Ambient IoT
07. 100
Epilogue
Foreword 01
FOREWORD
In recent years, the industry has adopted a number of IoT (Internet of
Things) technologies to fulfil the communication requirements for various
target usage scenarios. Nevertheless, as the boundaries of IoT usage
expand, more scenarios emerge and features like ultra-low cost, very small
form-factor, low power consumption and maintenance-free are becoming
of increasing importance for field deployment and operation. The newly
emerged Ambient Power-enabled IoT technology (a.k.a. Ambient IoT tech-
nology), which integrates environmental radio power harvesting, ultra-low
power communication and low-power computing, is anticipated to meet all
the stringent requirements and to become one important candidate of
next generation technology for Internet of Things.
01
Foreword 02
The status-quo
1.1 of IoT
Wireless communication technology flourished since 1990s. Digital mobile communication system has evolved
from 2G, 3G and 4G to 5G today, with each generation to serve the needs of voice calls, messaging service and
mobile Internet during their time of service. Nevertheless, a trend is becoming more and more clear - requirement
on IoT communication rises together with societal shifts and economic development with an accelerating pace.
Technologies and standards related to the mobile IoT have been developed and evolved since 2010. Among
them, 3GPP (3rd Generation Partnership Project) has standardized series of IoT technologies such as MTC (Ma-
chine Type Communications), NB-IoT (Narrow Band IoT) and RedCap (Reduced Capability UE). Techniques used
in MTC and NB IOT, such as reduced bandwidth, single antenna, reduced peak rate, half duplex and reduced
transmission power have significantly lowered the cost of IoT terminals. Furthermore, power consumption of IoT
terminals is greatly reduced by the introduction of eDRX (enhanced Discontinuous Reception) and PSM (Power
Saving Mode). At the same time, existing IoT solutions like MTC/NB-IoT can also support a large number of termi-
nals for simultaneous network accessing to meet the need of massive connection.
Apart from 3GPP, standardization organizations like IEEE also released technologies tailored for IoT. IEEE 802.11
standardized a series of WiFi technologies, for example 802.11b/n are widely used for local area network based
IoT. GS1/ISO published RFID technology for short-distance stock take application in a point-2-point mode. LoRa
protocols defined by LoRa Alliance is another type of wide area IoT technology known by large coverage, low
power consumption, large capacity and low cost.
Unsatisfied communication
1.2 requirements of IoT
Existing technologies such as MTC and NB-IoT have achieved low cost, low power consumption and massive
connection of IoT terminals, thus they are able to meet the communication requirements of IoT in many scenarios.
However, there are still many situations where communication requirements cannot be satisfied with existing
technologies, such as:
Harsh communication environment
Certain scenarios of IoT may face extreme environmental conditions such as high temperature, extremely low
temperature, high humidity, high pressure, high radiation or high-speed movement. Some examples of these haz-
ardous circumstances include ultra-high voltage power stations, railways carrying high-speed trains, environmen-
tal monitoring in over/under temperature areas and industrial production lines. In these scenarios existing IoT
terminals will not work due to factors like failure of conventional power supplies. In addition, maintenance of IoT
devices (e.g., recharging or replacing batteries) becomes challenging under extreme conditions. Battery-free and
maintenance-free IoT devices are highly desired for these application scenarios.
03 Ambient Power-enabled IoT
Scenarios where ultra-small devices are requested
For some IoT communication scenarios, e.g., goods traceability, commodity circulation and smart
wearables, small form-factored devices are requested for practical use. For example, IoT devices
used for commodity management in the circulation are normally in the form of small tags and is em-
bedded into the commodity packaging. As another example, small and light-weight wearable devices
can improve the user experience while meeting communication needs.
Scenarios asking for ultra-low cost IoT devices
Many IoT scenarios ask for ultra-low cost IoT devices to enhance competitiveness over other alter-
native technologies. For examples in logistics or warehousing scenarios, in order to facilitate the
management of a large number of goods in circulation, devices can be attached to each package of
pallet in order to complete the accurate management of the whole logistics process. These scenari-
os require that the cost of IoT devices be sufficiently competitive.
In sum, abovementioned scenarios request battery-free, ultra-low power consumption, very small
size and ultra-low cost IoT devices. Meanwhile, typical applications demand IoT systems provide suf-
ficiently large network coverage and system throughput as well as flexible network deployment abili-
ty. Additionally, various IoT communication requirements exist across private consumption scenarios
such as enabling smart terminal devices to access IoT devices to support smart home and smart
wearable applications.
Existing IoT technologies are either limited by their power consumption levels so that have to rely on
wired or battery power, or cannot provide the requested communication range and networking abili-
ties. Cost of IoT devices and system maintenance also needs to drop to reduce total cost of owner-
ship (TCO). How to meet these unsatisfied communication requirements of IoT and to better serve
the economic and social development is a problem worthy of discussion and study.
Position and vision

1.3 of Ambient IoT
Ambient IoT technology utilizes key techniques such as RF energy harvesting, ultra-low power com-
munication and low-power computing. An Ambient IoT device powers up itself by harvesting and
accumulating small amount of energy contained in radio waves, light, heat and motion. In this way
the device does not rely on conventional batteries. Furthermore, ultra-low power communication
and low-power computing technologies make the device achieve an extremely simplified RF and
baseband circuit structure, which can greatly reduce the cost, size and energy consumption. There-
fore, Ambient IoT is expected to enable battery-free devices to meet the communication needs of
IoT with ultra-low power consumption, very small size and ultra-low cost.
Due to the battery-free nature, such kind of devices are also named as Ambient IoT devices and the
corresponding communication procedure may also be called Ambient IoT communication.
Foreword 04
Logistics, Retail, Manufacturing, Supply chain, Remote metering, Sensors,

Supply chain, Sensors, Smart Grid, Agriculture,Alarming,
Clothing,etc. Agriculture, Indoor positioning, Security etc.
Smart Home etc.
Existing loT
Ambient loT
RFID
Lower capabilities Higher capabilities
Figure 1.3-1 Technical position of Ambient IoT
Ambient IoT technology is devoted to fulfil abovementioned unsatisfied IoT communication require-
ments. Ambient IoT is not positioned to replace existing IoT technologies; in contrast it is designed to
provide the device and system capabilities which existing IoT technologies cannot provide. Just like
Figure 1.3-1 shows, there are clear differentiation between technical abilities of Ambient IoT and
those of existing IoT technologies. The scenarios Ambient IoT meets and the corresponding require-
ments also have little overlap with existing IoT technologies.
NB-IoT MTC RedCap Ambient IoT
0
Power Cost Latency Connections Throughput Size
Consumption
Figure 1.3-2 Comparison of IoT technologies
As shown in Figure 1.3-2, compared with existing technologies such as MTC, NB-IoT and RedCap,
Ambient IoT will have significant advantages in terms of device power consumption, size and cost.
For example, in term of power consumption, the device power consumption is expected to be
reduced from tens of milliwatts of NB-IoT terminals to dozens of microwatts or even several micro-
watts. In term of cost, the device cost is expected to be reduced from more than ten RMB yuan of
the cheapest NB-IoT terminal (lowest among existing technologies) to 1 RMB yuan or even lower.
Therefore, with the obvious differences between the above and other technologies, Ambient IoT is
expected to become an important candidate technology for the next generation of IoT.
MTC / NB-IoT RedCap Ambient loT
2010 2015 2020 2025
Figure 1.3- 3 Roadmap of IoT technology development
To sum up, Ambient IoT will be committed to meet the communication requirements that existing
technologies are still unable to meet and achieve a good complementarity with the existing technolo-
gies, so as to meet the multi-level and multi-dimensional communication requirements of IoT.
Typical use cases of Ambient IoT 02
TYPICAL USE
CASES OF AMBI-
ENT IOT
The primary technological advantage of the Ambient IoT is battery-free
communication. By utilizing key technologies such as environmental
energy harvesting, ultra-low power communication and ultra-low power
computing, Ambient IoT terminals can operate without batteries while
maintaining minimal hardware complexity. Therefore, the Ambient IoT
devices can meet the requirements of extremely low power consumption,
small size and extremely low cost. It is foreseeable that Ambient IoT will
have significant application advantages in a wide range of fields. For
example, applications such as smart manufacturing and industrial moni-
toring, smart logistics, smart warehousing, smart agriculture, and smart
power grid for vertical industries, as well as smart home and indoor posi-
tioning for individual consumers. In this section, we will select some typical
scenarios to demonstrate the potential applications of Ambient IoT in
these fields.
02
Intelligent manufacturing and

2.1 industrial monitoring
In the context of Industry 4.0, the aim of intelligent manufacturing and industrial monitoring is to
achieve automation and intelligence in the manufacturing process by using new information technol-
ogies and manufacturing technologies. More specifically, intelligent manufacturing is a kind of
advanced manufacturing method, which integrates intelligent inventory, intelligent classification and
intelligent production. Accordingly, industrial monitoring refers to a real-time perception of the pro-
duction process, production quality, equipment operation and environmental safety, so as to obtain a
comprehensive overview of the industrial system. Intelligent manufacturing and industrial monitoring
support each other to jointly improve the flexibility, agility and intelligence of manufacturing system
such that it is able to cater for the rapidly changing market demands and the highly competitive envi-
ronments.
In order to meet the requirements of automation and intelligence in the scenario of intelligent manu-
facturing and industrial monitoring, on the one hand, it is necessary to perform real-time tracking,
positioning and recording for the status of materials and products, and on the other hand, it is also
important to monitor temperature, humidity, production lines and dangerous events through sensor
nodes. In view of the advantages for Ambient IoT devices (i.e., ultra-low power consumption,
extremely small size and extremely low cost), Ambient IoT will have a broad prospect in intelligent
manufacturing and industrial monitoring. For example, the extremely small and battery-free tags are
suitable to be embedded in materials and products to complete inventory and tracking during the
process of production and transportation. For another example, industrial sensor nodes in the tradi-
tional IoT system may be deployed in harsh environments and special locations, or even in extremely
dangerous environments (e.g., high/low temperatures, moving or rotating parts, high vibration condi-
tions, high humidity environments). In these scenarios, legacy battery-based devices may not work
well because of the physical and chemical characteristics of the battery. In addition, it may lead to a
high cost for network maintenance when using legacy battery-based devices because it is hard or
even impossible to replace the battery in these extreme environments. However, Ambient IoT devic-
es can easily fit in these extreme environments due to its batter-free characteristic and ultra-low
power consumption. Furthermore, the issue of network maintenance can also be addressed
because Ambient IoT devices are energized based on wireless power or environmental energy and
do not need to have the battery replaced (i.e., maintenance-free). By the way, most of Ambient IoT
devices have very low cost.
Therefore, the application of Ambient IoT technology in intelligent manufacturing and industrial mon-
itoring scenarios can not only achieve the goal of industrial system automation and intelligence, but
also reduce the cost of deployment and maintenance while expanding the application scenario.
Typical scenarios in intelligent manufacturing and industrial monitoring：
In-factory logistics Containers with integrated Ambient IoT tags are sent to suppliers to load materials.
The Ambient IoT tags are activated for inventory when returning to the factory, and
hence material information is obtained and recorded in the inventory system. The reg-
istered container is then delivered to a designated area within the factory. For a target
product, the smart factory automatically generates a pick list. Accordingly, the smart
forklift moves to the designated area to obtain materials according to the pick list and
inventory system, during which it may be necessary to use the Ambient IoT tag for
positioning. Finally, the smart forklift delivers the material to the production process.
Production line For products processed on a production line, Ambient IoT tags can also be embedded
monitoring in these products. With the help of Ambient IoT tags, real-time tracking for the status
of the products on the production line can be achieved. For example, it is able to know
which process the product is currently located in or which process has been complet-
ed, so as to ensure that the product has completed all of required steps before
moving to the next step. For another example, Ambient IoT tags can also record the
production information of each step, such as the manufacturer and responsible per-
sonnel of the step, for subsequent quality traceability.
Environmental In some special environments (e.g., high temperature, high pressure, extreme cold,
information collection radiation), Ambient IoT tags are used to collect environmental information. For exam-
ple, it can monitor the temperature and humidity information in the data center or
equipment room to prevent network service quality from being affected.
Figure 2.1-1 Examples of Ambient IoT in intelligent manufacturing and industrial monitoring
Typical requirements of intelligent manufacturing and industrial monitoring
Requirements for devices： In most cases, Ambient IoT devices are electronic tags with integrated
memory for data access and/or integrated sensor for information acquisition.
Since Ambient IoT devices are usually deployed in a large scale (e.g., each
asset or device will be equipped with a tag), the cost and power consumption
need to be considered.
Power consumption：lower than 1mW, battery-free, maintenance-free.

Work environment：Work well in special environments (e.g., high temperature,
high pressure, extreme cold, radiation).
Size：extremely small size for large scale application.
Coverage：tens of meters to hundreds of meters.
Device Type：paper tag or anti-metal tag.
Requirements for network： Deployment：Flexible deployment based on network infrastructure. Network

equipment can be deployed on outdoor light poles or with digital indoor
system to provide fundamental coverage. Additional deployment can be
considered for dead zone or coverage enhancement.
Coverage：Single site needs to provide the coverage of >30m for indoor
scenario and >100m for outdoor scenario.
Security：Authorization-based reading and writing to protect privacy and
data security.
Capacity：Support sufficient system capacity and support a large number of
devices to read data.
2.2 Logistics and warehousing

With the sustained and stable development of the economy, the economic volume is getting larger
and larger, which is followed by the further expansion of the scale for logistics. Logistics is a very
important link in the supply chain of commodity circulation, and occupies an important position in
the national economy, and warehousing is the core link of modern logistics.
In logistics and warehousing scenarios, lots of goods need to be frequently transferred, stored,
loaded, unloaded and inventoried in logistics stations or warehouses (tens of thousands of square
meters). With the occurrence of warehouse ordering, entering warehouse, goods management and
leaving warehouse, a large amount of warehousing information will be generated, which has the
characteristics of frequent data reading and large quantity.
In order to manage packages/goods digitally and improve the management efficiency of logistics
and warehousing, it is usually necessary to paste the identification of terminal on the surface of
packages/goods for obtaining logistics information and managing the whole logistics process.
Therefore, the small terminal size is more conducive to industry applications. At the same time, the
company of express delivery or warehouse suppliers can only accept the very low-cost terminal due
to the huge amount of goods, cost performance and competitiveness.
Ambient IoT devices have the characteristics of extremely low cost, small size, maintenance-free,
durable, long life and so on. In logistics and warehousing, Ambient IoT devices can be used to record,
save and update the information of goods. That is the Ambient IoT based logistics and warehousing
system, which is beneficial for reducing costs, significantly improving the efficiency of logistics and
warehousing management, and realizing of smart logistics and smart warehousing. In addition, sen-
sors can be integrated with Ambient IoT devices to be used to collect environmental information
during the process of transportation and warehousing especially for items such as fresh products
that have extremely high requirements for the transportation and storage environment, so as to
avoid product decay and deterioration.
Figure 2.2-1 Ambient IoT devices in smart logistics and smart warehousing
More specifically, Ambient IoT can enable smart logistics and warehousing and improve efficiency and productivity by:
Batch and wide Ambient IoT tags support large numbers of simultaneous reading and wide range
range reading reading. When the goods arrive at a warehouse, the Ambient IoT tags attached to the
goods can be read in batch (e.g., thousands of tags per second) to obtain accurate
information about the goods, such as size/weight, manufacturer, expiration date,
serial number, production line, etc. The Ambient IoT tags attached to the goods or
containers in the warehouse will save their basic information and location information.
By setting up a central network node in the warehouse, all the goods in the ware-
house can be identified in a timely and rapid manner, which is convenient for manag-
ers to understand the inventory distribution/total amount of goods and predict the
corresponding storage demand.
Transportation When the goods with Ambient IoT tags are moving in the warehouse, network equip-
management ment can recognize and update the information of tags in time. Furthermore, network
equipment is able to obtain the position information of tags [1]. When the correspond-
ing goods need to be picked, the location of the goods can be quickly achieved within
the entire warehouse. In this way, the sorting efficiency of the goods is improved
significantly.
Cold chain Ambient IoT tags integrated with sensors are affixed in containers or on the surface of
transportation items for cold chain transportation. The Ambient IoT tags are used to monitor the tem-
perature, humidity or product status information in transportation, so as to detect
abnormal situations in time and ensure the quality and safety of fresh food.
Figure 2.2-2 Ambient IoT in code chain transportation
Typical requirements of smart logistics and smart warehousing
Requirements In most cases, Ambient IoT devices are electronic tags. Since Ambient IoT
for devices： devices are usually deployed in a large scale (e.g., each asset or device will
be pasted with a tag), the cost, size and power consumption need to be
considered. For monitoring conditions such as temperature and humidity in
transportation or storage through Ambient IoT tags, corresponding inte-
grated sensors are needed.
Power consumption：battery-free, maintenance-free.

Cost ： extremely low cost due to large scale deployment in warehouse,
integrated sensors are needed for environment monitoring.
Size：extremely small size for large scale application.
Coverage：tens of meters to hundreds of meters.
Work environment：Work well in special environments (e.g., high tempera-
ture, high pressure, extreme cold).
Requirements Deployment：Flexible deployment based on network infrastructure.

for network： Network equipment can be deployed on outdoor light poles or with digital
indoor system to provide fundamental coverage. Additional deployment can
be considered for dead zone or coverage enhancement.
Coverage：Single site needs to provide the coverage of >30m for indoor
scenario and >100m for outdoor scenario.
Security：Authorization-based reading and writing to protect privacy and
data security.
Capacity：Support sufficient system capacity and support a large number
of devices to read data (e.g., thousands of tags per second).
2.3 Smart home

Smart home takes the house as the platform, connects various devices in the home through the
Internet of things to build an efficient livable system. Smart home uses automatic control of house-
hold appliances, lighting control, temperature control, anti-theft and alarm control and other func-
tions and makes the home environment safer, more convenient and comfortable.
Ambient IoT can achieve battery-free (i.e., no need to charge), hence it can greatly increase the life of
the corresponding equipment in smart home and reduce maintenance costs. At the same time, it
can be deployed very flexible in smart home, such as embedded in walls, ceilings and furniture, or
affixed to keys, passports, clothes, shoes due to its ultra-low cost, extremely small size, washable,
flexible/foldable and other characteristics [2] . Based on the above advantages, Ambient IoT can
expand the application of smart home scenarios, and has great potential in the field of smart home.
PASS
Figure 2.3-1 Ambient IoT in smart home

Typical scenarios in smart home：
Item Finding A small, washable, flexible and foldable Ambient IoT device can be attached to some
easily lost items in the home, such as keys, passports, bank cards, wallets, etc. When
you need to find these items, you can quickly locate and find the lost items.
Environment Sensors can be integrated with Ambient IoT devices to monitor the temperature and
monitoring and alarm humidity of the house, and can also be used for emergency situations such as gas
leaks. The battery-free characteristic of Ambient IoT devices can increase the life of
the equipment and achieve maintenance-free;
Intelligent control Sensors can be integrated with Ambient IoT devices to achieve intelligent control of
home devices. For example, automatic control of washing machines, air conditioners,
televisions and curtains are executed based on the information collected by the
sensor. It is also possible to navigate home robots with tags embedded/attached to
doors or furniture and provide more refined control [3].
Typical requirements of smart home：
Requirements In most cases, Ambient IoT devices are electronic tags with integrated memory for
for devices： data access and/or integrated sensor for information acquisition. In smart home, the
cost, power consumption, size, waterproofness and foldability of Ambient IoT devic-
es ought to be considered.
Power consumption：battery-free, maintenance-free.

Delay ：tens of milliseconds to hundreds of milliseconds for intelligent control,
hundreds of milliseconds to thousands of milliseconds for indoor positioning.
Device Type：paper tag, anti-metal tag, foldable tag or washable tag.
Size：extremely small size for smart home application.
Coverage：tens of meters for indoor scenario.
Requirements Deployment：Smart terminal can be used as gateway or Ambient IoT devices

for network： connect to base station directly.
Coverage：10-30m for indoor scenario.
Security：Authorization-based reading and writing to protect privacy and data
security.
Number of connections：tens to hundreds.
Power charging：The existing wireless equipment (e.g., smart phone, CPE, AP)
can be regarded as the source of power. In this case, no additional equipment
for power supply is needed.
2.4 Indoor positioning

At present, the outdoor positioning technology is maturing rapidly, and GPS positioning or cellular
positioning technology has well met the positioning and navigation requirements in people’s daily life.
However, in regard to the increasing demand for indoor positioning and navigation, there is still a lack
of effective and low-cost technical solutions. For example, in daily life, shopping is one of the indis-
pensable activities for people. Large global shopping centers are constantly emerging, however, it is
often difficult for people to find target shops and restaurants quickly in such large shopping malls.
Another example is the underground parking, where people often can't find free parking spaces in
time and their tourism experience can be significantly diminished. In order to solve these problems,
we can provide indoor positioning and navigation services with the help of Ambient IoT technology,
helping customers find the required information and target places more efficiently.
Ambient IoT technology harvests the energy in the surrounding environment, such as light, heat or
RF signals, to enable devices that do not require batteries or use only capacitors to work for a
decade or even longer time. This technology is maintenance-free, low complexity, lightweight, and
low cost, making it ideal for high-density deployments.
Take a 200,000 m2 shopping mall as an example. It consists of one or more buildings with multiple
floors above and below ground, and the underground parking lot has hundreds to thousands of
parking spaces. In order to meet the indoor positioning accuracy requirements, indoor positioning
and navigation systems can be deployed in such shopping centers. Specifically, we can distribute
reference tags in a high-density, uniform manner on every floor and in every room of the shopping
center, for example, deploying one tag every 2 meters. In this way, indoor positioning can be realized
on handheld devices (such as smart mobile phone). These handheld devices can communicate with
reference tags to provide users with accurate navigation and positioning services.
Fig. 2.4-1 Application of Ambient IoT technology in indoor positioning

In addition, this indoor positioning and navigation system is not only widely used in shopping centers, but also in many
other application scenarios. In the field of intelligent manufacturing, this system can help to accurately locate the products
on the production line or delivery system and accurately grasp the production stage of the products. In the industrial field,
for those dangerous areas containing toxic and harmful substances to workers' health, the system can be used to monitor
workers' positions in real time by attaching tags to workers, and immediately issue safety alarms when they enter danger-
ous areas to ensure workers' safety. In addition, in terms of logistics and warehouse management, the system can also be
used for inventory and attendance inspection to help managers quickly locate items or persons and improve manage-
ment efficiency.
Fig. 2.4-2 Indoor positioning in smart manufacturing and warehousing
In general, the indoor positioning and navigation system based on Ambient IoT technology has a wide application pros-
pect and huge market potential. It will bring more convenient and intelligent experience to our life and work, and become
one of the important trends in future.
Typical use cases for Indoor Positioning based on Ambient IoT technology are as follows:
Underground In each parking space, wall or ceiling of garage， one or more Ambient IoT tags can be
parking navigation deployed as reference anchor points to build an underground parking positioning and
navigation system. Users can obtain more accurate positioning and navigation ser-
vices via handheld devices (such as smart mobile phone).
Shopping mall one or more Ambient IoT tags can be deployed in interior walls of each store in the
navigation shopping mall, and these tags can be used as anchor points for auxiliary position-
ing to construct the indoor positioning and navigation system. Users can select
their targeting stores in this system, obtain accurate positioning and navigation
services, and find the target location efficiently. In addition, the Ambient IoT tags
can push advertisements or promotional information of merchants to the custom-
ers, which enhances people’s shopping experiences.
Manufacturing Ambient IoT tags can be attached to products on production lines or delivery systems
and Logistics to locate items for easy management. When goods are in transit, storage, loading /
unloading and inventory, the Ambient IoT tags can be attached to the cargo package
to provide real-time location information to help track the location and movement of
goods.
Typical requirements of Indoor positioning
Terminal requirements： The form of an Ambient IoT device can be an electronic tag. When construct-
ing the positioning system, the number of Ambient IoT devices required is
relatively large, which should have the characteristics of maintenance-free,
low complexity, portability and low cost.
Tag power consumption：Battery free, no battery replacement, charging

and other related maintenance issues.
Tag type：Paper tag or anti-metal tag.
Tag Size：Extremely small size.
Tag Price：Extremely low cost, suitable for high-density deployment;
Network requirements： Coverage requirements ：10m to 30m.

Positioning accuracy ：horizontal accuracy 1-3 meters.
Energy Harvesting signal ：The smart device carried by the user is used as
the transmitting device of the energy harvesting signal for the Ambient IoT
tags. No additional energy harvesting signal is needed and the network layout
is simplified. Meanwhile, dedicated nodes for assisted energy supply can be
deployed.
Network security：Authorization-based tag reading for privacy and data
security.
2.5 Smart grid

With around 80 million kilometers of transmission and distribution lines worldwide, electricity net-
works are the backbone of secure and reliable power systems. As stated in the World Energy Out-
look 2020 [4], significant investment takes place in new network capacity between 2019 and 2030 as
a result of growing demand for electricity, the addition of new renewable generation capacity and
the need to develop smart grids. The expansion of electricity networks to 2030 is about 80% more
over the past decade. Around 30% of the increase in transmission lines and 20% of the increase in
distribution network lines are attributable to the increase of renewables. Over the next ten years,
around 16 million km of existing distribution lines and 1.5 million km of transmission lines need to be
replaced or digitalized, together with switching equipment, transformers, meters and other crucial
components. In regions with older power systems, such as the United States and the European
Union, roughly one-fifth of current networks need to be replaced or digitalized; this corresponds to
2.7 million km in the United States and 3.7 million km in the European Union. More than 60% of
global line replacements and new lines are in emerging market and developing economies, with
China alone accounting for a third of what is needed (over 7 million km).
Smart grids with wide use of IoT devices have a vital role to play in supporting the penetration of
variable renewables electricity sources. IoT has great potential in a wide number of applications in
the energy sector, i.e., in energy supply, transmission and distribution, and demand [5]. In particular
substations are a significant part of the electrical power grid. Through these stations, the voltage
level is converted from high voltage to low voltage using transformer. The substation transfers
power to distribution stations by the transmission lines (see Figure 2.5-1). Monitoring electrical sub-
stations are necessary to detect faults and resolve them, because if left unattended, it may lead to
electrical problems and cause long-term consequences. These problems not only cause energy
losses but also lead to electrical outages and losses in expensive equipment, in addition to injuries
and accidents such as fire. Therefore, monitoring of substations and their equipment is important to
ensure safety, protection, and stability in the electric power networks. Different types of sensors
(e.g. temperature and humidity sensors) can be used in the outdoor ultra-high voltage substation
(see Figure 2.5-2) to detect the anomaly and trigger predictive maintenance. In addition, various
sensors can be used in other use cases in the power transmission and distribution networks (see
Figure 2.5-3/4/5) for remote monitoring and protection purposes [6].
For these use cases, the data acquisition process is typically not latency-critical, but a large number
of sensors have to be efficiently connected, especially considering many of these sensors have
limited power sources and relatively frequent data transmission (every 5-15 minutes) is expected in
some cases. Moreover, lifespans of the field IoT devices are expected to be one decade or longer,
which is one of the main differences compared with consumer products. Many production systems
are subject to regulatory approvals (e.g., safety certification), changes to a running production
system often need to be avoided. Often sensors are deployed in locations that are inaccessible,
where physical replacement would be unduly expensive. Research continues to develop efficient
communication techniques to meet the requirements, among which Ambient IoT [7] is very promising
to enable wireless communication with minimum energy consumption. The Ambient IoT devices typ-
ically are battery-less or with limited energy storage capability, and obtain energy from the environ-
ment. The communication power consumption of such Ambient IoT devices is expected to be a few
hundred μW [8][9][10][11]. Moreover, communication service availability with sufficient 5G coverage is
important especially for remote monitoring of the critical equipment.
Primary Distribution
Industrial Consumption
Secondary
Distribution
Network
Transmission Substation
network
Power Primary Distribution Transformation

Generation Network Center
Figure 2.5-1 Smart grid overview

Figure 2.5-2 Outdoor ultra-high voltage substation Figure 2.5-3 Indoor/outdoor shielding cabinet
Figure 2.5-4 Underground transmission and distribution lines Figure 2.5-5 Aerial transmission and distribution lines
Typical use cases for Smart Grid based on the Ambient IoT technology are as follows:
Monitoring of high-voltage Problems such as electric leakage and tilt of transmission tower can be
transmission line detected by monitoring high-voltage transmission line.
Substation equipment Using temperature sensors to detect whether the temperature of substa-
inspection tion equipment is abnormal, preventing accidents such as equipment aging
and short circuits.
Asset inventory of power Through communication with Ambient IoT tags on power plant equipment,
plant equipment asset inventory of power plant equipment can be realized in a completely
contactless way.
Typical requirements for Smart Grid based on the Ambient IoT technology are as follows:
Terminal Tag power consumption：Battery free, no battery replacement,

requirements: charging and other related maintenance issues.
Communication delay：Smart Sensor Data Acquisition, Second Level.
Tag Size： Extremely small size for large scale applications.
Communication distance：10 ~ 30m indoors or 50 ~ 200m outdoors.
Connection number：Support thousands to tens of thousands of
device connections per square kilometer
Network Flexible deployment ： Use intelligent terminal as gateway device or directly con-
requirements： nect with base station.
Coverage requirements：indoor 10 ～ 30m and outdoor 50 ～ 200m.
Network security：Authorization-based tag reading for privacy and data security.
Connection requirements ： Support sufficient system capacity and data reading
of a large number of terminals
2.6 Smart agriculture

Ambient IoT devices can be used in smart agriculture to monitor the environment and control the
facilities such as the irrigation system.
A smart greenhouse for tomato planting is built in a farm. Some sensing Ambient IoT devices are
placed in the smart greenhouse to monitor the air temperature and humidity, carbon dioxide concen-
tration, light, soil temperature, humidity and PH. Some operation Ambient IoT devices are placed in
the smart greenhouse to control the window and irrigation system. A pico cell or a reader UE (with
subscription of Operator ) is placed in the smart greenhouse to communicate with the Ambient IoT
devices.
These Ambient IoT devices power themselves by harvesting energy from the environment (e.g. solar,
RF energy). The maximum power consumption for Ambient IoT device could be limited (e.g. several
hundred micro-watts) [9][10][11].
Considering the greenhouse environmental control, crop growth characteristics and economic ben-
efits, the optimal scale of greenhouse construction is 8 ~10 meters span, 80~100 meters length. The
size of the greenhouse could be very huge, it is reported recently that a single greenhouse area
reaches nearly 70,000 square meters, equivalent to ten standard football field size. To provide
all-round technical support for a greenhouse of nearly 70,000 square meters, a "super brain" is
needed. This computer monitors tens of thousands of sensors in the greenhouse, people can sit in
front of the computer and know everything going on in the greenhouse. For example, the tempera-
ture and humidity in each area of the 70,000 square meters greenhouse, the temperature of the
underground heating tube, how much the concentration of carbon dioxide is, whether the fan is
opened, and whether the nutrition is enough for each tomato.
Figure 2.6-1 A picture for huge greenhouse

Typical use cases for Smart Agriculture based on the Ambient IoT technology are as follows:
Soil moisture Controls whether the intelligent irrigation system increases or decreases the
monitoring amount of watering by monitoring the soil moisture.
Air temperature, humidity Control opening or closing of greenhouse ventilation pipes or control opening or
and carbon dioxide closing of greenhouse windows by monitoring temperature, humidity and
monitoring carbon dioxide concentration in the greenhouse.
Agricultural Based on the monitoring results, Ambient IoT can be used to control agricultural
facility control facilities for operations such as watering, fertilizing, and ventilation.
Typical requirements for Smart Agriculture based on the Ambient IoT technology are as follows:
Terminal Tag power consumption：Battery free, no battery replacement, charging

requirements: and other related maintenance issues.
Communication delay：Smart Sensor Data Acquisition, less than 1 second.
Tag Size：Extremely small size for large scale applications.
Communication distance：Up to 30m indoors.
Connection number：Support one device connection per square meter;
Process thousands to tens of thousands of devices per second by reading
device ID information.
Network Flexible deployment ：Use intelligent terminals as gateway devices or directly

requirements： connect with base stations.
Coverage requirements ：indoor 30m.
Network security：Authorization-based tag reading for privacy and data
security.
Connection requirements：Support sufficient system capacity and data
reading of a large number of terminals.
2.7 References
[1]Liu W, Huang K, Zhou X, et al. Next generation backscatter communication: systems, techniques,
and applications[J]. EURASIP Journal on Wireless Communications and Networking, 2019, 2019(1): 1-11.
[2]Maselli G, Piva M, Stankovic J A. Adaptive communication for battery-free devices in smart homes
[J]. IEEE Internet of Things Journal, 2019, 6(4): 6977-6988.
[3]J. J. Pomárico-Franquiz, Y. S. Shmaliy, Accurate self-localization in RFID tag information grids using
FIR filtering. IEEE Trans. Ind. Informat. 10(2), 1317–1326 (2014)
[4]International Energy Agency (IEA), "World Energy Outlook 2020". Available online:
https://www.iea.org/reports/world-energy-outlook-2020.
[5]Motlagh, N.H.; Mohammadrezaei, M.; Hunt, J.; Zakeri, B. "Internet of Things (IoT) and the Energy
Sector." Energies 2020, 13, 494.
[6]3GPP TR 22.840 V19.0.0. Study on Ambient power-enabled Internetof Things .
[7]Pereira F., Correia R., Carvalho N.B. "Passive Sensors for Long Duration Internet of Things
Networks." Sensors 2017;17:2268. doi: 10.3390/s17102268.
[8]https://en.wikipedia.org/wiki/Electronic_paper.
[9]F. Palacio, et al. (2006). "Active RFID tag with sensing capabilities and low power consumption".
[10]T. O. John, H. C. Ukwuoma, S. Danjuma and M. Ibrahim, "Energy consumption in wireless sensor
network", Energy, vol. 7, no. 8, 2016.
[11]Tuna, G., and V. C. Gungor. "Energy harvesting and battery technologies for powering wireless
sensor networks." Industrial Wireless Sensor Networks. Woodhead Publishing, 2016. 25-38.
Technical principles of Ambient IoT communication 03
TECHNICAL PRIN-
CIPLES OF AMBI-
ENT IOT COMMU-
NICATION
Ambient power enabled IoT devices should be mainly able of energy har-
vesting, ultra-low power consumption of communication and computing,
as shown in Figure 3-1, so as to achieve the advantages of battery-free
and maintenance-free. Due to the low energy density of ambient power,
the energy collected by ambient IoT devices is very limited. Therefore,
ultra-low power communication and computing technology are essential
to transmit, receive and process data. Additionally, ultra-low power com-
munication technology at least includes backscattering, ultra-low power
transmission, ultra-low power reception.
Lower Low
Energy power-consumption
harvesthing power-consumption
communication comptuing
03
reader tag
Figure 3-1 The technical schematic of Ambient power enabled IoT communication
Ambient energy harvesting

3.1 and storage
For ambient IoT devices, energy can be harvested from different types of ambient power sources.
Ambient power includes radio waves, solar energy/light, thermal energy, mechanical vibration,
airflow and acoustic noise, etc.
3.1.1 Ambient energy harvesting
RF energy One of the most important functions of power harvesting is to collect electromagnetic waves and
convert radio frequency energy into direct current (RF-DC). In Ambient IoT communication, the col-
lected energy is used to drive the load circuit (low-power computing, sensors, etc.) to achieve bat-
tery-less communication.
Due to its controllable energy output and availability (e.g., RF transmitters can send power on
demand or at regular intervals), RF-based energy harvesting used to apply to logistics/warehousing,
manufacturing, smart home, health monitoring, etc.
There are several challenges regarding RF power harvesting.
1) It is difficult to collect RF energy in a wireless environment due to the low power density (e.g., less
than 10nW/cm2). The RF power that can be effectively collected shall exceed a certain input power
threshold, which can be called as RF power harvesting sensitivity of such device.
2) In order to drive logic circuits or chips, DC voltage converted from RF energy shall meet the mini-
mum output voltage requirements. It remains a big challenge to efficiently convert RF to DC under
the condition of very low input power.
3) Intelligently managing the collected or stored energy is also important for a good balance of com-
munication and computing.
Currently, it shows by results from experimental researches that the RF energy conversion efficien-
cy is different for different input power and energy harvesting circuit designs. For example, the
energy conversion efficiency at input power of -20dBm is often less than 10% while the conversion
efficiency at input power of -1dBm is close to 50%. When input power is less than -30dBm it is very
challenging to effectively collect RF energy and rectify it into a usable DC voltage.
Generally, the power required to drive ultra-low power circuit is at least 10uw. In order to meet
requirements of the basic low-power computing and communication, it can be seen that improving
the efficiency of energy collection and conversion under the condition of low input power is one of
the most important tasks in the research and development of ambient-power enabled IoT communi-
cation system [1].
Table 3.1-1 Input power vs. RF energy conversion efficiency [1]
Efficiency（%） Input power（dBm） Center frequency（MHz） Reflector unit
1.2 -14 950 0.3-μm CMOS convertor
5.1 -14.1 920 0.18-μm CMOS convertor
10 -22.6 906 0.25-μm CMOS convertor
11 -14 915 90-μm CMOS convertor
12.8 -19.5 900 0.18-μm CMOS , CoSi2 - Si Schottky
13 -14.7 900 0.35-μm CMOS convertor
18 -19 869 0.5-μm CMOS convertor
26.5 -11.1 900 0.18-μm CMOS convertor
47 -8 915 0.18-μm CMOS convertor
49 -1 900 Skyworks SMS7630 Si Schottky
The research of power harvesting circuits has gone through many years of development and explora-
tion, improving efficiency has always been the most concerned issue in circuit design. For RF-DC con-
version, the circuit designs have obvious impact on the efficiency. The proper use of the rectifier can
well convert the radio frequency energy into a stable direct current voltage (RF-DC). If the output volt-
age is low, further direct current conversion boost (DC-DC) is required. Voltage regulators and voltage
monitors are also commonly used to help boost and stabilize the output voltage. Diode-based rectifier
circuits are the most basic method for energy harvesting. And CMOS-based devices that usually
requires input power less than -20dBm can have better performance than discrete devices.
The typical power harvesting circuits include half-wave rectifier (as shown in Figure 3.1-1), single shunt
rectenna, single stage voltage multiplier (as shown in Figure 3.1-2), Cockcroft-walton/Greinacher
charge pump, Dickson charge pump, modified Cockcroft-walton/greinacher charge pump).
D1 C1 D2
RF Cload Rload Vout RF Cload Rload Vout
Figure 3.1-1 Half-wave rectifier Figure 3.1-2 Single stage voltage multiplier
Solar Energy Solar power/light can be transformed into electrical power using photovoltaic cells and it uses
/Light photovoltaic effect for energy harvesting with conversion efficiency of 10-40% [2] . For the out-
door case, solar energy is one of the most common ambient power, it can supply inexhaustible
clean energy and has high power density of up to 100 mW/cm2 [3] .
Solar power is unstable, inconsistent, and intermittent. It is highly dependent on the atmo-
spheric condition, surrounding obstructions, etc. It is available during daytime but inefficient on
a cloudy day or during the night. Solar energy harvesting can be mainly used for outdoor envi-
ronmental monitoring, agriculture, husbandry, transportation, etc. For the indoor cases, light
from the lighting equipment can be used. Although the power density is lower than solar, e.g.,
100uw/cm2, it is much stable and controllable. Energy harvested from light can be used for
manufacturing, indoor environmental monitoring etc.
I PV
ID I SH RS
IL R Sh V PV
Figure 3.1-3: The equivalent electrical circuit of a single diode solar PV cell [4]
Thermal Thermal energy is another ambient power source that are available for lots of use cases. Elec-
Energy trical power is directly generated by exploiting the temperature difference in thermoelectric
devices taking advantage of thermoelectric effects, such as the Seebeck effect or the Thom-
son effect. Thermoelectric generators have low efficiency (only about 5–6%) [5] . The power
density is 25~1000uw/cm2 depending the environment condition.
Although with low conversion efficiency, thermal energy can be used in many outdoor applica-
tions or indoor cases as long as temperature difference or temperature fluctuation can be
expected in the environment. For example, outdoor environmental monitoring, smart grid, agri-
culture, husbandry etc.
Heat source
Heat sink
Figure 3.1-4: Seebeck effect

Mechanical The piezoelectric effect generates electrical voltages or currents from mechanical strains,
Vibration such as vibration or deformation. Typical piezoelectric-based energy harvesters keep creating
power when there is a continuous mechanical motion, such as acoustic noises and wind, or
they sporadically generate power for intermittent strains, such as human motion (walking,
clicking a button, etc.). The volume of the piezoelectric power generators is relatively small and
typical output power density values of usual piezoelectric materials are around 250 μW/cm3
but they can create more power when a motion or deformation is intense [6][7] .
Clamped end
Piezoelectric layers Seismic mass
Electrodes
Shim layer
Figure 3.1-5: Piezoelectric energy harvesting generator [8]
From the discussion above, it can be seen that kinds of ambient power have the following
characteristics:
For typical ambient power, it can be observed the power harvested is very limited, e.g. from
1uW to 100mW (per cm2/cm3).
For some ambient power from artificial power source (e.g., light, RF waves), the power can be
stable and constant. But for some other kind of ambient power such as solar, heat or vibration,
the ambient power will be unstable (intermittent, not constant). It is impossible to use the am-
bient power as a direct power source for electronic devices.
Table 3.1-2 the comparation of energy sources
Energy source Type Typical power Note
Outdoor solar light Natural 100 mW/cm2 (outdoor) Uncontrollable but predictable
100 μW/cm2 (artificial light)

Indoor office light Artificial/natural Partially controllable
10 mW/cm2 (filtered solar light)
Artificial 0.1 μW/cm2 –10 μW/cm2 Controllable and predictable

Ambient
radio 0.001 μW/cm2 (WiFi)
frequency Artificial/natural Uncontrollable and
unpredictable
0.1 μW/cm2 （GSM)
Thermoelectric Artificial 60 μW/cm2 Partially controllable
4 μW/cm3 (human motion)

Vibration Uncontrollable
Artificial
but predictablz
800 μW/cm3 （machines)
Uncontrollable
Ambient airflow Natural/artificial 1 mW/cm2
and unpredictable
Uncontrollable
Acoustic noise Natural/artificial 960 nW/cm3
and unpredictable
3.1.2 Ambient energy storage
Based on the analysis in the previous sections, it can be seen that at least for some ambient IoT
devices, energy storage components are required due to the followings:
• The energy storage component can stabilize and control the power output, and can smooth the
power fluctuations of the collected ambient energy through peak shaving and valley filling.
• It is able to collect weak harvested power (e.g., microampere or even nanoampere level) and pro-
vide higher peak discharge current (e.g., tens of microampere to hundreds of microampere level)
for the ambient IoT devices.
There are many types of storage components suitable for ambient energy IoT devices, e.g., capaci-
tors and supercapacitors. Energy storage components make more types of ambient energy possible
for ambient energy IoT devices.
Radio
Harvesting Power
Sensors
functionality Management
Processor
Energy Storage
Figure 3.1-6 One example of Ambient IoT device Implementation
Capacitor can be considered as the basic energy storage elements for ambient IoT devices, in case
that power sources are stable and constant. Capacitor have limited power perseverance time and
storage capacity, which can restrict the ambient IoT application. For example, with a fully charged
capacitor of 24uF, it can drive the ambient IoT devices for 3.6k bits communication (1.5V, 10uA and
1kbit/s are assumed).
Printed solid-state batteries offer durability and higher capacity, making them a great additional stor-
age option. Taking a 1μAh, 1.5V solid-state battery as an example, it can even drive surrounding IoT
devices for 360kbit communication (assuming a circuit voltage of 1.5V and a current of 10μA).
When selecting energy storage components for ambient IoT devices, it is necessary to comprehen-
sively consider the two aspects of "source" and "consumption" in the real use case, at least including
the energy collection status, the consumption requirements of communication, computing, sensor
and other components of the device., etc. Since different data rates and Tx powers are required for
ambient IoT devices, the transitions between different operating modes such as energy harvesting,
dormant state or active state, are also very important.
ep tive ep tive ep tive ep

V sle ac sle ac sle ac sle
P (active)
P (sleep)
T
Figure 3.1-7 Examples of different operating modes for ambient IoT devices
Ultra-low power
3.2 communication
3.2.1 Backscatter transmitter
The backscatter technology enables signal transmission without an active transmitter. Similar as
radar technology, a part of electromagnetic waves will be reflected when they reach the surface of
an object. The strength of the reflected signal depends on the shape, material and distance to the
object. From the perspective of a radar, each object has its radar cross section (RCS) [9] . A tag
achieves signal modulation by changing its backscatter factor, which is equivalent to its RCS. The
backscatter transmitter modulates the reflected RF signal to transmit data without generating the
radio-frequency (RF) signal itself.
Backscatter was first proposed by Stockman in 1948 [10]. However, traditional backscatter communi-
cation cannot be widely used in data-intensive wireless communication systems due to the following
limitations.
1)Traditional backscatter communication requires the placement of the backscatter transmitter near
its RF source, thereby limiting the use and coverage area of the device.
2)In traditional backscatter communication, the receiver of the backscattered signal and the RF
signal transmitter are located in the same device, i.e., the reader, which results in self-interference
between the receiving and transmitting antennas, thereby degrading communication performance.
3)A backscatter transmitter passively responds only when inquired by a reader.
Recently, Ambient Backscatter Communication (AmBC) [11] has become a more promising technolo-
gy for low-power communication. It can effectively address the aforementioned limitations in tradi-
tional backscatter communication systems, making AmBC technology more widely adopted in prac-
tical applications.
The AmBC system typically consists of three parts: the ambient RF source, the backscatter device
(BD), and the reader. In AmBC systems, backscatter devices can communicate with each other by
utilizing wireless signals broadcast by ambient RF sources such as TV towers, FM towers, cellular
base stations, and Wi-Fi access points (APs). By separating the carrier transmitter and the backscat-
ter receiver, the number of RF components in the backscatter device is minimized. Furthermore, the
backscatter device can operate actively, i.e., the backscatter transmitter can send data without the
trigger of a reader when it has harvested sufficient energy from the RF source.
Backscatter Backscatter
transmitter receiver
Ambient RF
source
Legacy
receiver
Figure 3.2-1 Illustration of AmBC system [12]

In a typical backscatter communication process, Ambient IoT devices (such as backscatter tags)
receive carrier signals sent by readers and harvest energy through RF energy harvesting modules,
which is used to power the low-power processing modules. After obtaining energy, the backscatter
tags drive corresponding circuits to modulate incoming signals and perform backscattering.
Input carrier
Reader
Backscattered Signal
TX AMP
Energy
Logic
harvesting
RX LNA
Figure 3.2-2 Backscattering communication
In the backscatter communication system, load modulation is a common method used by tags to
transmit data. Load modulation adjusts the electrical parameters (such as resistance or capaci-
tance) of the tag's oscillating circuit according to the data stream, thereby changing the size and
phase of the tag's impedance, completing the modulation process.
There are two main types of load modulation techniques: resistive load modulation and capacitive
load modulation. In resistive load modulation, a resistor is connected in parallel with the load, called
the load modulation resistor. This resistor is turned on and off according to the clock of the data
stream, and the switch S is controlled by binary data encoding. In capacitive load modulation, a
capacitor is connected in parallel with the load instead of the load modulation resistor controlled by
binary data encoding.
R1 R2
S
L2 L1
C1 RL
C2
Figure 3.2-3 Resistance-based modulation

Taking resistance-based ASK modulation as an example, the terminal can switch between absorp-
tion state and reflection state by adjusting the load reflection coefficient. In the absorption state, the
terminal achieves impedance matching thus the input RF signal is completely absorbed by the
terminal. Hence, the signal received by the reader will be at low-level, which indicates a bit '0'. On the
contrary, in the reflection state, the terminal adjusts the circuit impedance that leads to a mismatch
of the impedance thus a part of the RF signal is reflected. Then the signal received by the reader will
be at high-level, indicating a bit '1'.
As shown in Figure 3.2-4, the terminal can implement ASK modulation in a simple way of impedance
switching. From the perspective of the receiver, ASK signals can be detected with low-complexity
envelope detector and comparator.
orm
avef
ier w
Carr
Load-based modulator antenna
load 1
load 2
1 0 0 1 1
Figure 3.2-4 ASK modulation
Similarly, the terminal can also change the response frequency of the circuit by adjusting the capaci-
tance of the circuit to implement FSK modulation. FSK has better BER performance than ASK. It is
often used to realize frequency division multi-access.
Therefore, backscatter communication cleverly utilizes impedance modulation to achieve extremely

low-complexity signal modulation and transmission. In contrast, backscatter devices do not require
complex RF components such as, high-precision crystal oscillators, duplexers, or high-precision
filters. They also do not need complex baseband processing. For example, envelope detection of the
signal can be used without the need for complex channel estimation and equalization operations.
Therefore, backscatter technology enables the realization of ultra-low complexity backscatter devices.
In the recent Ambient IoT study project carried out by 3GPP, different types of devices are
discussed [13]. For the devices supporting backscatter transmission, one type of device has no energy
storage or power amplifier, and only supports backscattering without being able to amplify the back-
scatter signal. Another type of device has certain energy storage and power amplifier and can use
the energy stored through energy harvesting to drive the power amplifier to amplify the backscatter
signal to enhance the coverage [14].
The typical topologies of Backscatter communication (BC) systems can include the following types: MoBC(Monostatic
backscatter communication), BiBC(Bistatic backscatter communication), and AmBC, as shown in the figure below [15]：
Carrier Signal
Backscattered Sig
nal
Reader Tag
(a) MoBC
Carrier Signal
Emitter Reader
Tag
(b) BiBC
Carrier
Signal
Reader
Tag
Emitter
(c) AmBC
Forward: Emitter-to-Tag Backscatter: Tag-to-Reader
Figure 3.2-5 Typical topologies of Backscatter communication systems
MoBC The topology includes one reader and one tag (Figure 3.2-5(a)), where the reader provides the carrier
for the tag's backscatter and receives the tag's backscattered signal. In this topology, the reader needs
to support full duplex with the capability to suppress interference between transmit and receive signals,
posing a higher demand on the reader. In order to prevent interference between the transmitted and
received signals, the power of the reader's received signal cannot be too weak compared to the power
of the transmitted signal. Additionally, the activation power threshold of the tag is relatively high, which
results in a shorter coverage distance for MoBC, making it suitable only for short-range applications.
BiBC The topology includes one reader, one tag, and a dedicated RF carrier transmitter (Figure 3.2-5(b)). The
dedicated RF carrier transmitter can be deployed close to the tag, solving the bottleneck of downlink
coverage. By separating the transmitter and receiver, the reader does not need to support full duplex.
Thus the issue of interference between transmit and receive signals can be bypassed and uplink cover-
age can be improved. Additionally, optimized deployment of the dedicated RF carrier transmitter can
increase the power of the carrier and the efficiency of energy harvesting by the tag, thereby increasing
the power of the backscattered signal and improving uplink coverage.
AmBC Compared to BiBC, this topology utilizes existing RF signal transmitters in the environment, such as
cellular base stations, to provide the carrier for the tag, eliminating the need for deploying a dedicated
RF signal transmitter. This reduces deployment costs and power overhead. However, because the am-
bient RF signals are uncontrollable and modulated signals, they can cause direct interference to the
reader's reception. The uncontrollability of ambient RF signals limits the performance of AmBC sys-
tems, imposing additional design and implementation challenges.
3.2.2 Active signal transmitter
For Ambient IoT devices that transmit signals through backscattering, their power consumption can
be as low as a few μW to tens of μW. Energy harvest circuit has certain requirements on the strength
of RF signals, generally not less than -20dBm (10μW). This limits the coverage of carrier transmission
or wireless energy supply. If Ambient IoT devices have certain energy storage capacity (such as
equipped with energy storage components), the RF signal strength can be relaxed to -30dBm. Then,
Ambient IoT devices can store energy for use during operation through long time energy harvesting.
However, under the circumstances where the power of the energy supply signal is limited, the cover-
age of energy supply signal from network equipment is still relatively small, generally up to tens of
meters. The power of the backscattered signal also depends on the power of the carrier signal. There-
fore, due to the influence of downlink coverage, the uplink coverage of Ambient IoT devices that adopt
backscatter communication is also very limited. Additionally, as mentioned in the previous section,
since uplink signals and downlink carrier signals are typically in the same channel or frequency band in
general, there is also a serious issue of uplink-downlink duplex interference.
In recent Ambient IoT study projects in 3GPP and IEEE 802.11 Working Group (WG) [13], different types
of Ambient IoT devices were explored. Besides Ambient IoT devices that transmit signals through
backscatter, Ambient IoT devices with active signal transmission capabilities can also be supported to
improve uplink coverage or solve the duplex interference problem. These Ambient IoT devices gener-
ally have large capacitors to store energy from the environment, and their power consumption can
support hundreds of μW to several mW. Their transmitters have components such as oscillating
circuits, signal amplification circuits, and RF transmit units that are required for traditional transmitters.
To achieve active signal transmission with extremely low power consumption,the following tech-
nical challenges need to be addressed:
1） To achieve extremely low-power signal transmission, an extremely low-power oscillator is

required to generate the RF carrier instead of a traditional crystal oscillator. This is because the
power consumption of a crystal oscillator is generally several mW [16] , which cannot meet the
extremely low-power requirement of Ambient IoT devices.
2） The RF carrier generated by an extremely low-power oscillator typically has poor accuracy and
requires the development of corresponding frequency synchronization or frequency offset compen-
sation technologies.
3） The output signal power of low-power active transmitters is -20dBm to -10dBm, which signifi-
cantly increases the coverage of the UL signal compared to the backscatter case (considering the
reflection loss, the backscattered signal power is generally -35dBm to -25dBm or lower) [see Sec-
tion 4.1], meeting the coverage requirements in most scenarios. Such a low output power level is
advantageous in terms of controlling the total peak power consumption of Ambient IoT devices.
However, it also presents new technical challenges. This is because at the bias voltage levels com-
monly used today (e.g., 1.5V), a low transmit output power actually reduces the energy efficiency of
the entire transmit circuit. Therefore, corresponding techniques to enhance the efficiency of trans-
mission circuits are also required.
Ambient IoT devices with active signal transmission capabilities can independently send signals
through the collection and storage of ambient energy, without relying on carrier signals provided by
other devices. Their transmitters' amplifiers can increase the signal transmission power, unlike in the
case of backscatter transmission, which is limited by the power of the carrier signal. The signal receiv-
ing end devices, such as the receivers of network equipment, can achieve a sensitivity level of
-100dBm to -110dBm. These advantages can greatly improve the coverage of uplink signals, especial-
ly suitable for applications that require large uplink coverage. At the same time, for scenarios that
require single-node reception (such as scenarios where smartphones communicate with Ambient IoT
devices), it also avoids the problem of full duplex interference in the receiver in case of backscatter.
3.2.3 Ultra-low power receiver
Besides low power transmitter, Ambient IoT devices also need to consider low power consumption
and low complexity of receiver to meet the overall power consumption target of Ambient IoT devic-
es. The typical receiver architecture of Ambient IoT devices includes RF envelope detection, hetero-
dyne intermediate frequency (IF) envelope detection, and zero-IF baseband envelope detection
architectures, as shown in the following figures:
Matching Comparator/ BB Digital

Network ADC processing
RF BPF LNA RF Envelope BB AMP BB LPF

Detector
Figure 3.2-6 Architecture with RF envelope detection

RF BPF IF AMP IF Envelope BB AMP BB LPF

Detector
LO
Figure 3.2-7 Heterodyne architecture with IF envelope detection

RF BPF BB AMP BB LPF
LO
Figure 3.2-8 Zero-IF architecture with baseband envelope detection
The RF envelope detection receiver receives the amplitude-modulated RF signal through a match-
ing network and RF filter. It directly converts the Amplitude modulation (AM) RF signal to a base-
band signal through RF envelope detection. This detection method does not require the receiver to
have a local oscillator or phase-locked loop (PLL), and can achieve extremely low power consump-
tion with low-precision analog to digital converters (ADCs) and optional amplifiers. Therefore,
among the three receiver architectures mentioned above, the RF envelope detection receiver has
the lowest power consumption, e.g., a few μW. The receiver needs to use high-Q matching networks
and RF filters to suppress adjacent channel interference. However, due to complexity and power
consumption, RF analog filters with weak filtering capabilities based on matching networks and
filters can not receive the target signal only and reject wide interference signals, resulting in a wide
noise bandwidth and poor receiver sensitivity of around -40dBm. The RF envelope detection receiv-
er has the highest noise figure among the three receiver architectures, generally ranging from 12dB
to 22dB.
The heterodyne IF envelope detection receiver first converts the filtered RF signal through a mixer
to an IF signal, and then performs envelope detection on the IF signal to convert it to a baseband
signal. This detection method requires the receiver to have a local oscillator for mixer operation and
a corresponding PLL. The added local oscillator and PLL are the main sources of power consump-
tion. Power consumption can be reduced by reducing the precision and stability requirements of the
local oscillator, using a frequency-locked loop (FLL) instead of a PLL, and employing low-precision
ADCs. Among the three receiver architectures mentioned above, the heterodyne IF envelope
detection receiver has the highest power consumption, generally exceeding 100μW. It is more effec-
tive for heterodyne IF envelope detection receiver to use IF filters to suppress adjacent channel
interference, compared to matching networks and RF filters. It also has RF and IF amplifiers that can
improve receiver sensitivity. The sensitivity of the receiver architecture can achieve a sensitivity of
-90dBm to -80dBm. The heterodyne IF envelope detection receiver has the lowest noise figure
among the three receiver architectures, generally ranging from 9dB to 15dB.
The zero-IF baseband envelope detection receiver converts the filtered RF signal directly to a base-
band signal through mixing. This detection method also requires the receiver to have a local oscilla-
tor and PLL. The methods to reduce power consumption are similar to those of the heterodyne IF
envelope detection receiver. Among the three receiver architectures mentioned above, the power
consumption of the zero-IF baseband envelope detection receiver is slightly lower than that of the
heterodyne IF envelope detection receiver. The zero-IF baseband envelope detection receiver uses
high-Q matching networks, RF filters, and baseband filters to suppress adjacent channel interfer-
ence. Its receiver sensitivity is slightly lower than that of the heterodyne IF envelope detection
receiver. The noise figure of the zero-IF baseband envelope detection receiver is slightly higher than
that of the heterodyne IF envelope detection receiver, generally ranging from 10dB to 16dB.
Low-power
3.3 computing
The main characteristics of Ambient IoT communication is to realize low-power communication by

modulating the incoming carrier waves. At the same time, it can also drive digital logic circuits
through RF power harvesting to achieve signal encoding, encryption or calculation.
As mentioned in section 3.1， the conversion efficiency of RF energy is often less than 10%, which
means that the power required to drive the digital logic circuits cannot be too high. Figure 3.3-1
shows the number of computing times that 1 microjoule of energy can support. Although with the
improvements of the material and optimizations of designs, executions per microjoule can be greatly
improved, complex computation using very limited energy remains challenging.
COMPUTATIONS PER MICROJOULES
10 5
LA
PT
OP
(
10 4 DE SIC with
LL OR out
DIM TE sc
GA EN X S reen
TE TIO C5 )
WA 83
10 3 YP N2 2
3( 7 40
33 0
MH
DE z)
10 2 IBM LL
PS O P
/2 T I PL
INT E& EX
SU GX
10 1 EL NS I
48 S1
6/ 00
25
1computation per 1μj &4 0
10 0 MA 86
/3
CT 3
NT
IBM OS
H1
10 -1 CO PC IBM 28K
,IB
MM M
OD PC PC-X
OR -XT T
10 -2 E6
CR 4,AP
AY PL
DE I EI
CP Ie
10 -3 DP AL
-11 TAIR
/2
0
10 -4 SD
S9
20
10 -5
UN
IVA
CI
10 -6 II
10 -7
UN
IVA
10 -8 CI
UN I
IVA
CI
10 -9 ED
VA
EN C
IAC
10 -10
10 -11
1940 1950 1960 1970 1980 1990 2000 2010
Figure 3.3-1 Computation develops with low power consumption[17]

In order to design the Ambient IoT communication system, low-power computing is usually
considered from the following aspects:
Low power- Low power-consumption chips generally include MCUs and sensors. There are minimum input
consumption voltage requirements for circuits that drive digital processing chips. As the power consumption
chip is in the order of microwatt for most of the MCUs available in the markets, the harvested
energy is usually not sufficient to support complicated communication and computations.
Therefore, it is critical to select MCUs and other active components that meet the power
budget of the whole system.
Low power User privacy and data security shall be considered in many Ambient IoT application scenarios.
security
On the one hand, it is strongly concerned by customers to protect information such as identifi-
cation, location etc. For example, in logistics scenarios, it is necessary to write or read item
information of Ambient IoT devices; in home scenarios, the location information of personal
items is not expected and should not be exposed to non-trusted devices. In order to meet spe-
cific requirements for different scenarios, Ambient IoT devices need to support low-power
security mechanisms to ensure the confidentiality and reliability of data communication.
On the other hand, due to the low complexity and low power consumption of Ambient IoT
devices, whether existing security algorithms (e.g., AES128 encryption and decryption algo-
rithm commonly used in WiFi) can be reused needs to be re-evaluated.
Assuming the following baseband architecture for AES128 [18], using a 65nm processing chip
and a data packet length of 128 bits, the power consumption of the AES module and the entire
chip are 3.8 μW and 6.5 μW, respectively. The AES128 security encryption algorithm brings
microwatt-level power consumption, which makes it possible for Ambient IoT devices.
In addition to AES128, other low-power security algorithms, such as AES 256 and Zu Chongzhi
algorithm, are also mainstream choices. In some implementations, random number generators
are also introduced.
CLK-in Clock management
Data-in Decoding AES
CRC Control
Buffer
Data-out Manchester coding EEPROM
Figure 3.3-2 Example of baseband architecture for evaluating AES128 algorithm [18]
Simple coding ASK (including OOK) and FSK can be used as the basic modulation schemes. Simple coding
and modulation schemes such as non-return to zero (NRZ) coding, Manchester coding, unipolar return to zero
coding, differential bi-phase (DBP) coding, miller coding, pulse interval coding (PIE) and other
coding methods can be considered. Overall, the use of simple coding and modulation can
greatly reduce the power consumption of a Ambient IoT system.
3.4 References
[1]Valenta C R, Durgin G D. Harvesting wireless power: Survey of energy-harvester conversion
efficiency in far-field, wireless power transfer systems[J]. IEEE Microwave Magazine, 2014, 15(4):
108-120.
[2]Kim S, Vyas R, Bito J, et al. Ambient RF energy-harvesting technologies for self-sustainable

standalone wireless sensor platforms[J]. Proceedings of the IEEE, 2014, 102(11): 1649-1666.
[3]Green M, Dunlop E, Hohl Ebinger J, et al. Solar cell efficiency tables (version 57)[J]. Progress in
photovoltaics: research and applications, 2021, 29(1): 3-15.
[4]Mishu M K, Rokonuzzaman M, Pasupuleti J, et al. An adaptive TE-PV hybrid energy harvesting

system for self-powered iot sensor applications[J]. Sensors, 2021, 21(8): 2604.
[5]Prauzek M, Konecny J, Borova M, et al. Energy harvesting sources, storage devices and system
topologies for environmental wireless sensor networks: A review[J]. Sensors, 2018, 18(8): 2446.
[6]Kim S, Vyas R, Bito J, et al. Ambient RF energy-harvesting technologies for self-sustainable

standalone wireless sensor platforms[J]. Proceedings of the IEEE, 2014, 102(11): 1649-1666.
[7]H. S. Kim, J. -H. Kim, and J. Kim, ‘‘A review of piezoelectric energy harvesting based on vibration,’’
Int. J. Precision Eng. Manuf., vol. 12, no. 6, pp. 1129–1141, Dec. 2011.
[8]Hoang T, Ferin G, Bantignies C, et al. Aging assessment of piezoelectric energy harvester using
electrical loads[C]//Journal of Physics: Conference Series. IOP Publishing, 2019, 1407(1): 012078.
[9]Nikitin P V, Rao K V S, Martinez R D. Differential RCS of RFID tag[J]. Electronics Letters, 2007, 43
(8): 431-432.
[10]H. Stockman, "Communication by means of reflected power", Proc. IRE, vol. 36, no. 10, pp.
1196-1204, Oct. 1948.
[11]Van Huynh N, Hoang D T, Lu X, et al. Ambient backscatter communications: A contemporary

survey[J]. IEEE Communications surveys & tutorials, 2018, 20(4): 2889-2922.
[12]Lu X, Niyato D, Jiang H, et al. Ambient backscatter assisted wireless powered communications
[J]. IEEE Wireless Communications, 2018, 25(2): 170-177.
[13]3GPP TR 38.848 V18.0.0 (2023-09), Study on Ambient IoT (Internet of Things) in RAN.
[14]F. Amato et. al., “Tunneling RFID Tags for Long-Range and Low-Power Microwave Applications”,
IEEE Journal of Radio Frequency Identification, vol. 2, no. 2, pp. 93-103, June 2018.
[15]D. Galappaththige, F. Rezaei, C. Tellambura, S. P. Herath, ‘‘Link Budget Analysis for

Backscatter-Based Passive IoT,’’ IEEE Access, vol. 10, pp. 128890-128922, 2022.
[16]Yu-Sheng Lin; Ching-Wen Hsu; Chun-Lin Lu; Yeong-Her Wang, A Low-Power Quadrature Local
Oscillator Using Current-Mode-Logic Ring Oscillator and Frequency Triplers, IEEE Microwave and
Wireless Components Letters ( Volume: 23, Issue: 12, December 2013).
[17]Radio-Frequency Rectifier for Electromagnetic Energy Harvesting: Development Path and

Future Outlook, Simon Hemour; Ke Wu, Proceedings of the
IEEE;2014;102;11;10.1109/JPROC.2014.2358691.
[18]IEEE 802.11-23/1189r0, Discussion on AMP Security.

31 Overall design of Ambient IoT communication system 04
OVERALL DESIGN
OF AMBIENT IOT
COMMUNICATION
SYSTEM
For different applications, Ambient IoT can use different frequency bands.
Different network deployment can be utilized for different communication
requirements. Coexistence with existing communication systems should
also be considered.
04
Overall design of Ambient IoT communication system 38
Frequency bands and link

4.1 budget of Ambient IoT
In the deployment of an Ambient IoT system, its communication frequency band should be appropri-
ately considered first. Selecting different frequency bands will directly affect the design of the over-
all architecture of the Ambient IoT system. In general, both unlicensed and licensed frequency bands
can be used for Ambient IoT.
For operation in unlicensed frequency bands, the spectrum resources can be used freely and flexi-
bly as long as the operation complies with the regulatory requirements. Typical unlicensed frequen-
cy bands are globally uniform. Therefore, unlicensed frequency bands can be used to reduce the
developing and operating costs and expand the application scenarios of Ambient IoT. According to
the latest progress of IEEE 802.11 AMP (Ambient Power enabled-IoT) Study Group (SG), IEEE 802.11
has considered standardizing Ambient IoT in the two unlicensed frequency bands of Sub-1 GHz and
2.4GHz.
For operation in licensed frequency bands, on the one hand, the spectrum resources of existing
operators can be fully utilized, and the maximum transmission power on the licensed frequency
band is relatively high, which facilitates achieving seamless cellular coverage and long-distance com-
munication. On the other hand, operators can reuse existing widely-deployed cellular network infra-
structure and cellular smart terminals to achieve rapid deployment and wide-area coverage for Am-
bient IoT. According to the 3GPP RAN Rel-19 Ambient IoT SID, Ambient IoT deployed on the licensed
FDD (Frequency Domain Duplex) band will be studied in Rel-19.
4.1.1 Licensed frequency bands
Corresponding specification requirements need to be followed for licensed frequency bands. The
basic requirement is to comply with local law and regulation to ensure legal, compliant, stable and
reliable use of spectrum resources. In addition, it is important for a new communication system
deployed on the licensed frequency band to protect existing communication services. The impact
on the existing communication system needs to be fully considered to avoid interference to other
systems when designing the Ambient IoT system.
From the perspective of spectrum deployment there are three modes for Ambient IoT, including
in-band deployment, guard-band deployment and standalone deployment. Different deployment
modes represent different spectrum location in the licensed frequency band. No matter which
deployment mode is adopted for Ambient IoT, it is necessary to meet the basic RF requirements of
the licensed frequency band and consider coexistence with the traditional cellular system. For
example, if guard-band deployment is adopted for Ambient IoT, since the division of the guard band
is to avoid mutual interference between devices deployed in adjacent frequency bands, potential
interference to adjacent frequency bands needs to be evaluated. The interference becomes more
severe when Ambient IoT downlink (DL) signal is used to provide wireless power to Ambient IoT
devices as the DL signal needs to have sufficiently high transmission power. For another example, if
in-band deployment is adopted for Ambient IoT, the existing cellular transmission may cause inter-
ference to the reception of Ambient IoT devices. Similarly, the Ambient IoT uplink (UL) transmission
also potentially causes interference to the legacy system due to the weak filtering capability. In a
word, the coexistence between Ambient IoT and the traditional cellular system needs to be further
studied. For relevant detailed analysis, please refer to the following section 4.4.
4.1.2 Unlicensed frequency bands
The related frequency regulation should be satisfied when wireless communication

equipment works on an unlicensed frequency band. In the USA, the relevant frequency
regulation is given in FCC 15.247. In Europe, wireless systems should follow the “ETSI
Harmonized Standards” . Theoretically, following these harmonized standards is not a
legal requirement. However, the introduction of new devices to the European market is
significantly much simpler if devices follow these harmonized standards.
Technically, wireless power transfer is mandatory for backscatter transmission. Mean-

while, it is also necessary to collect and transfer the wireless power for energy storage in
the case of active transmission. By taking these points into consideration, high transmis-
sion powers are required for energizing Ambient IoT devices. However, the maximum
transmission power is limited on unlicensed frequency band in most cases based on the
regulations in different countries/regions. Therefore, it is important to pay more attention
to the regulatory requirements about the maximum transmission power if the Ambient
IoT system operates on an unlicensed frequency band, so as to meet these requirements
during the design and deploy phase.
In the following discussion, the frequency regulation in China, the USA and Europe will be introduced, respectively.
Frequency regulation In China, the frequency band of 920-925MHz is mainly for RFID system [1]. The radio trans-
for 900MHz and mission equipment doesn’ t need to acquire a license on this frequency band but it is forbid-
2.4GHz in China den to generate large interference to other existing legitimate radio stations. More specifical-
ly, the maximum ERP (Equivalent Radiated Power) is 100mW (i.e., 20dBm) within the
frequency range of 920-920.5MHz. And the maximum ERP is 2W (i.e., 33dBm) within the
frequency range of 920.5-924.5MHz. Furthermore, the maximum ERP is also equal to
100mW (i.e., 20dBm) within the frequency range of 924.5-925MHz. In the meanwhile, the
maximum occupied channel bandwidth is restricted to 250kHz. As for frequency hopping
spread spectrum systems (FHSS), the maximum dwell time is 2s for each hopping channel.
The frequency band of 2.4GHz refers to 2.4-2.4835GHz in China and it covers Bluetooth,
WLAN (Wireless Local Area Network) and other kinds of wireless access systems [2]. The
maximum EIRP (Equivalent Isotopically Radiated Power) is 100mW (i.e., 20dBm) when anten-
na gain is lower than 10dBi, and the maximum EIRP is 500mW (i.e., 27dBm) when antenna
gain is equal or larger than 10dBi.
FCC 15.247 The frequency regulation in FCC 15.247 covers both relevant frequency bands of 900MHz
in the USA [3][4] and 2.4GHz (i.e., 902-928MHz and 2400-2483.5MHz). The maximum transmission power is
limited to 1W (i.e., 30dBm). Additionally, the maximum allowed antenna gain is 6dBi. This
results in an EIRP up to 4W (i.e., 36dBm).
For FH systems,
From 15.247 (a)(1)(i) The maximum allowed 20 dB bandwidth of the hopping channel is
500 kHz
From 15.247 (b)(2) The max output power shall be 1 watt for systems with at least 50 hopping
channels; otherwise, 25 Watts.
In addition, if the 20 dB BW of the hopping channel is less than 250 kHz, the system shall use
at least 50 hopping frequencies and the average time on a frequency shall not be greater
than 0.4 seconds within a 20 second period. Otherwise, the system shall use at least 25 hop-
ping frequencies and the average time on a frequency shall not be greater than 0.4 seconds
within a 10 second period.
For digital modulation techniques,
From 15.247 (a)(2) The minimum 6 dB bandwidth shall be at least 500 kHz
From 15.247 (b)(3) The max output power shall be 1 Watt
ETSI Harmonised There are multiple harmonized standards in Europe for different frequency bands and
Standards in Europe [3][4] it is assumed that the transmission has different characteristics on each frequency
band.
1）ETSI Harmonised Standards in the 2.4GHz band
ETSI EN 300 328 covers “Wideband transmission systems; Data transmission equip-
ment operating in the 2.4 GHz band; Harmonised Standard for access to radio spec-
trum”. This standard mainly applies to IEEE 802.11 which is referred as wide-band data
transmission. Furthermore, it covers frequency hopping spread spectrum systems
(FHSS) (e.g., Bluetooth). The frequency range is from 2400 to 2483.5MHz. The maxi-
mum transmission power is 100mW EIRP (i.e., 20dBm).
There are also some additional restrictions when transmission power is higher than
10dBm EIRP. For wide-band transmissions, the maximum power spectral density is
limited to 10dBm/MHz. FHSS systems have a maximum transmission TX-sequence
time of 5ms, followed by a TX-gap time of 5ms. Additionally, FHSS have to follow addi-
tional restrictions concerning the hopping frequency separation.
ETSI EN 300 440 covers “Short Range Devices (SRD); Radio equipment to be used
in the 1 GHz to 40 GHz frequency range; Harmonised Standard for access to radio
spectrum” . This standard targets RFID system in the 2.4GHz band and other commu-
nication systems not covered by ETSI EN 300 328. Unspecific communication devic-
es may transmit with a maximum transmission power of 10dBm EIRP without any
additional restrictions.
The allowed frequency range for RFID is between 2446 and 2454MHz, which results
in a maximum bandwidth of 8MHz. The maximum transmission power is 4W EIRP (i.e.,
36dBm). However, 4W EIRP “shall by technical means be restricted to in building use”
and has duty cycle restrictions of 15%. These restrictions do not apply for the commu-
nication with transmission power of 500mW EIRP (i.e. 27dBm). However, even in this
case the frequency regulation mandates an antenna with ±45 degrees horizontal
beamwidth in addition to a side-lobe attenuation of equal or more than 15dB. Further-
more, it requires “physical protection (e.g. antenna dome) which dimension limits a
power transfer from the RFID antenna to a quarter wave matched dipole at posi-
tioned at an extreme close proximity to +15dBm”.
ETSI Harmonised All in all, the regulation of ETSI on the 2.4GHz band has the following requirements for
Standards in Europe [3][4] transmission power:
• For EIRP equal or lower than 10dBm(i.e., 10mW), no additional restriction.
• 20dBm EIRP (i.e., 100mW) in case of frequency hopping or wide-band signals
• 27dBm EIRP (i.e., 500mW) with directive antennas, bandwidth of 8MHz only, can
be extended to 36dBm EIRP (i.e., 4W) for stationary indoor use (possible to
verticals not sensitive to the cost)
2）ETSI Harmonised Standards in the sub-GHz band
For sub-GHz, ETSI uses the ERP (equivalent radiated power) instead of the EIRP
(equivalent isotropically radiated power). A transmitter with 1mW ERP (i.e., 0dBm) is
identical to a transmitter with 1.64mW EIRP (i.e., 2.15dBm).
ETSI EN 302 208 covers “Radio Frequency Identification Equipment operating in the
band 865 MHz to 868 MHz with power levels up to 2 W and in the band 915 MHz to
921 MHz with power levels up to 4 W; Harmonised Standard for access to radio spec-
trum”.
The frequency range between 865 to 867MHz defines four channels, each channel
has a bandwidth of 400kHz. The maximum transmission power is 2W ERP (i.e.,
35.15dBm EIRP) with antennas having a beamwidth ≤90° . For omni-directional anten-
nas, the maximum power is limited to 500mW ERP (i.e., 29.14dBm EIRP). The maxi-
mum transmission time is 4s, followed by a break of at least 100ms. The maximum
transmission power of the tags is -20dBm ERP in addition to a maximum bandwidth
of 400kHz in one of the four available channels.
The frequency range of 915 to 921 MHz also defines four channels. However, it is cur-
rently available in few European countries only. Power levels of up to 4W ERP (i.e.,
38.17dBm EIRP) require antennas having a beamwidth ≤90 ° . For omni-directional
antennas, the maximum power is limited to 1W ERP (i.e., 32.15dBm EIRP). The maxi-
mum transmission time is again 4s, followed by a break of at least 100ms. The maxi-
mum transmission power of the tags is -10dBm ERP in addition to a maximum band-
width of 800kHz in one of the four available channels.
ETSI EN 300 220-2 covers “Short Range Devices (SRD) operating in the frequency
range 25 MHz to 1 000 MHz; Part 2: Harmonised Standard for access to radio spec-
trum for non specific radio equipment” . Relevant frequency band in the scope of this
document is from 863 to 870MHz and from 915 to 921 MHz. It should be noted that
the frequencies between 915 and 921MHz are available in few European countries
only. The maximum transmission power is typically limited to 10mW or 25mW ERP.
However, a higher transmission power of 500mW ERP (i.e., 29.14dBm EIRP) is allowed
on band O with a bandwidth of 250kHz from 869.4 to 869.65MHz. However, this
band is further limited by a duty cycle of ≤10%.
ETSI EN 303 659 covers “Short Range Devices (SRD) in Data Networks; Radio equip-
ment to be used in the frequency ranges 865 MHz to 868 MHz and 915 MHz to 919,4
MHz with power levels ranging up to 500 mW e.r.p.; Harmonised Standard for access
to radio spectrum”. This will also cover IEEE 802.11ah.
This frequency regulation allows for 500mW ERP (i.e., 29.14dBm EIRP) if the channel
bandwidth is ≤ 200kHz. The available four channels are identical to the four channels
defined for RFID usage in ETSI EN 302 208. Additional restrictions are a duty cycle of
10% for AP and 2.5% otherwise. Furthermore, also systems with a bandwidth of up to
1MHz are possible, e.g. IEEE 802.11ah. The parameters are mainly identical to the
parameters of the 200kHz mode, but the maximum transmission power is limited to
25mW ERP (i.e., 16.13dBm EIRP).
All in all, the regulation of ETSI on the sub-GHz band has the following
requirements for transmission power:
• Highly fragmented into many different bands with different parameters (e.g., duty
cycle, transmission power, bandwidth)
• 16.13dBm EIRP (i.e., 25mW ERP) for 1MHz wide channels with duty cycle limita-
tions
• 29.14dBm EIRP (i.e., 500mW ERP) for systems with up to 200kHz bandwidth with
duty cycle limitations, one channel with up to 250kHz bandwidth
• 35.15dBm EIRP (i.e., 2W ERP) for RFID systems with directive antennas, band-
width limited to 200kHz for downlink and 400kHz for uplink
• 38.17dBm EIRP (i.e., 4W ERP) for RFID systems with directive antennas in bands
that are only available in some EU countries, bandwidth limited to 400kHz for
downlink and 800kHz for uplink
4.1.3 Link budget for Ambient IoT
Like conventional communication, the network coverage of the Ambient IoT system is
limited by the transmission power of the network equipment, working frequency
band, equipment antenna gains and equipment receiver sensitivity. In addition, it is
particularly important to note that the coverage of the Ambient IoT system is closely
related to the power level of the wireless energy harvesting signal when energy har-
vesting is performed based on the RF signal.
Specifically, for the forward link (i.e., the downlink from the network node to the Ambi-
ent IoT device), considering that the power consumption of driving the low-power
circuit is about several microwatts to tens of microwatts, the signal strength of the RF
signal received by the Ambient IoT device can preferably be above -20dBm (10 micro-
watts), which is also the sensitivity of the current typical RFID tag energy harvesting.
It is worth noting that this value is much greater than the receiver sensitivity of tradi-
tional terminals (e.g. about -100dBm). If the Ambient IoT device has a certain energy
storage capacity, for example, it can be equipped with an energy storage capacitor.
The signal strength of the RF signal received by the Ambient IoT device can be
relaxed to -30dBm or even lower [5][6], and at this time, the terminal can reserve the energy to be
used during operation through energy harvesting for a long time. When RF signals of the same
power are provided at the transmitting end, the Ambient IoT device with energy storage capability
can have a larger coverage.
In actual deployment, the transmission power of the network node is restricted by regional regula-
tions, e.g., a maximum EIRP of 36dBm, i.e. allowed transmission power of 30dBm, plus the antenna
gain of 6dBi, is regulated in the ISM (Industrial Scientific Medical) band. This leads to approximately
50dB link budget, resulting in a fairly limited communication distance.
For the backward link, i.e., from the Ambient IoT device to the network node. The signal strength of
the backscattered signal from the antenna of the device would be usually 3~5 dB lower than the
input signal, a.k.a. wireless power sourcing signal. The communication distance is restricted by the
receiver sensitivity of network node. Fortunately, the receiver sensitivity can achieve as low as
-100dBm to -110dBm [ 7 ] for a typical network node as implemented in 3GPP. It is thus able to
increase the link budget for the backward link up to 80dB, yielding a 30dB coverage extension com-
pared to the forward link.
Based on the above analysis, it becomes obvious that the coverage of the Ambient IoT system is
primarily limited by the coverage of the wireless power in the forward link, that is, the forward link is
a coverage bottleneck.
In a typical radio frequency identification system, where the ISM frequency band is the target oper-
ation band, the maximum coverage would be no more than 10 meters. As seen from the typical use
cases in chapter 2, a service coverage distance of up to e.g., 100 meters is envisioned in some use
cases. For example, it shall cover a whole factory in the smart manufacturing and industrial monitor-
ing scenarios, and the whole logistics station or warehouse needs to be covered in smart logistics
and smart warehousing scenarios. Licensed frequency band can be used in these cases, the
allowed transmission power can be increased by about 10 dB on the licensed frequency band com-
pared with a similar implementation in ISM band, which results in about 3 times of coverage exten-
sion in the forward link (Consider the limited forward link coverage). Therefore, it also confirms that
the use of licensed frequency bands is conducive to the construction of Ambient IoT system that
meet the requirements of the vertical industry.
The antenna gain of the device also affects the coverage of the Ambient IoT network. It not only
affects the coverage of the forward link, but also affects the coverage of the backward link. In some
application, there will be relaxed restrictions on the size and cost of Ambient IoT devices. In order to
achieve extended coverage, Ambient IoT devices can use high-gain antennas (e.g. 12dBi receiving
antenna gain) to increase the distance of uplink / downlink communication.
In some applications, if the terminal can harvest environmental energy other than RF energy or is
equipped with the conventional batteries, the downlink coverage of the Ambient IoT device can be
greatly expanded, and the downlink coverage distance will no longer be limited by the signal
strength threshold of energy harvesting, but rather by the lower sensitivity of downlink receiver of
the Ambient IoT device. Based on the current research, the sensitivity of Ambient IoT device down-
link receiver can reach -50/-60dBm or even lower [7-11].
Take the following three Ambient IoT devices as examples to calculate the
link budget, as shown in Table 4.1-1:
Type 1 Ambient IoT device’s energy harvesting is based on RF signals of network nodes, and has no energy stor-
age capacity, i.e. the device does not have capacitor or other components for energy storage. It requires
that the power of received RF signal is not lower than -20dBm, and backscatter communication mode is
adopted.
Type 2 Ambient IoT device’ s energy harvesting is based on RF signals of network nodes, has energy
storage capacity, and requires that the power of received RF signals should be higher than
-30dBm. Backscatter transmission mode or low-power active transmission mode is adopted for
communication, and when low-power active transmission mode is adopted for communication,
the transmission signal power is about -10dBm ~ -20dBm.
Type 3 Ambient IoT device’ s energy harvesting is based on other environmental energy (such as light
energy, RF signals provided by third-party equipment, etc.), and has energy storage capability. Ambi-
ent IoT device’s requirement for the power of the received RF signal from the network node depends
on the receiver design (i.e., the receiver sensitivity, which may be considered as -45 dBm). Backscat-
ter communication mode or low-power active transmission mode is adopted for communication, and
when low-power active transmission mode is adopted for communication, the transmission signal
power is about -10dBm ~ -20dBm (LNA can be used);
A preliminary calculation of the link budget is given in Table 4.1-1. Considering the operational
frequency band, transmission power, transmission loss, EIRP of network equipment, antenna gain of
Ambient IoT device, backscattering coefficient (the ratio of the signal strength of terminal reflection
signal and power supply signal), low noise amplification and other factors, and the communication
distance of Ambient IoT is preliminarily evaluated as shown in the following table (calculated based
on Friis equation):
Table 4.1-1 Estimation of link budget (Based on Friis equation)
Type1 Type2 Type3
Case1 Case2 Case3 Case4 Case5 Case6 Case7 Case8 Case9 Case10
Backs catter Backs Active Backs Active Active transmis Active transmis Higher
Backs with higher Backs catter transm catter transm sion with sion with frequency
catter antenna gain catter with LNA ission with LNA ission high EIRP high EIRP band
System configuration
Carrier frequency (MHz) 920 920 920 920 920 920 920 920 920 2400
Network node
EIRP of Network node(dBm) 36 36 36 36 36 36 36 46 46 46
Receiver sensitivity of Network (dBm) -100 -100 -100 -100 -100 -100 -100 -100 -110 -110
Ambient IoT device
Antenna Gain (dBi) 2 12 2 2 2 2 2 2 2 2
Sensitivity of energy harvesting

based on RF signal (dBm) -20 -20 -30 -30 -30 -20/-30 -20/-30 -20/-30 -20/-30 -20/-30
Receiver sensitivity of RF signal (dBm) -45 -45 -45 -45 -45 -45 -45 -45 -45 -45
Backscatter transmission loss (dB) 5 5 5 5 — 5 — — — —
Tx power of active transmission (dBm) — — — — <-20 — <-20 <-20 <-20 <-20
Low Noise Amplifier factor for

backscatter communication (dB) — — — 15 — 15 — — — —
Communication distance
(No shadow fading loss) PRX = PTX + GTX + GRX - Lbf
Network => Ambient IoT (m) ≤20.61 ≤65.18 ≤65.18 ≤65.18 ≤65.18 ≤366.54 ≤366.54 ≤1159.11 ≤1159.11 ≤444.32
Ambient IoT => Network (m) ≥183.71 ≥580.93 ≥58.09 ≥326.68 ≤326.68 ≥326.68 ≤326.68 ≤326.68 ≤1033.06 ≤396.00
Communication distance
(With shadow fading loss) PRX = PTX + GTX + GRX - Lbf - Lshadowing fading loss
Shadowing fading loss (dB) 5
Ambient IoT => Network (m) ≥103.31 ≥326.68 ≥32.67 ≥183.71 ≤183.71 ≥183.71 ≤183.71 ≤183.71 ≤580.93 ≤222.69
Shadowing fading loss (dB) 8
Ambient IoT => Network (m) ≥73.13 ≥231.27 ≥23.13 ≥130.05 ≤130.05 ≥130.05 ≥130.05 ≤130.05 ≤411.27 ≤157.65
PTX is the power transmitted to the transmitting antenna. GTX is the antenna gain of Tx. GRX is the antenna gain of Rx. PTXGTX
Friis λ
equation： PRX = PTX GTX ( —— )2 GRX is related to transmit antenna. When the transmitting antenna is matched, the power transmitted to the transmitting anten-
4πd
λ
na is converted into radiated power. ( —— ) 2 is related to the transmission loss of radio waves in free space.
4πd
Table 4.1-1 is the link budget estimation based on Friis equation, and preliminary calculation is carried
out based on different shadowing fading loss. In the actual deployment environment, in addition to
the factors listed in the table, there are other factors such as obstructions and penetration loss. In
order to obtain the evaluation results closer to the actual deployment, it is necessary to model
different fading loss based on the actual channel. The actual communication distance between net-
work nodes and Ambient IoT devices may be shorter than the values in the table. Table 4.1-1 is only
used for qualitative analysis.
The following conclusions can be obtained by analyzing Table 4.1-1 (taking the shadow fading loss of
8dB as an example).
（1） In Case1, the Ambient IoT device of Type 1 harvests energy based on the RF signal sent by the
network node. The sensitivity of the receiver is -20dBm (the sensitivity of energy harvesting based
on RF signal), and the antenna gain is 2dBi. When the working frequency band is 920MHz and the
EIRP of downlink signal of the network node is 36dBm, the maximum communication distance from
the network node to the Ambient IoT device is 8.21m. The Ambient IoT device communicates with
the network node via backscattering. The Ambient IoT device located at the edge of the network
coverage area receives an RF signal of-20dBm, which has a loss of about 5dB during backscatter-
ing. At this time, the signal strength sent by the Ambient IoT device in backscatter mode is about
-25dBm. Considering that the sensitivity of the network node is -100dBm, the communication
distance between the Ambient IoT device and the network node can reach 73.13 m.
（2） In Case2, except for the antenna gain of 12dBi, the other assumptions of the Ambient IoT
device of Type 1 are the same as those of Case1. The communication distance from network node
to the Ambient IoT device is increased from 8.21m to 25.95m, and the communication distance from
the Ambient IoT device to the network node is increased from 73.13m to 231.27m, which is about
three times. It can be seen that for an Ambient IoT device, using a high-gain antenna can effectively
improve the communication distance if allowed. However, for Type1 Ambient IoT devices, the energy
harvesting coverage of DL is the main bottleneck affecting the communication distance between
network nodes and Ambient IoT devices.
（3）In Case 3, the Ambient IoT device of Type 2 adopts an energy storage unit, which can reduce the
requirement of RF signal strength for energy harvesting from -20 dBm to -30 dBm compared with
Case 1. Accordingly, the communication distance between network nodes and Ambient IoT devices
has increased from 8.21m to 25.95m, which is about three times. It can be seen that when the energy
storage unit is adopted and the threshold of signal strength required for RF energy harvesting of Am-
bient IoT device is lowered, the communication distance between network nodes and Ambient IoT
devices can be effectively increased. In case3, the Ambient IoT device of Type 2 adopts backscatter
communication mode. Considering the backscatter loss of about 5dB, the signal strength sent from
Ambient IoT device is about -35dBm. At this time, the communication distance between the corre-
sponding Ambient IoT device and the network node is reduced from 73.13m to 23.13m.
（4） In case 4, the Ambient IoT device of Type 2 adopts LNA on the basis of case 3. By using LNA
with 10dBi gain, under the condition that the communication distance from the network node to the
Ambient IoT device remains unchanged, it can effectively increase the backscatter communication
distance from 23.13m to 130.05m compared with case3. It can also increase the coverage range
from the Ambient IoT device to the receiving network node when different network nodes send and
receive.
（5） In Case 5, the Ambient IoT device of Type 2 adopts active transmission mode for communica-
tion, and has the same communication distance from the network node to the Ambient IoT device
as case 4. Because active transmission mode is adopted for communication, even the Ambient IoT
device located at the coverage edge can still accumulate enough energy over a period of time, and
use active transmission mode to send signals with signal strength of -10dBm ~ -20dBm. Taking the
active transmission signal strength of -20dBm as an example, the communication distance
between Ambient IoT devices and network nodes is about 130.05 m.
（6） In Case 6, the Ambient IoT device of Type 3 does not directly harvest energy based on the RF
signal of network nodes, it can harvest energy based on light energy and RF signals provided by
other equipment to obtain energy for communication. The communication distance from the net-
work node to the Ambient IoT device is mainly limited by the receiver sensitivity for RF signal of the
Ambient IoT device. Taking -45dBm as an example, the maximum communication distance from the
network node to the Ambient IoT device is about 145.92 m. When the Ambient IoT device uses
backscatter mode to communicate, if the energy of the energy harvesting signal is -30dBm, the
backscatter loss is considered to be 5dBm, and a 15dBm backscatter amplifier is used, the trans-
mission power of the backscatter signal is about -20dBm, and the communication distance from
the Ambient IoT device to the network node can be about 130.05m.
（7） In Case 7, the Ambient IoT device of Type 3 does not directly harvest energy based on the RF
signal of network nodes, it can harvest energy based on light energy and RF signals provided by
other equipment to obtain energy for communication. The communication distance from the net-
work node to the Ambient IoT device is mainly limited by the receiver sensitivity for RF signal of the
Ambient IoT device, taking -45dBm as an example. At this time, the maximum communication
distance from the network node to the Ambient IoT device is about 145.92m. If the Ambient IoT
device communicates in an active transmission mode, it can support a communication distance of
about 130.05 m.
（8） In Case 8, the same Ambient IoT device assumptions are considered as in Case 7. When the
EIRP of RF signal transmitted by network node is increased from 36dBm to 46dBm, the communi-
cation distance between the network node and the Ambient IoT device can be effectively increased
from 145.92 m to 461.45 m compared with Case 7. However, limited by the receiver sensitivity of the
network node and the signal transmission power of the Ambient IoT device, the communication
distance from the Ambient IoT device to the network node does not change.
（9） In Case 9, on the basis of the assumption of Case 8, when the receiver sensitivity of the net-
work device is increased from -100dBm to -110dBm, the communication distance from the Ambient
IoT device to the network node can be increased from 130.05m to 411.27m compared with Case 8
while keeping the communication distance from the network node to the Ambient IoT device
unchanged.
（10） In Case 10, on the basis of Case 9, a higher frequency band is used for communication, for
example, from 920 MHz to 2.4 GHz, and it can be seen that the effective communication distance is
significantly reduced.
4.2 Topology of Ambient IoT

According to different application scenarios and corresponding system requirements, it can support
different Ambient IoT network topologies. According to whether there is a relay node or an assistant
node for the data and control link between the Ambient IoT device and the reading node, the net-
work topology of Ambient IoT can be classified into three types: direct network topology, relay net-
work topology and hybrid network topology.
4.2.1 Direct network topology
In the direct network topology, the reader directly communicates with the Ambient IoT device for the
transmission/reception of data. Based on different deployment scenarios, direct network topologies
can be divided into two types: cellular direct topology and Local Area Network (LAN) direct topology.
4.2.1.1
Cellular direct topology
Cellular-based Ambient IoT systems can support large-scale deployment and centralized control of
Ambient IoT devices, aiming to solve the problems of short communication distances, high deploy-
ment costs, and low system efficiency of traditional point-to-point and point-to-multipoint IoT tech-
nologies (such as RFID). Thanks to the advantages of the cellular network in coverage distance,
access capability and resource utilization, the cellular-based Ambient IoT system can manage the
Ambient IoT devices in the network in a large range and in a centralized manner, which can greatly
improve system efficiency and save deployment costs. Therefore, cellular-based Ambient IoT sys-
tems are particularly suitable for certain application scenarios.
In the industrial sensor network scenario, the deployment environment of devices is strict, the
number of devices is huge, and the deployment and maintenance cost of traditional active devices
are high. The Ambient IoT network based on cellular direct connection can remotely and centrally
manage the Ambient IoT devices for control and information exchange. In logistics and warehouse
scenarios, a large number of goods need to be identified, tracked, and inventoried. Compared with
existing QR code or RFID based methods, cellular-based Ambient IoT systems can overcome the
inefficiency and low reliability of existing optical identification and short-range identification, greatly
simplifying the identification process, saving manpower and equipment investment, and reducing
costs. In the smart farming and animal husbandry scenario, livestock carrying Ambient IoT devices can
be managed in the farm through the cellular network, including statistics, positioning, tracking, etc.
As shown in Figure 4.2-1, a cellular-based Ambient IoT system can include the
following communication modes:
Case1
Cellular direct
g
zin g
gi rin
ner ge
E rig
/T
ta
da
L
U
UL
data
Trig
g erin
g on
ly
ing
giz
er Case2
En
Energized by
assistant node
UL data Energizing Triggering
Figure 4.2-1 Cellular direct topology
Case1： The base station and the Ambient IoT device can communicate directly with each other. The base
Cellular direct station provides wireless energy supply signals and trigger signals to the Ambient IoT device. The
wireless energy supply signal is used to provide energy to the Ambient IoT device; The trigger signal
can carry control information sent to the Ambient IoT device; Ambient IoT devices transmit informa-
tion to base stations.
It should be pointed out that the Ambient IoT device can send information to the base station by
backscattering or active emission. When using the backscatter method, it is necessary to consider
how to provide the backscattered carrier to the Ambient IoT device.
In a typical implementation, the backscattered signal has the same frequency as the incident
carrier. At this time, if the incident carrier is provided by the reader, the reader needs to send the
backscattered signal of the carrier and receive the Ambient IoT device at the same time, that is, it
has full duplex capability, which will significantly increase the implementation complexity of the
reader.
If Ambient IoT devices can support large frequency shifts during backscattering, the implemen-
tation complexity of reader can be reduced. However, in order to achieve the frequency shift
when backscattering, in one possible implementation, the Ambient IoT device needs to be able to
generate a local frequency shift carrier, based on which it is mixed with the incident carrier, and
achieve a single sideband backscatter modulation (such as a mirror-suppressing mixture), which
will also significantly increase the implementation cost and power consumption of Ambient IoT
devices.
If the Ambient IoT works in the FDD band and the reader is the base station, due to the restric-
tion of regional spectrum regulations, the carrier needs to be sent in the FDD downlink band, but
the Ambient IoT device needs to backscatter in the uplink band, and the Ambient IoT device
needs to realize the frequency shift from the downlink band to the uplink band during the back-
scattering. This requirement will greatly increase the difficulty of implementing Ambient IoT
devices.
Case2： Ambient IoT devices can not only obtain wireless energy from the base station with which they com-
Energized by municate, but also obtain energy from third-party devices. Through the wireless power supply to
assistant node
obtain energy, the strength of the energy supply signal to the device needs to meet a certain thresh-
old, such as -20dBm (when the device has no energy storage capacity) or -30dBm (when the device
has energy storage capacity), which causes limited coverage of the energy supply signal transmit-
ted by the network when the transmit power of signal is restricted, generally in the range of tens of
meters to 100 meters. From the perspective of cellular coverage, the coverage of wireless energy
supply is much smaller than that of information transmission signals. Therefore, the coverage of
wireless energy supply signal is a bottleneck. Wireless power supply through more network nodes
can significantly improve coverage, thereby maximizing the cell coverage of the Ambient IoT com-
munication. For this purpose, other nodes in the network can be used for wireless power supply.
Potential power supply nodes include smart phones in the network, Relay nodes, CPEs, routers, etc.
Where necessary, dedicated power supply nodes can also be deployed. The traditional wireless
communication signals sent by these nodes (such as synchronization signals, broadcast signals,
data channels, etc.) can provide wireless energy for Ambient IoT devices, or based on reasonable
scheduling methods, these energy supply nodes can also send dedicated wireless energy supply
signals.
The cellular direct topology with assistant energizing node can not only effectively expand the cov-
erage, but also well solve the above-mentioned full-duplex problem of the base station in the case of
backscattering and reduce the complexity when implementing Ambient IoT. With the help of the
energy supply signal and carrier signal sent by the third-party device, the base station device is only
responsible for the transmission and reception of DL and UL signals between the Ambient IoT devic-
es, without sending the energy supply signal and carrier signal, thus avoiding the full duplex require-
ment to the base station. Sending the carrier signal through a third party at UL also avoids the prob-
lem of local carrier generation for significantly frequency shifting with mirror-suppressing mixture.
4.2.1.2
LAN direct topology
An Ambient IoT based on a direct LAN connection topology enables direct communication between
an Ambient IoT device and other types of devices, such as smartphones, WiFi routers, CPEs, or
other IoT devices. In the topology of LAN direct connection, direct communication does not depend
on the cellular network. This deployment mode also has a wide range of application scenarios, espe-
cially for low-cost short-range communication requirements. For example, in the smart home sce-
nario, the direct communication between the intelligent device and the Ambient IoT device can
achieve functions such as finding objects, family asset management, environmental monitoring, and
intelligent control. In smart wearable scenarios, information reading or intelligent control of Ambient
IoT wearable device can be achieved.
An Ambient IoT system based on a LAN direct connection topology can include the following com-
munication modes:
Case 1: The Ambient IoT devices communicate directly with smart devices. The smart device sends the
LAN direct energy supply signal and trigger signal to the Ambient IoT device. Ambient IoT device directly trans-
mit information to smart devices to achieve direct communication. Smart devices can be mobile
phones, routers, or controller nodes (such as CPE). The topology maps to Case1-1 and Case1-2 in
Figure 4.2-2.
Case 2: In order to realize the direct connection between the Ambient IoT device and the
Energized by assistant node smart device, the wireless energy supply signal to the Ambient IoT device can not
come directly from the smart device, but from the third-party device. As shown in
Case2 in the figure, the wireless energy supply signal required for the direct connec-
tion between the Ambient IoT device and the mobile phone comes from the control
node (router or CPE in the figure). After collecting the energy sent by the third-party
device, the Ambient IoT device receives the trigger signal sent by the smart device
and transmits the information to the smart device.
Case1-1:
LAN direct
igg ing
ng
/tr ergiz
eri
ta
da
En
UL
UL d
ata
Trigg
ering Case2:
only
Energized by
nly assistant node
o
iz ing
erg
En
Case1-2: izing
Energ ng
LAN direct eri
/trigg
ta
UL da
UL data Energizing Triggering
Figure 4.2-2 LAN direct topology
It should be pointed out that for the backscatter type of Ambient IoT device, there is
also a full-duplex problem similar to that in the aforementioned cellular direct connec-
tion topology, which can also be solved by means of third-party node power supply.
4.2.2 Relay network topology
UL data
Triggering
ing
ger
Energizing trig
g/
rgizin
Uu data/control Relay Ene
network dat
a
UL
Trig
gerin
Uu data/control go
nly
UL
dat
a
Intermediate
node
Figure 4.2-3, Relay network topology
In the relay network topology, there is an intermediate node between the reading node and the
Ambient IoT device. The intermediate node can be a network relay device or an intelligent device.
The relay network topology has significant advantages in some application scenarios:
Greatly expand the coverage distance
Due to the complexity and power consumption limitations of Ambient IoT devices, the coverage of
the above-mentioned direct connection topology is limited. Intermediate nodes (relay network
nodes or smartphones) can be supported to work well under existing cellular network coverage.
Therefore, the use of relay network topology can greatly expand the coverage distance and realize
the support of Ambient IoT in the existing network topology.
Enable reusing of existing network deployments
With the help of intermediate nodes, the Ambient IoT device can work in the existing network
deployment architecture, and the existing network equipment can support the Ambient IoT through
software upgrade, so as to reuse the existing network infrastructure as much as possible, and the
intermediate node has the advantages of low complexity and easy deployment. Therefore, this
method greatly reduces the network deployment cost.
Provide better wireless power
Although the Ambient IoT device can collect a variety of environmental energy, such as light, heat,
kinetic energy, etc., in more scenarios, radio frequency energy is still the most important, most reliable,
and controllable source of environmental energy. For example, in the aforementioned smart home sce-
nario, the most convenient source of environmental energy for the Ambient IoT device affixed to
household items is still radio waves; In logistics and warehouse scenarios, radio waves are still the most
likely source of environmental energy, due to the size of the tag and the working environment.
However, the coverage of wireless power supply is relatively limited as analyzed in section 4.1, if only
relying on the cell base station directly connected to the Ambient IoT device to achieve wireless
power supply. In actual deployment, high base station deployment density is required, which will
significantly increase the network deployment cost. Relying on flexible deployment of intermediate
nodes to achieve wireless energy supply can greatly reduce the cost of wireless energy supply and
achieve better wireless energy coverage.
Support complex and extreme application scenarios
In the actual deployment scenario, there may be serious problems of blocking attenuation,
multi-path interference and multi-cell interference. As shown in Figure 4.2-4, in the warehouse sce-
nario, multiple layers of goods may block each other, resulting in severe attenuation of the wireless
energy supply signal that can be received by the Ambient IoT devices on some goods, while the
signals emitted by the Ambient IoT tags of some goods are also seriously blocked. With the help of
intermediate nodes, the problems of energy supply blocking and signal attenuation can be solved
flexibly.
gNB
Intermediate Intermediate
node node
Figure 4.2-4 Relay network topology in warehouse scenario
In the cold chain logistics scenario, real-time monitoring of the temperature and other information of
fresh goods transported on the road is required. With the help of intermediate nodes deployed on
the logistics truck, even in the process of high-speed transportation, the Ambient IoT device
attached to the fresh transportation box can still report real-time monitoring information of fresh
goods to the network.
Intermediate gNB
node
Figure 4.2-5 Relay network topology in cold chain logistics scenario

4.2.3 hybrid network topology
In the actual deployment of Ambient IoT systems, based on the requirements of different scenarios, you can consider the
flexible coexistence or combination of the direct network topology and relay network topology to build a hybrid network
topology, so as to match more potential application scenarios. The system block diagram of the environment of the
hybrid network topology is shown below. Multiple communication modes can be used, as shown in Figure 4.2-6:
Case1： In this topology, the DL trigger signal is provided by an intermediate node, wireless energy supply for
Energized/ wireless energy acquisition/backscatter communication can also be provided by this intermediate
triggered by node, and the UL signal sent by the Ambient IoT device is received by the base station. Among them,
intermediate
the power supply of intermediate nodes and the operation of sending the DL trigger signal can be
node
controlled by the base station through air interface.
With the help of intermediate nodes to send the DL trigger signal and wireless power supply/back-
scatter carrier, the topology can effectively expand the coverage of DL communication and wireless
power supply. When backscattering is used, because the intermediate node is closer to the Ambient
IoT device, it can also provide stronger incident carrier energy, so the communication distance of UL
is also improved. In addition, through the separation of DL wireless power supply and UL reception,
this method also avoids the full-duplex problem when using the backscatter transmission mode.
Case2： The base station provides wireless power supply/backscatter carrier for wireless energy acquisi-
Energized / tion/backscatter communication and sends the DL trigger signal to the Ambient IoT device, while
triggered by the UL signal sent by the Ambient IoT device is received by the intermediate node, such as an intelli-
network
gent terminal. Further, the intermediate node sends air interface data to the base station.
With the help of intermediate nodes to receive the UL signals, the topology can effectively expand
the coverage of UL communication. In addition, through the separation of DL wireless power supply
and UL reception, this method also avoids the full-duplex problem when using the backscatter
transmission mode.
Energ
izing
and tr
iggerin naling
g Uu sig
and
Ene riggerin
UL data
t
rgiz
ing g
UL
a dat
dat a
Uu
Case2: Case1:
Energized /triggered Energized /triggered
by network by smart device
UL data Energizing Triggering Signaling/Data
Figure 4.2-6 hybrid network topology
It can be seen from the discussion in this section that different Ambient IoT modes can be consid-
ered in different application scenarios based on actual needs. We need to further study the charac-
teristics of various deployment scenarios and communication methods to maximize the application
potential of Ambient IoT systems.
Ambient IoT device type

4.3 and functionality
Ambient IoT devices can be divided into different device types according to their own
characteristics, including energy storage capacity and RF signal generation capacity.
1.Categorization based on Type 1 Ambient IoT devices without long-term energy storage. This kind of devices
whether there is energy does not have the ability to store energy for a long time, for example, it cannot contin-
storage uously store the collected energy for more than several hours. This kind of equipment
can drive its own circuit to work by using the environmental energy instantly acquired.
As described in Section 3.1, the low energy density of the Ambient IoT device requires
very low power consumption of this type of equipment. It should be pointed out that
although this kind of equipment does not have the ability to store energy for a long
time, it can still use energy storage devices such as tiny capacitors to smooth the
instantaneous collected energy to cope with the fluctuation of the collected environ-
mental energy power level in extreme time (such as microseconds or milliseconds).
Type 2 Ambient IoT devices with long-term energy storage capacity. This kind of
devices has limited energy storage capacity, and can store the collected energy for a
long time, such as several hours or even several days. Ambient IoT devices collect
radio wave energy or energy in the environment (such as solar energy, thermal energy,
mechanical vibration energy, etc.), and store the collected energy in an energy storage
unit (such as capacitor or solid-state energy storage device). After obtaining enough
energy, it can drive Ambient IoT devices to send and receive signals, and conduct data
processing and other work. Although energy storage devices are used, these devices
are also real Ambient IoT devices because they do not need to use traditional batter-
ies and do not need manual maintenance work such as charging.
Type 3 Active Ambient IoT devices. Ambient IoT device uses ultra-low power commu-
nication technology for signal transmission. Although with the built-in battery, this
type of Ambient IoT devices has extremely low power consumption and complexity, so
it can use smaller capacity batteries to achieve smaller volume and save battery cost.
The extremely low device power consumption can also greatly prolong the service life
of the battery and reduce the maintenance cost of the device.
2.Categorization based on Type 1 Backscatter-based Ambient IoT devices. These devices do not have indepen-
uplink transmission mech- dent signal generators and transmit signals through backscatter. Therefore, when the
anism device performs data transmission, the network is required to provide a carrier, and
the device performs backscatter based on the carrier to realize data transmission.
Such devices can also use reverse amplifiers to amplify reflected scattered signals to
improve coverage
Type 2 Ambient IoT devices based on active transmitter. This kind of Ambient IoT
devices has the ability to generate RF signals independently. Therefore, this kind of
Ambient IoT devices can send data by using its own active transmitter when sending
data, without the need for network to provide carriers
Type 3 Ambient IoT devices with both backscatter and active transmitters. This type
of devices can support both backscatter and active transmission. The device can
determine which uplink transmission mode to use for data transmission according to
different conditions (such as residual power, available ambient energy) or based on
the scheduling of network.
3.Categorization based on Type 1 Ambient IoT devices with only uplink one-way communication. This device does
communication direction not need to receive downlink control information from base stations or other control
nodes before sending uplink information. When the device obtains enough energy
from the surrounding environment, it can send uplink information.
Type 2 Ambient IoT devices with only downlink one-way communication. This device
needs built-in receiver to control other devices (such as light switch, start of agricul-
tural facilities, etc.) by receiving downlink control information. In order to ensure the
correct reception of one-way control information, the probability of correct transmis-
sion can be improved by repeating multiple downlink transmissions.
Type 3 Ambient IoT devices with uplink and downlink two-way communication. This
device supports both downlink receiving control information and uplink transmitting
information. By receiving downlink control/data information, the device can obtain
control information and service data information. In addition, according to the instruc-
tion of downlink control information, the device will perform uplink transmission,
including uplink feedback, data transmission, etc.
4.Categorization based on Type 1 Different shapes of Ambient IoT devices. Considering the extensive application
shapes and materials scenarios of Ambient IoT, devices can be designed in different shapes to adapt to
different scenarios. For example, the device may be in the shape of a square or bar
tag, attached to the surface of a box item such as a courier.
Type 2 Different materials of Ambient IoT devices. In order to work normally in differ-
ent scenarios, device materials usually need to be specially selected or processed. For
example, when the device is attached to the metal surface of power device, it is nec-
essary to consider the influence of its own material on the dielectric constant and
design the metal-resistant material; Waterproof materials shall be considered when
the equipment is used under wet condition; When the device is used to work on soft
articles such as documents, it is necessary to consider the use of bending resistant
materials; When the device is used as a wearable device, the skin-friendly nature of
the material shall also be considered.
5.Categorization based on Type 1 Ambient IoT devices providing inventory function. In logistics and warehousing
functionalities scenarios, the Ambient IoT device can support reading a large number of tags at the
same time for cargo inventory. For example, the wireless tag affixed to the goods or
containers in the warehouse will record and save the basic information of the goods
(such as ID) and the location information in the warehouse. Through the built-in cen-
tral network node of the warehouse, the goods in the warehouse can be confirmed in
time and quickly for warehousing, outbound, storage and other operations, and the
quantity of goods can be counted to achieve real-time update.
Type 2 Ambient IoT devices providing environmental information collection function. Ambient IoT
devices can be used for environmental monitoring and environmental information collection, e.g.,
capturing information such as temperature and humidity in the data center or computer room,
monitoring the temperature of power devices in the power scene, finding abnormal situations in
time, and preventing accidents.
Type 3 Ambient IoT devices providing tracing and positioning function. Ambient IoT devices with
extremely small size and waterproof and folding resistance features can be stuck on some easily
lost items in the family, such as keys, passports, bank cards and wallets, thus can be used to quickly
locate and find lost items when looking for these items.
Type 4 Control-type Ambient IoT devices. By receiving control instructions, it can achieve the func-
tion of controlling a certain type of devices. For example, in the smart greenhouse scene of smart
agriculture, the Ambient IoT device can not only monitor the humidity of soil, the temperature and
humidity in the greenhouse, and the concentration of carbon dioxide, but also receive the control
command of the network to control the intelligent irrigation system to increase or decrease the wa-
tering amount, control opening or closing of ventilation pipes, or control opening or closing of
greenhouse windows, so as to realize the greenhouse environment control.
Coexistence of the Ambient IoT and

4.4 other communication systems
4.4.1 Coexistence with cellular communication systems
There are three deployment modes to consider including in-band deployment, guard-band deploy-
ment and standalone deployment, when deploying the Ambient IoT in the cellular communications
band. In order to avoid mutual interference between systems, it is necessary to study the coexis-
tence of the Ambient IoT and existing cellular communication systems (such as NR, LTE, GSM, etc.).
For the coexistence of the Ambient IoT communication systems and existing cellular communication
systems, the most important thing is to analyze the impact of Rx performance, including In Channel
Sensitivity (ICS), maximum input power, Adjacent Channel Selectivity (ACS), In-band/out-of-band/nar-
row-band blocking, and spurious response.
Out Of-Band In-Band ACS
5MHZ -15dBm
-35dBm
-52dBm
ICS
-81
Wanted
RefSens
Figure 4.4-1 Basic receiver RF specifications

Firstly, the impact on the receiver performance of conventional cellular systems needs to be consid-
ered. The communication signal and the energy transfer signal of the Ambient IoT device may fall
into the adjacent band or in-band of the conventional communication system, forming adja-
cent-band interference or blocking. In particular, for in-band mode, it is necessary to avoid co-chan-
nel interference between systems. As described in section 3.1, based on wireless energy acquisition,
the input power of the signal received by the Ambient IoT device needs to have a sufficiently high
signal strength that is generally greater than -20dbm/-30dbm, which requires the energy supply
node to transmit a sufficiently high power energy supply signal. However, there is a limit on the maxi-
mum signal strength that can be received by the existing terminal receiver. For example, the maxi-
mum input power of the terminal specified in the NR protocol is -15 dBm. Therefore, it is necessary
to investigate the impact of in-band deployment on existing terminals and how to avoid the impact.
Transmit Signal Out of band Blocker

Power
In-band Blocker
Desired
Signal
DC Freq
Tx Band Rx Band
Figure 4.4-2 Receiver blocking

Secondly, it is necessary to consider the impact of UL signals transmitted by the Ambient IoT devic-
es on the performance of receivers in conventional communication systems. The RF link of the Am-
bient IoT device is extremely simple, and when the Ambient IoT device uses backscatter transmis-
sion, the device is difficult to carry out fine filtering of the backscattered signal, resulting in
out-of-band leakage of the backscattered signal. For example, if the backscattered signal uses OOK
modulation, the spectrum diagram shown in Figure 4.4-3, and it can be seen that sidelobe leakage
appears on the spectrum diagram. Thus, it is necessary to assess the impact of these sidelobe leak-
age on existing cellular communication systems.
× 104
10
-2
1.12 1.112 1.124 1.126 1.128 1.13

× 105
Figure 4.4-3 Spectrum distribution of backscattered signal (OOK modulation)
Finally, the impact of coexistence on receiver performance of Ambient IoT devices needs to be con-
sidered. In order to reduce the device complexity and the device power consumption, a class Ambi-
ent IoT can use RF envelope detection receivers, which are extremely simple in structure. For exam-
ple, it does not require an oscillator to down-convert the RF signal, nor does it require a fine signal
filter. Thus, the power consumption of this type of receiver can be as low as a few uW. However,
even if the Ambient IoT bandwidth is very narrow, such as only a few hundred kHz, the Ambient IoT
device will still receive tens of MHz of signals including the Ambient IoT DL signal at the receiver due
to the poor overall filtering performance. Therefore, when the Ambient IoT and other cellular systems
are deployed in the adjacent frequency, the Ambient IoT device receiver will receive the signals of
other cellular systems on both sides together with the Ambient IoT communication signals, which will
form strong interference to the reception of the Ambient IoT device DL signal and affect the system
performance.
Ambient
IoT
Existing system Existing system
(GSM、 LTE、 NR) (GSM、 LTE、 NR）
Figure 4.4-4 The DL signal received by an Ambient IoT device contains

signals from other communication systems
4.4.2 Coexistence with the communication system

operating on an unlicensed frequency band
It is also necessary to consider the coexistence between the Ambient IoT system and the communi-
cation system operating on an unlicensed frequency band (such as Bluetooth, WiFi, RFID, etc.) when
deploying an Ambient IoT system on an unlicensed frequency band.
Systems operating in all unlicensed frequency bands are required to comply with spectrum usage
rules, such as maximum transmit power, frequency hopping, out-of-and leakage requirements, etc.,
as detailed in Section 4.1. Thus, the Ambient IoT system also needs to comply with the spectrum
usage rules of unlicensed frequency bands. More importantly, it is necessary to ensure the fair use
of channel resources between systems. Therefore, the devices operating on unlicensed frequency
bands are required to do channel listening before sending a signal, that is, to do Listen Before Talk
(LBT). However, for the Ambient IoT devices, due to the limited energy they collect, requiring them to
perform LBT before transmitting signal will lead to excessive energy consumption, and the limited
energy is difficult to support frequent LBT operations. Therefore, how to ensure the good coexis-
tence between the Ambient IoT systems and the existing unlicensed systems, while avoiding the
excessive power consumption of the Ambient IoT device to execute LBT is a problem that needs to
be solved.
A possible solution is to use channel sharing mechanism. In the Ambient IoT systems, when the
device needs to send signals, the network device should assist in occupying the channel and sharing
the transmission opportunity with the Ambient IoT device, so that the Ambient IoT device does not
need to perform LBT to send date during the time when the network device shares the channel.
CCA DL Frame
Network
device
UL Frame
Ambient IoT
device
Figure 4.4-5 Network devices share sending time with the Ambient IoT device
4.5 Reference
[1]Radio Regulations for Radio Frequency Identification (RFID) Devices in the 900 MHz Band, Radio
Administration Bureau of the Ministry of Industry and Information Technology.
[2]Announcement Regarding Adjustments to the Transmit Power Limit and Associated Issues in the
2.4GHz Frequency Band.
[3]IEEE 802.11-23/0436r0, Technical Report on support of AMP IoT devices in WLAN.
[4]IEEE 802.11-23/0089r0, Frequency Regulation Chapter for AMP TIG Report.
[5]Noghabaei S M, Radin R L, Savaria Y, et al. A high-efficiency ultra-low-power CMOS rectifier for

RF energy harvesting applications[C]//2018 IEEE International Symposium on Circuits and Systems
(ISCAS). IEEE, 2018: 1-4.
[6]Wu Z, Zhao Y, Sun Y, et al. A self-bias rectifier with 27.6% pce at-30dbm for rf energy harvesting
[C]//2021 IEEE International Symposium on Circuits and Systems (ISCAS). IEEE, 2021: 1-5.
[7]Ayers J, Mayaram K, Fiez T S. An ultralow-power receiver for wireless sensor networks[J]. IEEE
Journal of solid-state circuits, 2010, 45(9): 1759-1769.
[8]Moody J, Bassirian P, Roy A, et al. A− 76dBm 7.4 nW wakeup radio with automatic offset compen-
sation[C]//2018 IEEE International Solid-State Circuits Conference-(ISSCC). IEEE, 2018: 452-454.
[9]Salazar C, Kaiser A, Cathelin A, et al. 13.5 A− 97dBm-sensitivity interferer-resilient 2.4 GHz

wake-up receiver using dual-IF multi-N-Path architecture in 65nm CMOS[C]//2015 IEEE Internation-
al Solid-State Circuits Conference-(ISSCC) Digest of Technical Papers. IEEE, 2015: 1-3.
[10]Durante M S, Mahlknecht S. An ultra low power wakeup receiver for wireless sensor nodes
[C]//2009 Third International Conference on Sensor Technologies and Applications. IEEE, 2009:
167-170.
[11]Magno M, Benini L. An ultra low power high sensitivity wake-up radio receiver with addressing
capability[C]//2014 IEEE 10th International Conference on Wireless and Mobile Computing, Net-
working and Communications (WiMob). IEEE, 2014: 92-99.
Key techniques and challenges of Ambient IoT 05
KEY TECHNIQUES
AND CHALLENGES
OF AMBI-ENT IOT
Ambient IoT devices rely on the energy obtained externally. In order to support
Ambient IoT, RF radio energy will be a very important source of ambient energy.
In Section 5.1, we will analyze the requirements and challenges of wireless
power supply for supporting Ambient IoT. Further, in order to adapt to the mini-
malist hardware structure and ultra-low complexity for the Ambient IoT device,
we will analyze the challenges of data transmission of the Ambient IoT device in
Section 5.2, including potential modulation scheme, coding, multiple access,
resource allocation and synchronization. The dependence on wireless power
and the minimalist software and hardware structure of the device also requires
lightweight protocol stack and lightweight security mechanism, which will be
described in section 5.3 and section 5.4 respectively. Finally, Ambient IoT also
imposes new requirements for network architecture. We will discuss the simpli-
fied network architecture suitable for Ambient IoT in section 5.5.
05
The requirements and challenges of

5.1 wireless power sourcing
As described in Section 3.1, the Ambient IoT device itself does not need to contain a battery. There-
fore, before communication, the device needs to harvest ambient power to obtain the energy
needed. As explained in Section 3.1, based on different application scenarios, different ambient
energy sources (e.g., light, heat, etc.) can be used for energy harvesting. Nonetheless, RF radio wave
is favored to be used as a source of ambient energy in many application scenarios due to its control-
lability and widespread availability in wireless communication networks.
In order to support the Ambient IoT device to communicate in the wireless communication network,
it needs to provide wireless power supply to the Ambient IoT device. Compared with the traditional
communication system, how to effectively supply energy to the Ambient IoT device in a reasonable
way to provide appropriate network coverage is a new challenge for the Ambient IoT system.
Case1
ng
ci
o ur ing
S er
er g
w trig
Po &
g
att ack
in
er
Sc B
ling Bac
igna k Sc
Uu s atte
ring
Trig
data gerin
g
Case3
Case2 ly
pp r
r su ing we y
owe igger Po ppl
P tr su
&
ck g
Ba erin
att
Sc
Backscater Power supply Triggering Uu Signaling/Data
Figure 5.1-1 Various schemes for wireless power supply
Fortunately, there are plenty of network nodes in wireless communication network such as base
station equipment, user device equipment, relay equipment, routers, CPE, etc. One of the main func-
tions of all these network nodes is to transmit wireless signals. Therefore, these devices can be used
to provide wireless power to Ambient IoT consumption devices.
As shown in Figure 5.1-1, when the Ambient IoT device is deployed in the cellular network, the base
station can directly send wireless power supply signal to the device, as shown in case 1 in the Figure.
In this case, each time before the base station communicates with the device, it needs to send a
wireless power supply signal to charge the device so that it obtains sufficient energy to be activated.
In the communication process, the wireless power supply signal also needs to be continuously trans-
mitted to enable the device to obtain the energy required to maintain normal operation. For example,
in the process of downlink communication, the device needs to receive the wireless power supply
signal to obtain the energy needed to maintain the device for downlink signal reception, signal
demodulation and other operations. Since the downlink signal carrying information also carries wire-
less energy, it can also be used as a wireless power supply signal. Similarly, in the uplink communica-
tion process the device sends data to the network equipment, it is also necessary to continuously
send the wireless power supply signal to the device. At this time, the wireless power supply signal
provides the energy required by the device for data acquisition (such as reading data from the
sensor or memory), coding and other operations. In addition, the wireless power supply signal can
also be used as the carrier signal for backscatter Ambient IoT devices, so that the Ambient IoT devic-
es can perform backscatter of the wireless power signal, thereby completing the uplink signal trans-
mission.
The wireless power supply signal not only goes through the downlink channel, but also carries the
backscattered uplink signal through the uplink channel. Especially for passive devices, when the
wireless power supply signal arrives at the device, its signal strength shall not be lower than - 20 /-
30dBm (when the device can reserve the harvested energy). These requirements and limitations
make it a challenge to build Ambient IoT networks to provide sufficient coverage. As analyzed in Sec-
tion 4.1, when the base station directly supplies energy to the device, the communication distance is
relatively short, which is usually suitable for building a cell with a cell radius of tens of meters to 100
meters. Such kind of cell is suitable for covering logistics centers, storage stations, industrial plants
and other scenarios.
In order to further improve the network coverage and expand the IoT applications to more scenarios,
e.g., large industrial plants, agriculture and animal husbandry, it can be considered to deploy dedicat-
ed wireless power supply nodes to supply energy to Ambient IoT devices. As shown in case 3 in
Figure 5.1-1. The dedicated wireless power supply node is responsible for sending wireless power
supply signals when the base station needs to communicate with the device, so as to perform the
wireless power supply function of the network. With this kind of deployment, the two functions of
wireless power supply and Ambient IoT communication can be decoupled. The distributed wireless
power supply nodes in the cellular network provide wireless power supply, which alleviates the chal-
lenge of network coverage due to wireless power supply, so as to provide relatively large network
coverage. Furthermore, dedicated power supply node is mainly used for wireless power supply, so
its complexity and deployment cost will be much lower than that of base station. Therefore, it is a
relatively economic way to deploy Ambient IoT network with relatively large coverage.
For application scenarios such as smart wearable or smart home, it generally needs short-range
communication. For example, the coverage requirement of smart wearable network is less than 5m
and the coverage of smart home is generally about 10 meters. Therefore, as shown in case 2 in
Figure 5.1-1, the Ambient IoT network that is centered on smart phones or CPE nodes provides a
very attractive short-range personal communication network. The use of nodes such as smart
terminals or CPE with transmit power complying with the regulation requirements is also sufficient
to provide adequate Ambient IoT communication coverage (refer to Section 4.1).
From the perspective of wireless power supply signal, at least the following requirements shall be considered.
To provide sufficient For the passive device without a battery, when the wireless power supply signal
wireless power arrives at the device, its signal strength shall not be lower than a certain strength such
as -20dBm (when the device has no energy storage unit) or -30dBm (when the device
has energy storage unit).
Efficiency of wireless From the perspective of signal waveform, although any kinds of radio waveform can
power supply provide energy for the Ambient IoT consumption device, it can be further investigated
whether there are differences in the efficiency for different waveforms, and reason-
able waveforms based on that can be designed.
Stability of wireless When the Ambient IoT device is working, it is necessary to provide stable wireless
power supply power supply. Continuous sine waves can provide radio waves with stable power
because of their constant amplitude. In the case of wireless power supply using a
downlink signal carrying information, the power supply signal is a modulated wave-
form. Based on the coding of information bits, it is inevitable that the signal amplitude
changes. From power supply stability perspective, it is required that the power level of
the power supply signal cannot be too low for a long duration. Therefore, in order to
ensure the stability of wireless power supply, it requires that the modulated waveform
should be carefully considered when selecting coding and modulation scheme.
Compatibility with When the Ambient IoT system is deployed in the same frequency band with other sys-
other systems tems, the impact on other systems needs to be considered. For example, generally
the wireless power supply signal needs to be transmitted with high power, so the
interference to the adjacent system needs to be considered. When the cellular
system is deployed, especially when it coexists with other systems, the wireless signal
from other existing systems can be used for wireless power supply thus the source of
wireless power supply can be expanded. The base station or smart phone can per-
form wireless power supply without changing the waveform of the transmitted signal.
Therefore, from the perspective of wireless power supply, compatibility with other
systems is also worth studying.
Compatibility When the wireless power supply signal is used as the carrier wave for backscattering,
with backscattering it needs to provide enough radio energy. In addition, the impact on the modulation
during backscattering should also be considered. For example, when it uses ASK or
PSK for backscatter transmission, the sine wave signal of a single frequency is an
ideal candidate. However, when the power sourcing signal experiences amplitude or
phase modulation such as ASK or PSK modulation, it will produce complex and mixed
waveforms, which need to be carefully evaluated for its impact on the demodulation
at the receiver. For example, when FSK modulation is used in backscattering, it may
be necessary to maintain a stable frequency of the wireless power sourcing signal.
The introduction of the wireless power supply function will have an important impact on the man-
agement and allocation of wireless network resources. Similar as legacy resource dimensions, e.g.,
time domain resources, frequency domain resources and code domain resources in existing wireless
networks, wireless power has become a new resource dimension in Ambient IoT system. The net-
work nodes in Ambient IoT system can allocate (transmit) or schedule wireless power according to
the communication requirements, so that the battery-less device can still complete the wireless
communication function. From the perspective of energy management, the Ambient IoT system
changes the power supply from a distributed manner for traditional devices into the centralized
manner for Ambient IoT devices. From power consumption point of view, centralized energy supply
distributes and uses energy according to demand and it will make the energy use of wireless net-
work more efficient and avoid unnecessary energy waste when there is no communication. In the
future, we can further investigate how to maximize the advantages of centralized power supply in
Ambient IoT system so as to enable green and low-carbon communication networks.
Key PHY technologies

5.2 of Ambient IoT
5.2.1 Modulation and coding of Ambient IoT communication
With the development of wireless communication technology and the improvement of component
technology, more complex signal modulation can be used in new communication systems. For
example, in addition to supporting low-order modulation methods such as BPSK and QPSK and
high-order modulation methods such as 16QAM and 64QAM, ultra-high-order signal modulation
technologies such as 256QAM and even 1024QAM are adopted in LTE and NR systems[3]. Similarly,
the Forward Error Correction channel coding technology has developed rapidly, and Convolutional
code, Turbo code, LDPC code and polar code had been adopted in LTE and NR systems[2]. These
modulation and coding techniques play a key role in supporting LTE and NR to realize ultra-wide-
band and ultra-high speed data transmission.
In the existing technologies of IoT, such as MTC, NB-IoT and RedCap, although the terminal capabili-
ty is significantly reduced compared with LTE terminals or NR terminals, it basically inherits these
traditional modulation or coding methods. For example, MTC/NB-IoT can support modulation meth-
ods such as BPSK, QPSK and 16QAM, as well as Turbo and Convolutional codes, while RedCap can
also support BPSK, QPSK, 16QAM and 64QAM, LDPC and Polar codes.
However, these modulation and coding methods for ordinary terminals are great challenges for Am-
bient IoT devices. As described in chapter 3, an Ambient IoT device has a simple RF and baseband
structure, for receiving transmitting data in an ultra-low power mode. Therefore, these two aspects
will bring strong constraints to the signal modulation and coding methods available to Ambient IoT
devices. Specifically, the simple RF and baseband structures make it difficult for Ambient IoT devic-
es to process phase and amplitude modulation and de-modulation. That makes it hard to support
QPSK and QAM modulation. Despite the excellent signal encoding and decoding performance, For-
ward Error Correction channel coding methods such as Turbo, Polar and Convolutional are difficult
to support for Ambient IoT devices with ultra-low complexity and low power consumption.
As described in chapter 3, keying modulation technology can be well combined with backscattering
technology, so that the terminal realizes ASK, FSK or PSK modulation with extremely simple hard-
ware structure and realizes backscatter data transmission. Using keying modulation technology, the
Ambient IoT device only needs to adjust its circuit impedance, capacitance or phase delay on the
hardware to implement signal modulation and backscatter transmission. In the other direction, simple
signals such as ASK, FSK or PSK can also be received through simple hardware structure. For exam-
ple, as mentioned in chapter 3, the solution modulation of ASK can be processed through a compara-
tor. That avoids complex baseband signal processing and greatly reduces power consumption.
The channel coding of Ambient IoT devices also needs to match the hardware and software capabil-
ities of Ambient IoT devices. Therefore, more suitable coding methods are based on binary coding,
including: non-return to zero (NRZ) coding, Manchester coding, unipolar return to zero coding,
differential bi-phase (DBP) coding, miller coding, pulse interval coding (PIE) and other coding meth-
ods. These coding methods are simple in baseband processing and generally use high-low electrical
level conversion to represent 0 and 1, so they can also be well combined with simple modulation
methods such as ASK, FSK or PSK.
On this basis, we can further explore whether more complex signal modulation and coding methods
can be further supported under the conditions of minimal hardware and ultra-low power constraints,
such as QPSK.
5.2.2 Multiple access methods
For different application scenarios, the Ambient IoT system needs to support varying numbers of
terminal devices. Within the coverage area of the Ambient IoT network, multiple Ambient IoT devices
may be simultaneously activated and communicating with network devices. Therefore, without intro-
ducing a reasonable multiple access method, the backscattered signals from multiple Ambient IoT
devices on the same frequency may cause mutual interference. In this case, the network cannot
distinguish between different Ambient IoT devices and correctly demodulate the backscattered
signals of each device. Therefore, the Ambient IoT system needs to support efficient multiple access
methods. Common multiple access methods include Time Division Multiple Access (TDMA), Fre-
quency Division Multiple Access (FDMA), Code Division Multiple Access (CDMA), and Non-Orthog-
onal Multiple Access (NOMA). This section will discuss multiple access methods suitable for the Am-
bient IoT system.
TDMA can be considered as a candidate multiple access method for communication in the Ambient
IoT system. The Ambient IoT system divides multiple time slots in the time domain. Different Ambient
IoT devices can communicate in different time slots based on different time delays. Different time
slots isolate the UL signals of different Ambient IoT devices, avoiding mutual interference between
UL signals of the devices. TDMA only requires determining a certain time delay and communicating
based on this time delay. Therefore, for Ambient IoT devices, TDMA is a simple and feasible multiple
access method.
FDMA stands for Frequency Division Multiple Access, which involves dividing several channels with
different frequencies in the frequency domain, allowing different users to communicate using differ-
ent frequency channels. For the Ambient IoT system, it is necessary to explore the feasibility of using
the FDMA method. For example, the following aspects need to be considered:
Does the hardware of The prerequisite for using FDMA for multi-user multiplexing is that terminal devices
Ambient IoT devices have the ability to operate on different frequency channels. This seemingly simple
have the capability to requirement for ordinary terminal devices may pose certain challenges for Ambient
operate on different IoT devices. Constrained by the requirements of hardware complexity and ultra-low
frequency channels? power consumption, Ambient IoT devices are not easily equipped with high-precision
crystal oscillators and high-precision phase-locked loop. Therefore, generating
low-complexity and low-power consumption frequencies and, based on these
frequencies, implementing the ability to work on different frequencies will be an
important challenge for Ambient IoT devices.
Signal interference levels When Ambient IoT devices send signals to network devices via backscattering, due to
between Ambient IoT the complexity of devices and circuits, backscattered signals generally do not undergo
devices on different complex shaping filtering before leaving the Ambient IoT devices. Therefore, potential
frequency channels adjacent channel interference may occur for Ambient IoT devices on adjacent chan-
nels. How to ensure reliable communication on multiple channels while allowing for the
existence of adjacent channel interference is also a worthwhile issue to explore. For
example, it may be considered to reasonably reserve frequency guard bands between
channels and design receivers reasonably.
Influence on the channel Ambient IoT communication is a passive communication method, where Ambient IoT
access process of Ambi- devices need to obtain energy through energy harvesting to drive the circuit into an
ent IoT Devices active state before communicating with the network. Therefore, unlike traditional
communication methods where communication processes can be initiated by both
network devices and terminal devices, in Ambient IoT communication, communication
generally needs to be initiated by network devices. Furthermore, unlike traditional
communication processes where terminal devices actively search for networks
through cell search processes, Ambient IoT devices remain in a powered-off state
until they receive wireless energy provided by the network, thus only becoming acti-
vated when network wireless energy is received.
When different Ambient IoT devices support operation on different frequency chan-
nels, it becomes necessary to consider how network devices can efficiently trigger
the communication process. For example, efficient searching of the channels where
Ambient IoT devices dwell and subsequent communication processes, as well as pro-
viding appropriate energy supply, are important issues for network devices.
CDMA is a commonly used multiple access method in traditional communication sys-

tems, allowing multiple users to communicate using orthogonal multiplexing codes on
the same time-frequency resources. Therefore, CDMA communication possesses
advantages such as strong anti-interference capability, robustness against fading,
good secrecy, and large system capacity. For Ambient IoT communication, adopting
CDMA could be beneficial for eliminating interference between users, effectively
increasing system capacity, and enhancing communication performance. Particularly,
in the case of uplink transmission in Ambient IoT communication, whether using back-
scatter or active transmission methods, the signal power level is generally weak. Using
CDMA, i.e., spread spectrum communication, can also help improve the communica-
tion coverage distance.
However, adopting CDMA transmission in Ambient IoT communication may encounter

significant technical challenges. For example, CDMA systems have high requirements
for synchronization among multiple terminal devices and impose limits on the power
difference of received backscattered signals from multiple Ambient IoT devices. Addi-
tionally, constrained by the ultra-low complexity, Ambient IoT devices lack high-precision
clock generation capability, resulting in weaker synchronization capability. When using
backscattering, flexible power control is also difficult to achieve. Therefore, effective
implementation of CDMA transmission requires further exploration.
5.2.3 Data transmission and resource management
The traditional cellular mobile communication system supports flexible duplex mode,
flexible multiple access mode and data transmission of various types of services. There-
fore, data transmission supports flexible and efficient resource management and alloca-
tion. For example, the traditional cellular mobile communication system can support
resource allocation with variable time, frequency, code and even power domain resource
granularity. From the perspective of the signaling for resource allocation, it can support
dynamic and semi-static scheduling grant (for example, NR supports configured grant
transmission). In addition, the system can flexibly schedule the data transmission of termi-
nals according to the system load, the number of terminals, the UE type and priority of
services.
As described in Chapter 2, in the deployment scenario of Ambient IoT, a large number of

terminal deployments can be realized due to its advantages of low cost and battery free.
The traffic of Ambient IoT is characterized by uplink dominated and small data transmis-
sion. For different Ambient IoT scenarios, there are different data transmission require-
ments and characteristics. For industrial monitoring scenarios, it is mainly used for indus-
trial data reporting and environmental monitoring and the traffic is often periodic. For
smart home services, such as home asset management and switch control, the traffic is
often one-shot and bursty. For logistics and warehouse scenarios, the network needs to
obtain a large amount of information of Ambient IoT devices in a short time duration. In
this case, the resource allocation and scheduling need to support a large number of Am-
bient IoT devices to transmit within a short time duration.
For different scenarios of the Ambient IoT system, the methods of data transmission are
also different. Data transmission in the Ambient IoT system can be roughly divided into
three types: Device-terminated (DT) DL data transmission, Device-originated-autono-
mous (DO-A) data transmission initiated by the terminal and Device-originated-de-
vice-terminated triggered (DO-DTT) data transmission.
Device-terminated This type of data transmission usually involves a network or other device sending data to
(DT) data transmission an Ambient IoT device, but the Ambient IoT device does not need to send data. For
example, in the scenario of smart control of home equipment, instructions can be sent
through the mobile phone to control the turning on or off of home appliances. At this
time, you only need to send commands to the home equipment through the mobile
phone. Generally, it is not necessary for the home equipment to send data. In some sce-
narios, in order to ensure that the terminal device receives the instruction correctly, the
terminal device can send information indicating whether the instruction message is
received correctly. In DT type data transmission, since it is mainly downlink data trans-
mission, it is necessary to ensure that the Ambient IoT device can receive the downlink
data sent by the network. The downlink data sent by the network can be sent through
broadcast, multicast or unicast.
Device-originated- Data is generated on the Ambient IoT device, and data transmission is initiated auton-
autonomous (DO-A) data omously by the Ambient IoT device. For example, in the smart home scenario, an Am-
transmission bient IoT sensor is placed in the kitchen to monitor whether there is a gas leak. When
the gas concentration exceeds a threshold, the Ambient IoT sensor actively initiates
data transmission, such as initiating an alarm or sending data to the house owner
through the network. For another example, in the smart grid scenario, Ambient IoT
sensors are used to monitor the temperature, humidity, pressure, vibration and other
data of the power grid system, and periodically report the monitoring data to the net-
work, thereby determining whether the entire power grid system is working normally.
In DO-A type data transmission, data transmission is triggered by an Ambient IoT
device. It can be sent based on event triggers or periodically. In order to ensure that
the Ambient IoT device can send data quickly, it needs to support fast access mecha-
nism, or allocate semi-static transmission resources to Ambient IoT devices, so that
Ambient IoT devices can obtain corresponding transmission resources when they
have data to transmit.
Device-originated – This type of data transmission is triggered by the network signaling to trigger the Am-
device-terminated bient IoT device to perform uplink data transmission. For example, in logistics and
triggered (DO-DTT) data warehouse scenarios, when goods arrive at the warehouse, new goods need to be
transmission registered for goods stocktake so that the system can determine which goods are
stored in the warehouse. At this time, the network sends a trigger command, and the
Ambient IoT device reports identification information to the network based on the
command information, so that the network side can maintain and update the invento-
ry list. In DO-DTT type data transmission, when the network sends trigger signaling,
there are usually a large number of Ambient IoT devices that need to transmit uplink
data at the same time in a short time. At this time, how to avoid collision, interference
among Ambient IoT devices needs to be resolved.
Because the service type, energy supply and terminal characteristics of Ambient IoT
are different from those of traditional cellular mobile communication systems, it
imposes new challenges to the data transmission and resource management of Am-
bient IoT.
Challenge 1: For traditional terminals, data transmission depends on communication requirements

Restriction of power and network resource scheduling. The energy supply does not need to be considered
supply for Ambient in the procedure of data transmission. For the Ambient IoT device, any data transmis-
IoT communication sion of the terminal depends on external power supply. The stability and availability of
wireless power supply, energy storage status and energy storage capacity will affect
data transmission. Resource allocation and scheduling need to consider the impact of
these factors. The wireless power supply shall guarantee the reliable transmission of
data. In addition, the resource overhead of wireless power supply on the network side
shall be as small as possible to realize on-demand wireless power supply. Therefore,
from the perspective of resource management, the resource allocation of Ambient IoT
system needs to avoid the interference between users, inter cell interference and the
interference of different systems on Ambient IoT, so as to make the Ambient IoT under
ideal conditions, ensure the success of data transmission and avoid data retransmis-
sion as much as possible. On the other hand, the data transmission of Ambient IoT
also needs to rely on the reasonably designed power supply signal and efficient coop-
eration between the power supply process and the data transmission process.
Challenge 2: For example, for logistics scenarios, data reading and reporting from thousands
Data transmission or even tens of thousands of Ambient IoT devices need to be completed in sec-
from a large number onds. In such scenarios, how to reasonably control the access of a large number
of terminals in a very of terminals is a problem to be solved. Different terminals should transmit in an
short time orderly manner on different resources, while avoiding the collision of data from
different terminals and mutual interference. In addition, in such scenarios,
before the Ambient IoT device is triggered by the network, all terminals are
unknown to the network. A reasonable system access and communication pro-
cedure is needed to enable a large number of Ambient IoT devices to be quickly
identified and scheduled for efficient data transmission. For this reason, in the
Ambient IoT system, it is necessary to consider multiplexing as many users as
possible with the permission of terminal capabilities. Typically, several multiple
access modes can be supported. Besides TDMA, FDMA and CDMA can be
considered to provide sufficient system capacity for Ambient IoT. As shown in
Figure 5.2-1, data transmission for different Ambient IoT devices are effectively
allocated to different resource units to avoid inter-user interference. A reason-
able access control mechanism needs to be introduced to allow efficient
resource management while guaranteeing the requirements of services.
Unit6 Unit12 Unit18 Unit24

Resource Units in frequency domain

F_unit
T_unit
Resour Units in time domain
Figure 5.2-1 Resource allocation of Ambient IoT

Challenge 3: Firstly, the resource utilization capacity of the Ambient IoT device has an impact on data
Impact of ultra-low transmission. In traditional cellular systems, flexible resource allocation can meet the
cost and complexity requirements of different service type with diverse peak rate, latency and so on. It allows
of Ambient IoT higher resource utilization and spectral efficiency. For Ambient IoT devices with extremely
devices on data low cost and complexity, the flexibility of resource allocation that they can support will be
transmission greatly reduced due to the limitations of transmission bandwidth, communication time,
power supply, spectrum shift, time-frequency synchronization, etc. Compromise is required
between the flexibility of resource allocation and the capability of Ambient IoT devices. Net-
works can allocate relatively less-flexible resources for Ambient IoT devices. As shown in
Figure 5.2-1, the network configures multiple relatively fixed resource units for Ambient IoT
devices in a certain mapping manner. The frequent changing of the mapped resource units
should be avoided for an Ambient IoT device to reduce the complexity of resource allocation.
Secondly, the weak synchronization capability of Ambient IoT device also has impacts on
data transmission. In the traditional cellular mobile communication system, in order to
reduce the interference between multiple users and the interference between uplink and
downlink in TDD system, the transmission needs to be under a stringent requirement of
synchronization. Traditional terminals, such as smartphones, MTC devices and other IoT
devices, can meet the synchronization requirements of cellular systems. For Ambient IoT
devices with ultra-low cost and complexity, it is generally not possible to be equipped with
an oscillator with high accuracy. A simple oscillator with small size and low power consump-
tion, such as RC oscillators, is usually used to provide the clock. However, the accuracy of
RC oscillators is poor, with errors up to 1% or even higher. Whether the clock accuracy with
such large error can meet the timing requirements of Ambient IoT and how to adapt to that
is a problem that needs further study.
Further, for Ambient IoT devices, they don’ t have batteries. The energy needed for its
circuit to work comes from wireless power supply. Because of the instability of wireless
power supply, even if the Ambient IoT device has a simple oscillator, the oscillator will stop
working if wireless power supply is not available. The oscillator cannot work continuously to
provide a stable and continuous clock. Regarding the effect of time-frequency synchroniza-
tion on data transmission, please refer to section 5.2.4.
5.2.4 Time-frequency synchronization
The accuracy of time-frequency synchronization achieved by the terminal equipment has

certain influence on the design of time-frequency multiplexing of resources and the perfor-
mance of data decoding. In the pursuit of high data throughput and transmission rate com-
munication systems (such as cellular communication systems, IEEE 802.11ac systems, etc.),
OFDM is usually used. OFDM technology requires strict synchronization between subcarri-
ers and is very sensitive to timing deviation and frequency deviation. For example, in the 5G
cellular system, the timing deviation of the user device is required to be preferably within
the cyclic prefix (CP) range to avoid inter-symbol interference; The precision of the crystal
oscillator is required to be ±0.1ppm to ensure the orthogonality [1] between the subcarriers
to avoid intercarrier interference. Correspondingly, the 5G cellular system has also
designed a relatively complete and complex synchronization mechanism. During the initial
access process, user devices can obtain the initial time-frequency synchronization through
the 127 bit Primary Synchronization Signal (PSS) and Secondary Synchronization Signal
(SSS). In the follow-up communication process, the Tracking Reference Signal (TRS) can be used to
track and compensate the time-frequency deviation, which can meet the time-frequency synchroni-
zation requirements of the system. To achieve the reception of these synchronous signals the termi-
nal device needs to have complex signal processing capabilities such as high sampling rate, IQ
demodulation, FFT transformation, and so on, which cannot be applied to the Ambient IoT systems.
For systems that communicate with simpler waveforms, the synchronization requirements can be
relaxed. For example, the IEEE 802.11ba system using MC-OOK waveform can achieve a timing devi-
ation of ±250ns or even 0ns [2] in most cases through a synchronous reference signal with a
sequence length of 32 bits, which is less than the MC-OOK symbol length (the symbol length is 2us
in high data rate scenarios and 4us in low date rate scenarios), which does not have much negative
impact on subsequent date communication. Such synchronization mechanism does not require the
terminal equipment to carry out complex signal processing, and has important reference signifi-
cance for the synchronization scheme design of Ambient IoT. Specific parameter design, including
the structure of synchronization signal, type of synchronization sequence, sequence length, modula-
tion and coding mode, needs to be adapted according to the actual synchronization needs, while
taking into account the delay and complexity of synchronization process. Generally, the longer the
synchronization reference signal, the higher the synchronization accuracy, but the corresponding
detection delay and complexity will also increase.
Although the synchronization mechanism of IEEE 802.11ba system can recover timing information
more accurately, it cannot realize frequency synchronization. The WUR (Wake Up Radio) terminal
device uses a low-power wake up Radio device to monitor the WUR data sent by the network and
maintain the basic connection, and wakes up the Main Radio (MR) with high time-frequency syn-
chronization accuracy to send and receive data when necessary. Since WUR only involves the
reception of DL signal, and DL signal adopts MC-OOK waveform, WUR terminal device can use
envelope detector for non-coherent demodulation without requiring high precision frequency syn-
chronization. In some practical applications, even if the center frequency of WUR is not accurately
synchronized with the network, WUR can still receive DL signals correctly with some loss of receiver
sensitivity. In addition, since WUR acts as an auxiliary receiver, it is possible to share the power
supply and even the clock circuit with the main receiver. Therefore, on the one hand, WUR can toler-
ate a larger power budget to achieve a more accurate clock circuit, and on the other hand, it can also
obtain a more accurate clock by sharing the clock circuit of the main receiver. However, frequency
offset is a problem that cannot be ignored for Ambient IoT.
Different from the WUR device, in order to strictly control the complexity and power consumption of
the Ambient IoT device, the Ambient IoT device is only equipped with a set of RF devices with weak
capabilities. It usually contains a simple local oscillator or even no oscillator, and based on such RF
devices, it is necessary not only to complete the downstream data reception, but also to achieve the
uplink data transmission. For downlink data reception, the above MC-OOK waveform, which is rela-
tively insensitive to frequency bias, can be used. For uplink data transmission, especially for Ambient
IoT devices using active transmission mode, large frequency offset may cause uplink transmission to
deviate far from the owning channel and interfere with the signals in the adjacent channels, which
greatly reduces the reliability of uplink data. Therefore, for the Ambient IoT system, it is necessary to
design a reasonable frequency synchronization scheme, so that the Ambient IoT device can achieve
sufficient frequency synchronization accuracy under the condition of low power consumption. For
example, the terminal device calculates the phase rotation based on repeated reference signals to
determine the frequency offset [3]. On this basis, a small amount of frequency deviation can be toler-
ated by appropriately reserving frequency protection bands in the UL channel, which slightly reduc-
es the frequency spectrum usage efficiency (generally speaking, for licensed frequency bands,
spectrum efficiency is an important indicator of system design), but it can reduce the power con-
sumption and device complexity of Ambient IoT devices.
In addition to the above time-frequency synchronization functions, the design of the synchronization
signal of the Ambient IoT system can also have other functions, so that the ambient IoT device can
obtain some other important information while detecting the synchronization sequence. For exam-
ple, the synchronization signal of the existing 5G cellular system carries the Physical Cell Identifier
(PCI) for cell identification through 1008 combinations of PSS and SSS sequences; The IEEE 802.11-
ba system indicates the data transfer rate through two different synchronization sequences. In the
process of standardization discussion, whether the Ambient IoT system needs to synchronize signal
carrying information and what information it carries needs to consider factors such as network
deployment, date channel design, and business requirements.
5.2.5 Positioning technology for the Ambient IoT communication system
At present, the mainstream positioning technology for satellite system and cellular system can
achieve high accurate positioning in outdoor scenarios, while indoor positioning has not been effec-
tively solved due to complicated environmental factors such as multipath and occlusion. However,
positioning function is indispensable in some typical application scenarios, such as smart parking lot,
warehouse and logistics, smart home and so on. This subsection will analyze the following traditional
positioning methods one by one, and screen out the potential solutions suitable for the Ambient IoT
system.
TDOA (Time Difference of Arrival)

the basic principle is to construct multiple hyperbolic constraint equations through the propagation
time difference between multiple known reference nodes and the target node, and then position of
the target node could be estimated.
Multi-RTT (Multi-Round Trip Time)

the basic principle is to construct multiple distance constraint equations through the round-trip time
difference of signal transmission between multiple reference nodes and targets node, and then
position of the target node could be estimated.
Angle-based positioning method (such as AoA/AoD)

the basic principle is to construct triangular equations by referring to the angle information
between the reference nodes and the target node, and then the position of the target node could
be estimated.
Cell ID-based positioning method (CID)

the basic principle is to select a serving cell according to the signal strength. Considering that the
signal strength is greater with closer distance, the position of the selected cell could be considered
as the estimated position of the target node.
Phase-based positioning method

the principle is to estimate the propagation distance based on the phase change of signal between
reference nodes and the target node, and then the position of the target node could be estimated.
The performance of any positioning method depends on the related measurement accuracy. Both
TDOA and Multi-RTT positioning methods need to measure the time of arrival, of which the accura-
cy is positively correlated with the signal bandwidth and sampling clock rate at Rx side. The larger
the signal bandwidth and the higher the sampling rate, the more accurate the corresponding time
measurement is. The resolution of angle-based measurement is related to the number of antennas
and their arrangement. Theoretically, the higher the number of antennas, the higher the resolution of
angle-based measurement. Moreover the angle information measured by two-dimensional antenna
array is more abundant than that measured by one-dimensional linear antenna array. However, nei-
ther large bandwidth nor high sampling rate can be implemented in Ambient IoT devices. For net-
work equipment, configuring antenna array will also increase deployment costs. In some scenarios,
such as using a smart mobile phone to locate Ambient IoT devices, it is difficult to deploy antenna
array within the smart mobile phone. In summary, the positioning method based on time measure-
ment is not suitable for Ambient IoT systems, and the application scenarios of angle-based position-
ing method will be quite limited.
Cell ID-based positioning method is simple and easy to implement. The device is only required to
measure the signal strength, such as RSRP or RSSI, of multiple candidate base stations. According
to the measurement results, the base station with the best signal quality will be selected and its posi-
tion will be considered as the estimated position of the device. Although the measurement accuracy
of signal strength such as RSRP is highly related with the number of reference signal resources, it
could be improved by multiple measurements in the time domain rather than mandating a larger
bandwidth for the reference signal. Due to the large coverage of a cellular cell, which could be hun-
dreds or even thousands of meters, large positioning error obvious exists when using cell ID-based
positioning method. Fortunately, the low-cost characteristic of the Ambient IoT devices makes it pos-
sible to deploy massive devices as positioning reference tags with high density, such as at intervals of
2 meters or even denser for the indoor scenario, thereby significantly improving positioning accuracy.
Figure 5.2-2 shows a positioning method similar as the above-mentioned cell ID-based positioning
method. The target device will measure the signal strength of multiple Ambient IoT reference tags,
select the reference tag with the highest signal strength and take its position as the estimated posi-
tion. This positioning method is suitable for the indoor navigation case. Users can determine their
own position according to the reference tags around them, and determine the approximate walking
direction and route according to the target position and passing reference tags.
The other typical positioning scenario is object search, in which the handheld mobile phones and the
network devices could measure the signal strength of the target device and the reference tags
respectively, and the position of reference tag whose measurement result is the nearest to that of
the target device is the estimated position. To further improve the reliability and accuracy of position-
ing, the first k nearest reference tags can be selected, and the K-Nearest Neighbor (KNN) algorithm
can be used to determine the position of the target device [4]. Figure 5.2-3 shows the Cumulative Dis-
tribution Function (CDF) of positioning errors by using the KNN algorithm when the reference tags
are deployed at intervals of 4m, 2m and 1m respectively. It shows that the positioning accuracy of
6m could be achieved in ideal conditions.
（x0,y0,RSRP0）（x4,y4,RSRP4）（x8,y8,RSRP8）（x12,y12,RSRP12）
Reference tag
with the best RSRP
moving trajectory target position
the measured RSRP

values of reference tag
#1/2/5/6 are the closest
to the measured RSRP （x1,y1,RSRP1）（x5,y5,RSRP5）（x9,y9,RSRP9）（x13,y13,RSRP13）
value of target measurement
node
Figure 5.2-2 Signal strength-based positioning method with reference tags

ideal SNR, random LOS states

1
X:2.333
Y:0.9
0.9
X:0.6357 X:2.934 X:6.145
Y:0.9 Y:0.9 Y:0.892
0.8
0.7
0.6
0.5
F(x)
0.4
0.3 ref Tag spacing = 4m, Los corr dist = 1m

ref Tag spacing = 2m, Los corr dist = 1m
0.2 ref Tag spacing = 1m, Los corr dist = 1m
0.1
0
0 2 4 6 8 10 12 14 16 18
x
Figure 5.2-3 CDF of positioning errors for signal strength-based positioning method
High accurate positioning can be achieved based on phase information, which is currently widely used in Bluetooth and
satellite positioning systems. Similarly, the measurement accuracy of the phase information will also affect the final posi-
tioning performance. As shown in the figure 5.2-4, phase measurement is performed by comparing the received signal y
with the local signal x to determine the phase difference between them. In theory, the positioning reference signal in this
method can be a simple sine wave, which is very suitable for the narrow band characteristics of Ambient IoT systems. A
tricky problem for phase-based positioning method is how to solve the influence of random initial phase. The difference
between the initial phase θTx and θRx at the Tx and Rx sides will cause severe deviation in phase difference measurement,
so it is necessary to eliminate the influence of random initial phase.
This problem can be avoided if backscatter transmission is adopted by the Ambient IoT device, where the network
device sends and stores the signal x , and the Ambient IoT device will not generate its own signal, but backscatter the
signal from the network. At this time, the phase difference φ of y received by the network device compared with the
2π 2d
original signal x is related to the distance d and the frequency f , that is, φ = c* * f . Because of the 2π periodicity of
phase, only the fractional part can be measured and the whole cycle part cannot be obtained in practice. Then, the
range of single-frequency measured distance is half wavelength 2λ = 2f c
. Even with low frequency such as f=1GHz, the
range is only 15cm, which is very limited. In order to solve the problem caused by the ambiguity of the whole cycles, the
dual-frequency phase difference method [5] can be used by measuring the phase change {φ1,φ2 } of two frequencies
{f₁,f₂ } . In this case, the phase difference ∆φ=φ1-φ2 depends on the distance and the frequency difference ∆f=f1-f2 , that
is, ∆φ = 2πc* 2d *∆f. As long as the frequency difference is small enough, the equivalent half-wavelength ∆λ = c could be
2 2∆f
increased, thereby extending the distance range. For example, the distance range could be up to 150 meters when ∆
f=1MHz, satisfying the demands of Ambient IoT positioning scenarios with medium and short distance. Figure 5.2-5 indi-
cates the positioning error for phase difference based positioning under the case that the coverage radius of positioning
gNB is equal to 70m. According to the simulation result, the accuracy of the phase difference ranging with dual-frequen-
cy can reach less than 6 meters. With the distance between target node and multiple reference nodes, it will be easy to
estimate the absolute position of the target node by using the classic multi-lateration positioning algorithm, thus realizing
indoor navigation and object search functions.
The traditional phase positioning model is as follows:
Rx
The signal sent y = cos ( 2πft + φ + θTx )
by transmitter ＇Phase difference＇
associated with
cos ( 2πft + θTx ) distance d 本地参考信号
2π * d x = cos ( 2πft + θRx )
φ= f
c *
Illustration of traditional phase-based ranging
Local signals sent and

saved by the Reader
Signal received by Tag
cos ( 2πft + θTx )
y’ = cos ( 2πft + φ’ + θTx )
Backscatter signal 2π * d
Received by Reader φ’ = *f
distance d c
y = cos ( 2πft + φ + θTx ) Tag backscatter based
on the received signal
2π * 2d
φ= *f
c
Illustration of phase-based ranging with backscattering
Figure 5.2-4 Illustration of phase positioning model
Empirical CDF
1
X:5.158
Y:0.9
0.9
X:6.534
Y:0.9
0.8
0.7
0.6
0.5
cdf
0.4
0.3
0.2 Tx power = 30dBm, 1 PRS sym, = 10RB

Tx power = 0dBm, 1 PRS sym, = 10RB
0.1 Tx power = 30dBm, 1 PRS sym, = 5RB
Tx power = 0dBm, 1 PRS sym, = 5RB
0
0 2 4 6 8 10 12 14 16 18
distance error[meter]
Figure 5.2-5 CDF of positioning error for phase difference based positioning
Requirements and challenges of

5.3 lightweight protocol stacks
In the Ambient IoT system, the Ambient IoT device needs to collect energy in the environment (such
as radio-frequency energy, luminous energy, thermal energy) before it is capable to work. Therefore,
an Ambient IoT device keeps in ‘power-off’ state, i.e., out-of-service, before energy is obtained. This
section will introduce the lightweight protocol stacks provided that the Ambient IoT device collects
the RF energy from the radio wave sent by network.
In some use cases, islanding is observed in the deployment of Ambient IoT systems, where full cov-
erage cannot be achieved. As a result, Ambient IoT devices are likely to be out-of-service due to lack
of coverage, e.g., in logistics and warehouse scenarios.
Therefore, Ambient IoT devices are disconnected from network when there is no power source or
out of coverage. Under such harsh conditions, smaller processing delay, lower memory consumption
and more efficient data transmission schemes are required in the Ambient IoT system to enable
faster data communication process.
5.3.1 State management
In traditional communication systems, service types of devices are complex, diverse, generally
requiring service continuity and large data volume. According to whether there is data to be trans-
mitted, network adopts different strategies in resource allocation and device management. In this
regard, RRC states and NAS states are defined. When a UE is in RRC_IDLE state, it is essential to
support mobility management and paging reception, while dedicated radio resources are not allo-
cated for devices since there is no requirement of data transmission. When a UE goes into RRC_-
CONNECTED state, dedicated resources shall be configured by network in order to perform uplink
and downlink data transmission. For subscribed users, NAS procedure is used by the core network
to facilitate management, while RRC procedure is used by base stations.
Considering the characteristics of small memory, low processing capacity, battery-less, small data
transmission and massive deployment, traditional multiple-layered protocol stack and complicated
state management are no longer suitable for Ambient IoT devices.
For logistics and warehouse scenarios, complicated state management and transition procedures
are not needed for devices that only needs to send single-packet data. Therefore, state-less concept
is more beneficial for efficient small data transmission, which is also helpful to reduce the cost and
complexity.
5.3.2 Lightweight protocol stack and efficient data transmission
The massive deployment of Ambient IoT devices makes the allocation of IP address one of the bot-
tlenecks. The potential solution is to support non-IP data transmission which can not only simplify
the data communication process, but also avoid unnecessary session management procedure.
In traditional communication systems, with the assumption that devices are with sufficient power,
multi-layer protocol stack is defined to realize functional modularization. For example, SDAP is used
to map QoS flow to DRBs. PDCP is used for header compression, data security, data delivery, etc.
RLC realizes ARQ, data segmentation etc. MAC is for data multiplexing and de-multiplexing. Modu-
larized multi-layer protocol stack is designed for diverse QoS requirements, complicated service
types and large amount data transmission.
In Ambient IoT communication, due to the dependence on power supply and the limitation of device
capacity (such as small memory and low calculation capability), lightweight protocol stack shall be
designed to reduce the power consumption and computation complexity. In addition, small-volume,
infrequent and delay-insensitive data is supposed to be supported by Ambient IoT devices. Accord-
ing to the above requirements and characteristics, flattening and sinking protocol stack is worth
considering. Meanwhile, integration of control plane and user plane can further reduce the complexi-
ty of UE and accelerate the procedure of data transmission.
In addition, Ambient IoT devices can simplify the protocol stack structure adaptively according to
different use cases. The demand for data transmission for Ambient IoT devices is relatively small. For
example, in the warehouse and logistics scenario, the Ambient IoT device only needs to report its
identification or collected information. Hence complicated protocol stack becomes unnecessary. For
this kind of Ambient IoT device requiring simple transmission of its own state information, session
management is not needed except for mobility management. For the transmission of small data
such as reporting its own state, data transmission can be simply realized via mobility management
stratum. Considering the different requirements of mobility, we can further study how to reduce the
protocol stack on mobility management.
ZP-RAN ZP-CN
UE NB CN
IOT service
RRC NGAP
IOT data
Radio NG-C
Figure 5.3-1 Lightweight protocol stack
Data transmission can be carried out based on non-state transition and non-dedicated bearer. The
identification of Ambient IoT devices can be used for data receiving and forwarding, which will
further simplify the data transmission procedure and achieve efficient data transmission.
Furthermore, various service types exist in the Ambient IoT system, such as logistics, warehouse,
smart wearables, etc. Different service types may have different requirements on protocol stack for
data transmission. To be specific, in logistics or warehouse scenario, data segmentation and
(de-)multiplexing are not needed since the data amount is extremely small, single packet data trans-
mission is expected, and the time interval of data traffic is long. Another example is smart wearable
where data has the characteristics of continuous transmission, so it is necessary to consider
sequential delivery of data packets.
Therefore, flexible and adaptive protocol stack is promising for the Ambient IoT system.
5.3.3 Mobility management
For different scenarios with different UE mobility characteristics and service QoS requirements, the
mobility requirements of Ambient IoT devices are different. For most of scenarios, such as industrial
wireless sensing, smart home, warehouse, etc., Ambient IoT devices are generally stationary where
no mobility requirements are observed.
However, mobility management is essential for scenarios such as logistics, tracking, manufacturing
monitoring. As an example, on a production line, the network needs to communicate frequently with
mobile Ambient IoT devices. Compared with the traditional NR system, if there is no delay require-
ment and lossless requirement, mobility management with limited functions is enough.
In conventional communication systems, mobility can be controlled by either UE or network. Net-

work-based mobility can guarantee data continuity. UE-based mobility can realize load balancing
and enable proper cell selection. No matter which kind of mobility, the mobility management is com-
pleted in real time based on the measurement results of downlink reference signal transmitted by
the network.
In Ambient IoT communication, devices may be out-of-service easily due to unstable power supply.
Therefore, it is difficult to get location information for mobility management as in traditional commu-
nication systems, e.g., with periodic location updating.
Therefore, the requirements on mobility management for Ambient IoT devices needs to be investi-
gated and how to execute necessary mobility management needs further study.
The requirements and challenges of

5.4 lightweight security mechanisms
Similar as in other IoT scenarios, trusted access and secure transmission remain important for Ambi-
ent IoT. However, due to the extremely limited hardware and software capabilities, very small
memory capacity, ultra-low power consumption and the dependence on external power supply, it is
difficult for Ambient IoT devices to use complex security mechanisms as in LTE or NR. Therefore, a
lightweight security mechanism suitable for Ambient IoT needs to be studied.
5.4.1 Trusted access and secure transmission

for resource-limited device
The value security and trustworthiness of networks and devices will be highly recognized in 6G era.
For personal devices, smart home devices, or industry devices, trusted access and secure transmis-
sion are expected when performing Ambient IoT communication.
The security mechanism in the 4G/5G era guarantees the above two security requirements[6][7] ：
1） In order to ensure trusted access, the terminal side uses the pre-stored 256-bit root key K in the
USIM card to calculate the authentication vector, IK and CK, using the f1-f5 function. Based on the
authentication vector, the terminal and the network are mutually authenticated during the authenti-
cation process.
2） In order to ensure secure transmission, the terminal uses the KDF function, IK and
CK in the ME to generate dozens of secret keys according to the requirements of differ-
ent communication scenarios, and use these secret keys to perform operations to
ensure the confidentiality and integrity of the data and signaling transmission.
However, in the Ambient IoT scenario, the computing and transmission resources sup-
ported by Ambient IoT devices are very limited. The traditional security mechanism is
challenging for Ambient IoT devices due to resource constraints. It is necessary to
research on how to provide trusted access and secure transmission under limited
terminal resource conditions.
5.4.2 The security requirements of low cost

and distributed scenario
Ambient IoT communication needs to support uplink data transmission, such as report-
ing information from Ambient IoT devices themselves or collecting information, which
can be initiated by devices or based on control commands from network devices.
Based on these two characteristics, the security threats of Ambient IoT can be divided into:
Downlink threat: If a fake base station mis-triggers an Ambient IoT device to perform uplink data trans-
false trigger mission, or a forged trigger signaling mis-triggers an Ambient IoT device to perform
uplink data transmission, the possible harm may not only lead to waste of transmission
resources and energy, but also cause data leakage or even user privacy disclosure.
Uplink threat: If the uplink transmission data is eavesdropped or leaked, malicious attackers may
data leakage obtain sensitive data or personal privacy data, which not only compromises the rights of
the data owner, but also may cause compliance risks or legal risks for business opera-
tors or network operators.
Eavesdropper
Uplink Threats: BS
Eavesdropping or temper
Fake BS
Tag Tag
Downlink Threats:
Fack BS or forged signaling
Figure 5.4-1 Security threats of Ambient IoT
The advantages of Ambient IoT devices are small size, light weight, low cost thus there
is a wide range of application scenarios. Ambient IoT can be widely used in scenarios
such as smart homes, logistics, manufacturing, personal wearable, etc., providing
trillions of links for the Internet of Things.
Therefore, the security requirements for Ambient IoT can be divided into:
Low-cost security In the Ambient IoT scenario, terminal equipment tends to be extremely simplified, with
requirements extremely low power consumption, and the cost, protocol stack and computing capa-
bility will be extremely reduced compared with current terminals. Therefore, it is nec-
essary to research on the low-cost trusted access and secure transmission schemes
that are compatible with extremely low-complexity terminal capabilities. From the
terminal side, it is necessary to simplify the authentication calculation, the key man-
agement scheme, confidentiality and integrity protection calculations.
The requirements of For logistics/manufacturing and other industry scenarios, Ambient IoT devices are
distributed/scenario- connected within a specific factory area, which needs distributed authentication.
diversified authentication Meanwhile, the business authorization for different scenarios as mentioned above
and authorization shall be considered.
5.4.3 Possible security way forward for Ambient IoT
Facing the massive number of links and devices for the 6G Internet of Things, efficient
distributed authentication and authorization need to be redesigned on the current
centralized trust mechanism to ensure trusted identity management, flexible authori-
zation and distributed authentication. Block chain is a feasible technology choice,
meanwhile the infrastructure construction and ecosystem maturity are required to
support multi-scenarios, multi-services, and multi-users trusted security mechanisms.
For Ambient IoT devices, trusted identity management and reliable secure transmis-
sion are mandatory to ensure rights and protections for business, networks, and users.
It is necessary to consider the characteristics of ultra-low cost and ultra-low complexi-
ty in order to optimize the transmission security mechanism based on traditional secu-
rity mechanisms, study the hierarchical protection mechanism of data transmission,
and research on the enhanced security scheme that combines the physical layer and
the transport layer.
Requirements and challenges to

5.5 simplify network architecture
On the one hand, in many Ambient IoT scenarios, it is required for the network to be able to manage
the Ambient IoT devices flexibly and efficiently, so as to enable and guarantee the realization of net-
work functions (such as remote control, remote positioning and environmental monitoring) and facili-
tate the operation of Ambient IoT business. On the other hand, it is difficult to reuse the existing
complex network architecture for Ambient IoT devices with minimal terminal capacity, very low
power requirements and dependence on wireless power. Therefore, a simplified network architec-
ture suitable for Ambient IoT needs to be studied.
5.5.1 Requirements to the simplify network architecture
With the increase in the number of users and the use of a large number of IoT devices, the network
scale has been rapidly expanded, and business requirements have also shown diversified character-
istics, which has resulted that network architectures become more and more complex. However, the
existing complex network architecture is no longer adapted to the characteristics of Ambient IoT. The
main reasons are as follows:
1） Complex network architecture will bring high operating costs to Ambient IoT, thus hindering the
development of Ambient IoT.
2） Complex network architecture will also affect the power consumption of Ambient IoT devices,
which brings new challenges.
3） The complex network architecture also makes the network deployment complicated and inflexi-
ble, which is not conducive to deploy Ambient IoT networks in a simple and rapid way.
In order to reduce network deployment costs, power consumption, and operating costs, a simplified
network architecture is required to be considered for the network architecture of Ambient IoT. The
simplified network architecture can simplify the types of network elements and combine network
functions, making the deployment of network elements meet the requirements of Ambient IoT. Fur-
thermore, the interface protocols between different network element types are also as simple as
possible.
The following sections describe the characteristics of a network architecture suitable for Ambient IoT.
5.5.2 Support of simplified controlling signaling or transport plane
First of all, Ambient IoT does not require personalized QoS requirements, and signaling interaction is
greatly reduced. For small data that needs to be sent, it can be sent in the mobility management pro-
cedure, so that it can reduce the interactive signaling for establishment of separate data channels.
Alternatively, if the default destination data center is configured at the Ambient IoT device, the Ambi-
ent IoT device can send in a stateless manner, and then send uplink data only when triggered by the
network. The network can establish a dedicated data channel for Ambient IoT devices or services,
which can avoid the establishment of a dedicated Ambient IoT data channel for each terminal.
In many cases, Ambient IoT only requires simple partial communication. In order to achieve such a
function, a simple Non-Access-Stratum layer processing functions can be deployed in the base sta-
tion, so that integrated communication with the air interface can be realized.
nt
mbie
n ne l for Avice
Tu er
IoT s
Network Date
Function network
Tunne
l for
IoT se Ambient
rvice
Network
Function
Figure 5.5-1 Simplified tunnel for Ambient IoT
5.5.3 Network architecture supporting hierarchical control
Ambient IoT devices can be used in logistics and warehouse. Due to limited power consumption of
Ambient IoT devices, a hierarchical network architecture can be adopted. For example, the Ambient
IoT device first sends the data to a certain data cache point in logistics or warehouse (data cache
points can be installed or deployed in logistics vehicles or inside the warehouse), and the data cache
points in logistics or warehouse are periodically or quantitatively reported with the data sent by the
Ambient IoT device to the network. For downlink data, the network can also first send the data to a
data cache point in logistics or warehousing for buffering, and then the data cache point can send
the downlink data to a group of Ambient IoT devices at a fixed time or via a paging triggering.
Secondary
Cache
Figure 5.5-2 Network architecture supporting hierarchical control
5.5.4 Support flexible and efficient network selection function
Ambient IoT is mainly used in industrial sensor networks, logistics and smart homes. Therefore, Am-
bient IoT can ignore complex network environments such as roaming scenarios, and the demand for
network selection is weakened. In this regard, the Ambient IoT device can perform flexible and effi-
cient network selection functions to minimize the power consumption.
5.6 References
[1]3GPP (2020). 3GPP TS 38.101: User Equipment (UE) Radio Transmission and Reception. 3rd
Generation Partnership Project; Technical Specification Group Radio Access Network.
[2]S. Shellhammer, A. Asterjadhi, and Y. Sun, IEEE 802.11ba: Ultra-Low Power Wake-up Radio
Standard. Wiley-IEEE Press, January 2023.
[3]Sourour E, El-Ghoroury H, Mcneill D. Frequency offset estimation and correction in the IEEE
802.11a WLAN[J]. IEEE, 2004.DOI:10.1109/VETECF.2004.1405033.
[4]Ni L M, Liu Y, Lau Y C, et al. LANDMARC: indoor location sensing using active RFID[J]. Wireless
Networks, 2003.DOI:10.1109/PERCOM.2003.1192765.
[5]P. V. Nikitin, R. Martinez, S. Ramamurthy, H. Leland, G. Spiess and K. V. S. Rao, "Phase based
spatial identification of UHF RFID tags," 2010 IEEE International Conference on RFID (IEEE RFID
2010), Orlando, FL, USA, 2010, pp. 102-109, doi: 10.1109/RFID.2010.5467253.
[6]3GPP TS 33.401: “System Architecture Evolution (SAE); Security architecture”.
[7]3GPP TS 33.501: "Security architecture and procedures for 5G system".

Standardization and future trends of Ambient IoT 06
STANDARDIZATION
AND FUTURE TRENDS
OF AMBIENT IOT
06
Industry status and standardization

6.1 progress of Ambient IoT
6.1.1 Industry status of Ambient IoT
As mentioned earlier, there are clear and extensive application scenarios and urgent market
demands for the Ambient IoT technology in the industry. However, there is currently no comprehen-
sive Ambient IoT standard that can address the requirements of multiple scenarios. Prior to this, both
the industry and academia have made many efforts to meet the application requirements of some
scenarios. Next, we will introduce the current standardization activities and industry status of Ambi-
ent IoT.
RFID RFID (Radio Frequency Identification) is a technology that uses radio frequency for contactless bidi-
rectional data communication. A typical RFID system includes electronic tags, readers, and applica-
tion systems. The electronic tags store relevant information about objects, while the reader is
responsible for reading from and writing to the tags, and the application system is used to process
and manage this information.
RFID, with the help of RF energy harvesting and ultra-low-power backscatter communication tech-
nology, has realized ultra-low complexity and ultra-low cost RFID tags. Currently, RFID technology is
widely used in logistics, retail, manufacturing, healthcare, transportation, and other fields. Its applica-
tion helps improve efficiency, reduce errors, and lower costs.
However, RFID also has some technical limitations, preventing its application in more scenarios. For
example, in the process of backscatter communication, the carrier signal transmitted by the RFID
reader and the signal reflected by the RFID tag are in the same channel, resulting in full-duplex inter-
ference between them. This limits the communication distance of the RFID system, generally not
exceeding 10 meters, restricting its usability in various work environments. Additionally, the RFID
protocol design greatly simplifies the complexity of RFID tag for system access, but it also limits the
capacity of the RFID system, with a limited number of RFID tags that can be simultaneously
accessed by the system. Finally, RFID systems are typical point-to-point communication systems
that lack flexible and effective networking capability, leading to limitation of working efficiency in real
life applications. For example, for the logistics and warehousing scenario, the efficiency and reliability
of inventory via RFID tags still need to be improved.
EnOcean EnOcean is another wireless communication system based on energy harvesting technology, partic-
ularly suitable for smart homes and building automation fields, such as wireless control for lighting
fixtures. EnOcean technology supports ultra-low-power module operation and can self-power by
collecting energy from the environment (such as mechanical energy, light energy, thermal energy,
etc.), thus operating without batteries. This energy harvesting capability makes maintenance simple
or even unnecessary, especially suitable for applications where power sources are difficult to
replace.
EnOcean's communication protocol design is streamlined, employing a communica-

tion mechanism without the need for handshaking, making the communication pro-
cess efficient and extremely low-power. In a typical EnOcean communication system,
IoT devices act merely as transmitters. After collecting environmental energy, the
devices can send sensor data and control signals to receivers. Receivers typically
include gateways powered by cables, switch devices, control devices, etc. EnOcean
data transmission uses a subtelegrams mechanism, where IoT devices repetitively
send data packets to ensure proper reception by receivers and to prevent interfer-
ence with other IoT devices.
EnOcean technology has obtained certification from the International Standards

Organization ISO/IEC, including specifications for the physical layer, data link layer,
and network layer.
Due to the design of the EnOcean protocol, which makes it more suitable for unidirec-
tional control scenarios, its range of application scenarios is somewhat limited.
Ambient IoT based on In addition to the aforementioned Ambient IoT standards for specific scenarios, there
existing standards are also explorations based on existing standards (such as Bluetooth, LoRa, WiFi, etc.)
such as Bluetooth, driven by urgent industry demands.
LoRa, etc. In these explorations, Ambient IoT devices communicate based on current protocols
(such as Bluetooth, LoRa, etc.), and in different application scenarios, these devices
can utilize energy sources like solar energy, thermal energy, etc., for energy harvesting
to obtain the required energy for communication. Although some successful practices
have been achieved in individual application scenarios, the design of these protocols
does not specifically optimize for ultra-low power communication targeting the
extremely low energy density of ambient energy. Therefore, in practical use, either
larger ambient energy harvesting modules need to be equipped for the devices to col-
lect sufficient ambient energy, or the communication frequency of Ambient IoT devic-
es needs to be restricted to reduce total energy consumption. Consequently, the
application scenarios for such Ambient IoT devices are also very limited.
6.1.2 Standardization progress of Ambient IoT technology

In order to meet the demands of the industry, the two major global standardization
organizations, 3GPP and IEEE, are currently conducting research and standardization
efforts for Ambient IoT technology.
Standardization progress Over the past few decades, 3GPP has standardized a series of wireless communica-
of Ambient IoT tion technologies, including GSM, WCDMA, LTE, NR, etc., greatly meeting people's
technology in 3GPP needs in areas such as voice calls, mobile internet, and effectively promoting the
development of society and the improvement of people's living standards. In the past
decade, 3GPP has also standardized technologies such as MTC, NB-IoT, and RedCap,
specifically targeting IoT requirements, and has achieved certain commercial success.
However, 3GPP has not yet standardized a solution that meets the requirements of
Ambient IoT-related scenarios.
In past standardization practices, 3GPP has accumulated a wealth of technologies and solutions
related to low-power communication, low-complexity communication, massive connectivity, etc.
The industry and academia have also made fruitful achievements in various environmental energy
harvesting technologies (such as RF energy harvesting, micro photovoltaic energy harvesting, tem-
perature difference-based thermal energy harvesting, etc.). Therefore, the timing for 3GPP to conduct
research and standardization of A,bient IoT technology has matured.
Starting from May 2021, OPPO took the lead in submitting a standardization proposal for Ambient
IoT to 3GPP SA1 [1-7], suggesting that 3GPP study the application scenarios and technical require-
ments of Ambient IoT. This proposal was approved at the SA1#97 meeting in February 2022 [8],
marking the beginning of 3GPP's research on Ambient IoT.
Below, we will introduce the standardization progress of 3GPP Ambient IoT:
• In February 2022, at the 3GPP SA1#97 meeting, the initiation of research on Ambient IoT pro-
posed by OPPO was approved. As of now, this project has completed all research and standard-
ization work. The project has studied and agreed upon over 30 application scenarios and their
technical requirements for Ambient IoT [9], and has developed relevant technical specifications [10].
• In September 2022, at the 3GPP RAN#97 meeting, agreement was reached for the study item
(SI) on Ambient IoT at the RAN level proposed by Huawei. This research project primarily focuses
on studying deployment scenarios and network topologies of Ambient IoT, types of devices,
design goals, etc. As of now, this SI has also completed all research work and produced a technical
report [11].
• In November 2023, at the SA#102 meeting, a SI on the network architecture of Ambient IoT, led
by Huawei and OPPO was agreed [12]. This SI will formally commence related research and stan-
dardization work from February 2024.
• In November 2023, at the RAN#102 meeting, a SI on RAN for Ambient IoT, led by China Mobile,
Huawei, and T-mobile [13] was approved. This research project will also formally commence related
research and standardization work from February 2024.
Support for cellular-based Ambient IoT offers at least the following advantages：
• Cellular communication networks have achieved wide-area seamless coverage in major regions
globally, thereby possessing the potential to provide wide-area coverage for Ambient IoT net-
works.
• Within cellular networks, there are not only base station equipment and relay nodes but also
numerous intelligent terminal devices. These devices and nodes can serve as power supply
nodes and communication nodes for Ambient IoT devices.
• Operators possess the high-quality sub-1 GHz frequency band, with minimal loss in the low-fre-
quency bands. This is advantageous not only for enhancing communication distances but also
for extending wireless power supply distances, thus enabling better coverage for Ambient IoT
networks.
• Cellular networks possess strong device management and mobility management capabilities,
supporting IoT communication in complex scenarios (such as logistics).
Standardization progress Over the past 20 years, IEEE 802 work group (WG) has standardized a series of
of Ambient IoT WLAN communication standards and achieved significant commercial success.
technology in IEEE Through continuous technological updates and standard iterations, WiFi stan-
dards have achieved characteristics such as high speed, large capacity, and low
latency, effectively meeting the demand for high-speed wireless local area com-
munication in scenarios such as homes, offices, and shopping malls. In past prac-
tices, WiFi standards such as 802.11b/n have also been widely applied in IoT sce-
narios such as industrial monitoring and industrial automation. In response to IoT
communication requirements, 802.11 has specifically developed 802.11ah to
achieve lower device costs, a greater number of connections, lower power con-
sumption, and longer battery life. However, IEEE 802 still lacks a standard for Am-
bient IoT communication.
Research and standardization of Ambient IoT technology within IEEE 802 offers
the following advantages：
• The widespread deployment of WLAN infrastructure provides a solid founda-

tion for the deployment of Ambient IoT networks.
• Free frequency bands can reduce the deployment and operation costs of Ambi-
ent IoT.
• IEEE 802 focuses on short-range and local communication scenarios such as

homes, smart wearables, and industrial local communication, better matching
the requirements of many scenarios in Ambient IoT technology.
• 802 WG has already standardized related technologies such as 802.11ah and

802.11ba, which can serve as the basis for standardizing Ambient IoT.
Based on the considerations mentioned above, in May 2022, OPPO submitted a

standardization proposal for Ambient IoT to IEEE 802.11 [14], proposing to study the
application scenarios, technical requirements, and key technologies of Ambient
IoT. This proposal was approved at the May 2022 IEEE meeting, initiating the
research process of IEEE 802.11 on Ambient IoT.
Below, we will introduce the standardization progress of IEEE 802.11 Ambient IoT：
• At the May 2022 IEEE meeting, approval was granted for the Technical Interest
Group (TIG) on Ambient IoT initiated by OPPO, establishing the AMbient Pow-
er-enabled (AMP) TIG. During the TIG phase, research on application scenarios,
technical requirements, key technologies, etc., for Ambient IoT was completed [15].
To verify the technical feasibility, the TIG phase also showcased multiple demos
provided by many companies [16].
• At the March 2023 IEEE meeting, approval was given for the establishment of a
Study Group for the IEEE 802.11 Ambient IoT proposed by companies such as
OPPO [17]. During the SG phase, further research was conducted on deploy-
ment methods, key technologies, and operating modes of Ambient IoT. At the
IEEE meeting in November 2023, approval was granted for the PAR [18] and
CSD texts [19] of the Task Group stage. In March 2024, the 802.11bp Task group
(TG) was formally established, signaling that the project will enter the Task
Group phase in the first half of 2024, officially commencing the development of
Ambient IoT standard specifications.
6.2 The future trends of Ambient IoT

6G technology is in the ascendant. At present, many kinds of candidate technologies for 6G have
been widely developed by the industry. As a new communication technology, Ambient IoT is expect-
ed to integrate with other 6G candidate technologies to build a green, energy-efficient, intelligent
and efficient mobile communication network.
6.2.1 Integration of Ambient IoT and symbiotic radio
In a typical AmBC (Ambient Backscatter Communications) system, Ambient IoT devices can use
radio waves in space to realize backscatter communication[20]. As shown in Figure 6.2-1, when a
router and an intelligent device in a primary communication system is communicating, Ambient IoT
device backscatters the downlink signal sent by the router in order to transmit information to the
reader. A secondary communication system supported by backscatter communication technology is
formed by the Ambient IoT device and the reader. Because of its potential value, backscatter com-
munication was rated as one of the top ten breakthrough technologies in the 2016 MIT Technology
Review.
Tag Reader
Legacy Receiver
Figure 6.2-1 Illustration of Ambient Backscatter Communications system
However, in the above AmBC system, because the same spectrum is used by the primary communi-
cation system and the secondary communication system, the communication of the secondary
communication system will interfere the primary communication system. That is, the backscatter
signal of the Ambient IoT device will be mixed into the signal received by the receiver of the primary
communication system and degrade its decoding performance. Therefore, although the use of
backscattering benefits the secondary communication system, it may sacrifice the performance of
the primary communication system.
In order to solve the above issue, the concept of symbiotic radio is proposed [21] Symbiotic radio
works on the basis of environmental backscattering through good coordination between the prima-
ry system and the secondary communication system. It not only eliminates the interference from the
secondary communication system to the primary system, but also converts the backscatter signal
into a signal which is beneficial to the primary communication system.
PTx PRx
STx SRx
Figure 6.2-2 Symbiotic radio
As shown in Figure 6.2-2, a symbiotic radio system includes the primary communication system and
the secondary communication system. The primary transmitter PTx and the primary receiver PRx
constitute the primary communication system while the secondary transmitter STx and the second-
ary receiver SRx constitute the secondary communication system. STx realizes backscattering using
signals transmitted by PTx. In order to enable symbiotic radio, one of the most important things is
that the chip width Cp of the signal backscattered from the secondary communication system and
the chip width Cs of the primary communication system need to satisfy a K-fold relationship, namely
Cp=K*Cs. Therefore, the backscattering signal does not change in the duration corresponding to K
chips of the primary communication system. Thus, when the primary communication system per-
forms coherent demodulation, the backscatter signal from the secondary communication system is
equivalent to a multipath signal for the primary communication system[22]. With such constraints, it
not only eliminates the interfere from the secondary communication system, but also improves the
performance of the primary communication system by providing multipath signals[23]. Since such
kind of relationship of between the primary communication system and the secondary communication
system is similar to biological symbiosis, the communication system model is named symbiotic radio.
S1(n) S2(n) ... Sk(n)
(a) Signal of the primary system
c(n)
(b) Signal of the secondary system
Figure 6.2-3 Relationship between the signals from the primary

communication system and the secondary communication system
Symbiotic radio enables Ambient IoT to share the spectrum of traditional communication and coex-
ist well with traditional communication on the same spectrum. Therefore, symbiotic radio is expect-
ed to become an important manner to realize Ambient IoT communication.
6.2.2 Integration of Ambient IoT and ISAC

（Integrated Sensing And Communication）
While transmitting information in wireless channels, ISAC can realize the sensing functions of target
positioning, detection, imaging and recognition by actively recognizing and analyzing the character-
istics of the channel and using wireless signals to sense the physical characteristics of the surround-
ing environment, so as to excavate communication capabilities and enhance user experience. In
terms of hardware architecture, a great challenge for the ISAC system is to adapt to high-precision
sensing requirements. Thus, it will greatly increase the dynamic range and complexity of hardware
system. How to design a green and energy-saving hardware system architecture that meets the
two-way requirements of ISAC is one of the important challenges in the future [24].
In ISAC, it is necessary to sense various users and environmental backgrounds. Usually, these tasks
are done with the help of special sensors. However, these special sensors often need external power
source to run. Because the energy stored by the battery is limited, the battery either needs to be
charged or replaced during the use of the equipment, which is not only inconvenient and costly, but
also impossible to implement in some deployments (extreme environments such as high tempera-
ture, low temperature, radiation, etc.).
The integration of Ambient IoT technology (mainly using energy harvesting and ultra-low power
communication) and ISAC can significantly improve the energy efficiency and meet the green
energy saving goal of ISAC [25]. On the one hand, energy harvesting technology can obtain energy
from the surrounding environment and fundamentally eliminate the dependence on batteries. On
the other hand, when communicating based on ultra-low power communication technology, only
microwatt power consumption is required to achieve extremely low power consumption. At present,
backscattering technology has been applied in many fields, such as using radio frequency backscat-
tering for food and liquid quality test [26], using battery-free mobile phone for communication [27],
using backscattering assisted vehicle network[21], using backscattering for underwater monitoring [28],
using visible backscattering for gesture sensing [29], etc.
Ambient IoT also provides an effective means for ISAC. An Ambient IoT unit is installed on the target
object, which can trigger the Ambient IoT communication process when the network triggers the
sensing signaling of the target object, so that the information of the target object is reported to the
network side through the Ambient IoT communication mode, realizing accurate sensing function.
In addition, the energy harvesting mode can also be used to reflect the environmental characteris-
tics. At present, there has been research on the sensing of the surrounding environment based on
energy harvesting signal. Therefore, the energy harvesting module in Ambient IoT can be used as a
virtual sensor to detect the surrounding environment. Based on this, the industry has extended the
application of signals used for energy harvesting [30], which can be used not only for circuit driving
but also for sensing through energy harvesting signals. For example, kinetic-powered wearable IoTs
are able to detect and count the user’ s steps, as the energy harvester generates distinguishable
peaks in the energy harvesting signal each time the leg hits the ground [31]. Similarly, a thermoelectric
energy harvester is able to detect any changes in surface temperature simply from the variations in
the generated energy harvesting signal [32][33].
Context Sensing
Sensors
Environmental energy power

Energy Harvester MCU
(kinetic,thermal, solar, RF energy)
Radio
Figure 6.2-4 Application Extension of Energy Harvesting Signal
In recent years, the energy harvesting modes from kinetic, thermoelectric, solar, and RF energy harvesters have been
demonstrated to detect a variety of contexts [30]. Irrespective of the type of energy source, there are two main approach-
es for sensing from energy harvesting signals. The first approach analyzes the patterns of the instantaneous power gen-
erated by the energy harvesting transducer, while the second approach analyzes the amount of the total energy accu-
mulated in the storage over a specific period of time.
Typical applications of sensing based on energy acquisition signals [30] are shown in Table 6.2-1:
Table 6.2-1 Typical applications of sensing based on energy acquisition signals
Energy Harvesting Patterns Typical applications
Human Activity Recognition (such as walking and running).
Transportation Mode Detection (such as car, bus, or train).
Energy Harvesting
Estimation of Calorie Expenditure.
from Kinetic energy
Step Counting.
HVAC (Heating Ventilation Air Conditioning) Airflow Monitoring, etc.
(harvests energy from the pipe’ s thermal gradient, i.e., the temperature difference
between the pipe and the room temperature, when hot water flows through the
Water Flow Detection [34] pipe. The harvested thermal energy is used to wake up the sensor from deep sleep
mode as well as to compensate battery energy expenditure.).
Energy Harvesting from
Thermoelectric energy
Heat Appliance Monitoring.
Chemical Reaction Detection, etc.
Energy Harvesting from Solar energy Localization and Positioning, Gesture Recognition, Visible Light Communication, etc.
Energy Harvesting from RF energy RFID tags for integrated sensors
The integration of Ambient IoT and ISAC can realize the integration of communication, sensing and energy transmission.
On the one hand, battery-free and energy saving can be achieved through energy harvesting and ultra-low power com-
munication technology, which is convenient for deployment in some complex environments. On the other hand, it can
also sense the surrounding environment based on the energy harvesting module.
6.2.3 Integration of Ambient IoT and AL/ML
AI (Artificial Intelligence) is the simulation of human intelligence processes by computer system. It is

usually defined as the science of making computers to perform tasks that require intelligence like
humans [35]. ML (Machine Learning) is one of the most popular applications of AI, which can optimize
system performance and reactions to environments by learning, inferring, fitting and categorizing
from large amount of data. Generally, ML can be divided into supervised learning, unsupervised
learning and reinforcement learning. Taking supervised learning as an example, where artificial
neural networks (ANN) is a typical algorithm, the weighting coefficients between neural network
nodes are trained based on set of prior data. When the training is converged, multi-layered neural
network is able to identify and infer new data. In general, more essential characteristics can be
learned by increasing the number of hidden layer nodes, which is beneficial to improve the accuracy
in classification and prediction. Deep learning is achieved by this way. It is foreseen that AI technolo-
gy can effectively improve system performance, reliability and adaptability. Currently, AI has been
widely used in various fields to create more excellent operation efficiency.
For Ambient IoT system, terminals may experience more complicated and stringent communication
conditions. In some scenarios such as logistics and product line detection, a large number of termi-
nals need to communicate with network using limited channel resources, which brings challenges to
coordinate the communication between terminals. In addition, it is not suitable to perform extra
channel measurement or reporting procedure due to limited power of Ambient IoT devices. Without
sufficient channel information, how to improve date transmission performance and perform access
control effectively is another challenge. Therefore, AI technology can be considered to improve the
performance of Ambient IoT system in these scenarios. Possible directions are listed as follows:
1：Optimizing AI training model takes environment information, feedback information from network (e.g., whether
communication the UL transmission is correctly received) and system performance indicators as inputs, and com-
strategy of Ambient munication strategy at Ambient IoT devices as output, in order to optimize the operation strategy.
IoT with AI/ML [36][37]
For example, based on the training results, Ambient IoT devices can adaptively switch between
operations such as energy collection, backscattering or active transmission with ultra- low power
consumption, channel estimation and equalization etc. according to different environments. When
communicating with network, the device can adjust time, data rate, power and other parameters
with the help of AI in combination with environmental interference and its own power level. Reliabili-
ty and robustness of the Ambient IoT system can be improved with the aid of AI.
AI service
Environments
Strategy
Rewards
Carrier wave Backscattered signal Tag Server
Figure 6.2-5 Optimizing communication strategy of Ambient IoT device with AI/ML
2：Optimizing resource As an example, with the assistance of AI, Ambient IoT network nodes train
allocation method, access resource allocation methods and wireless power supplying strategies under
control and energy supplying transmission conditions. The coverage can be enhanced by selecting appropriate
strategy of network with
node pairing strategy to ensure efficient energy collection and communication.
AI/ML [38][39]
In addition, the performance of AI systems largely depends on sufficient data
sources. Ambient IoT devices have the advantages of small size, low cost, low
power consumption and are convenient for large-scale deployment. Therefore,
a low-cost data collection scheme can be provided by Ambient IoT devices,
which can be used to improve the performance of AI system. For example, in
smart factory scenario, Ambient IoT devices are used to collect environmental
information, such as temperature, humidity, pressure, motion attitude, vibration
frequency, etc. With this kind of data, AI can predict the changes in advance,
such as environment, working state, etc., in order to trigger warning or provide
indication information for other intelligent devices. In the smart home scenario,
Ambient IoT devices collect information such as the position of human body,
daily living habits, temperature, light, etc. AI can be used to realize intelligent
linkage between various home devices to make life more comfortable.
Environmental
Smart monitoring Smart
community energy
Smart Smart
medical agriculture
Smart
factory
In-vehicle
intelligence
Smart
logistics
Payment
Smart
transportation
Wearable Smart
device home
Figure 6.2-6 Low-cost data collection for

AI by using Ambient IoT devices
6.3 References
[1]3GPP, S1-211121, Motivation of support wireless power sourcing enabled communication
services in 5GS, OPPO.
[2]3GPP, S1-213047, Motivation of support wireless power sourcing enabled communication

[3]3GPP, S1-213046, New SID: Study on Wireless Power Sourcing enabled Communication
Services in 5GS, OPPO.
[4]3GPP, S1-214035, Motivation on support Energy Harvesting enabled Communication

[6]3GPP, S1-220119, Motivation on support Energy Harvesting enabled Communication

[8]3GPP, S1-220192, New SID: Study on Ambient power-enabled Internet of Things, OPPO.
[9]3GPP, TR 22.840 V19.0.0, Study on Ambient power-enabled Internet of Things.
[10]3GPP, TS 22.369 V19.0.0, Service requirements for ambient power-enabled IoT.
[11]3GPP, TR 38.848 V18.0.0, Study on Ambient IoT (Internet of Things) in RAN.
[12]3GPP, SP-231803, New SID: Study on Architecture support of Ambient power-enabled

Internet of Things, OPPO.
[13]3GPP, RP-234058, New SID: Study on solutions for Ambient IoT (Internet of Things) in
NR, Huawei.
[14]IEEE 802.11-22/0645r0, Ambient Power Enabled IoT for Wi-Fi.
[15]IEEE 802.11-23/0436r0, Technical Report on support of AMP IoT devices in WLAN.
[16]IEEE 802.11-22/1960r6, Summary and recommendation for AMP IoT.
[17]IEEE 802.11-23/0063r3, Proposal for consensus straw poll.
[18]IEEE 802.11-23/1006r4, IEEE 802.11 AMP PAR Proposal.
[19]IEEE 802.11-23/1212r2, IEEE 802.11 AMP SG Proposed CSD.
[20]V. Liu, A. Parks, V. Talla, S. Gollakota, D. Wetherall, and J. R. Smith, “Ambient backscatter:
wireless communication out of thin air,” in Proc. of ACM SIGCOMM, Hong Kong, China, Jun.
2013, pp. 1–13.
[21]Y.-C. Liang, Q. Zhang, Erik G. Larsson, and G. Y. Li, Symbiotic Radio: Cognitive Backscattering
Communications for Future Wireless Networks, IEEE Trans Cognitive Communications and Net-
working (TCCN), 2020.
[22]W. Liu, Y.-C. Liang, Y. Li, and B. Vucetic, Backscatter Multiplicative Multiple-Access Systems:
Fundamental Limits and Practical Design, IEEE Transactions on Wireless Communications, Sept
2018.
[23]R. Long, Y.-C. Liang, H. Guo, G. Yang, and R. Zhang, “Symbiotic radio: A new communication
paradigm for passive internet-of-things,” IEEE Internet Things J., 2020.
[24]https://mp.weixin.qq.com/s/I2KhygV3Rnzpm4gl7Zb_xg
[25]Toro U S, Wu K, Leung V C M. Backscatter Wireless Communications and Sensing in Green

Internet of Things[J]. IEEE Transactions on Green Communications and Networking, 2021.
[26]U. Ha, J. Leng, A. Khaddaj, and F. Adib, “Food and liquid sensing in practical environments using
rfids,” in 17th {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 20),
2020, pp. 1083–1100.
[27]V. Talla, B. Kellogg, S. Gollakota, and J. R. Smith, “Battery-free cellphone,” Proceedings of the
ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 1, no. 2, pp. 1–20, 2017. [5] P.
Wang, L. Feng, G. Chen, C. Xu, Y. Wu, K. Xu, G. Shen, K. Du, G. Huang, and X. Liu, “Renovating road
signs for infrastructure-tovehicle networking: A visible light backscatter communication and net-
working approach,” in Proceedings of the 26th Annual International Conference on Mobile Comput-
ing and Networking, 2020, pp. 1–13.
[28]R. Ghaffarivardavagh, S. S. Afzal, O. Rodriguez, and F. Adib, “Ultrawideband underwater back-

scatter via piezoelectric metamaterials,” in Proceedings of the Annual conference of the ACM
Special Interest Group on Data Communication on the applications, technologies, architectures, and
protocols for computer communication, 2020, pp. 722–734.
[29]A. Varshney, A. Soleiman, L. Mottola, and T. Voigt, “Battery-free visible light sensing,” in Proceed-
ings of the 4th ACM Workshop on Visible Light Communication Systems, 2017, pp. 3–8.
[30]D. Ma, G. Lan, M. Hassan, W. Hu, and S. K. Das, “Sensing, computing, and communications for
energy harvesting iots: A survey,” IEEE Communications Surveys & Tutorials, vol. 22, no. 2, pp.
1222–1250, 2019.
[31]S. Khalifa, M. Hassan, and A. Seneviratne, “Step detection from power generation pattern in
energy-harvesting wearable devices,” in Proc. IEEE Int. Conf. Data Sci. Data Intensive Syst., 2015, pp.
604–610.
[32]P. Martin, Z. Charbiwala, and M. Srivastava, “DoubleDip: Leveraging thermoelectric harvesting

for low power monitoring of sporadic water use,” in Proc. 10th ACM Conf. Embedded Netw. Sensor
Syst., 2012, pp. 225–238.
[33]B. Campbell, B. Ghena, and P. Dutta, “Energy-harvesting thermoelectric sensing for unobtrusive
water and appliance metering,” in Proc. 2nd Int. Workshop Energy Neutral Sens. Syst., 2014, pp. 7–12.
[34]P. Martin, Z. Charbiwala, and M. Srivastava, “DoubleDip: Leveraging thermoelectric harvesting

for low power monitoring of sporadic water use,” in Proc. 10th ACM Conf. Embedded Netw. Sensor
Syst., 2012, pp. 225–238.
[35]C. Wang, M. D. Renzo, S. Stanczak, S. Wang and E. G. Larsson, "Artificial Intelligence Enabled
Wireless Networking for 5G and Beyond: Recent Advances and Future Challenges," in IEEE Wireless
Communications, vol. 27, no. 1, pp. 16-23, February 2020, doi: 10.1109/MWC.001.1900292.
[36]N. Van Huynh, D. N. Nguyen, D. T. Hoang and E. Dutkiewicz, "“Jam Me If You Can:” Defeating
Jammer With Deep Dueling Neural Network Architecture and Ambient Backscattering Augmented
Communications," in IEEE Journal on Selected Areas in Communications, vol. 37, no. 11, pp.
2603-2620, Nov. 2019, doi: 10.1109/JSAC.2019.2933889.
[37]F. Jameel, W. U. Khan, S. T. Shah and T. Ristaniemi, "Towards Intelligent IoT Networks: Reinforce-
ment Learning for Reliable Backscatter Communications," 2019 IEEE Globecom Workshops (GC
Wkshps), 2019, pp. 1-6, doi: 10.1109/GCWkshps45667.2019.9024401.
[38]Y. Nie, J. Zhao, J. Liu, J. Jiang and R. Ding, "Energy-efficient UAV trajectory design for backscatter
communication: A deep reinforcement learning approach," in China Communications, vol. 17, no. 10,
pp. 129-141, Oct. 2020, doi: 10.23919/JCC.2020.10.009.
[39]T. T. Anh, N. C. Luong, D. Niyato, Y. Liang and D. I. Kim, "Deep Reinforcement Learning for Time
Scheduling in RF-Powered Backscatter Cognitive Radio Networks," 2019 IEEE Wireless Communi-
cations and Networking Conference (WCNC), 2019, pp. 1-7, doi: 10.1109/WCNC.2019.8885426.
Conclusion 07
EPILOGUE
Time flies, and it has been over two years since the first edition of the "Zero Power Communi-
cation" white paper was released. Over the past two years, we have worked together with vari-
ous industry partners to promote the research, standardization, and industrialization of such
technology, and rename it as Ambient IoT technology based on its unique technical features.
We have delved into the frontline of the industry to understand the real demands from all par-
ties. Through our joint efforts with various parties, we are pleased to see that many industry
players, including numerous operators, potential industry application parties, equipment ven-
dors, and chip manufacturers, have shown strong interest in Ambient IoT technology. The
industry is eagerly anticipating the rapid implementation and application of this technology. In
response to industry demand, we have been working with all parties to promote the technical
research and standardization of Ambient IoT technology in 3GPP and IEEE 802.11, and have
made significant progress. We also showcased Ambient IoT tags at MWC 2023 and received
widespread attention. This technology was also honored to be selected as one of the best
inventions of 2023 by Time Magazine.
In the past two years of research and standardization practice, we have closely communicated
with various industry partners about Ambient IoT technology. During this period, we have
repeatedly re-examined the industry positioning and potential market value of Ambient IoT
technology. We conducted multiple rounds of research on Ambient IoT application scenarios
and worked with the industry to identify the most likely typical application scenarios for imple-
mentation. Through communication with all parties, we have further improved the technical
framework and technical roadmap of Ambient IoT and deepened our understanding of its key
technologies. In standardization discussions, for the convenience of technical exchange, indus-
try parties have reached a consensus on the standardization naming of this technology—Ambi-
ent Power-enabled Internet of Things. Based on this, we realize the need to update the "Zero
Power Communication" white paper to reflect these latest changes and developments.
We are pleased to see that the first edition of the "Zero Power Communication" white paper
has become a bridge for us to meet many friends from the industry. Friends from various indus-
tries, universities, and research institutes from all over the world have come together with us
after reading this white paper to jointly promote the standardization and industrialization of Am-
bient IoT technology. In this process, many friends and teachers have given us strong support
and selfless assistance, for which we are grateful. We hope that the updated version of this
white paper, titled "Ambient Power-enabled Internet of Things," can contribute to the develop-
ment of our cause. We believe that through everyone's concerted efforts, the commercializa-
tion of the Ambient IoT is imminent, and it will play a positive role in improving social production
efficiency and people's living standards.

3 Ambient-Power-enabled-IoT-White-Paper

Uploaded by

Copyright:

Available Formats

3 Ambient-Power-enabled-IoT-White-Paper

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3 Ambient-Power-enabled-IoT-White-Paper

Uploaded by

Copyright:

Available Formats

Ambient Power-enabled IoT

Harsh communication environment

Scenarios where ultra-small devices are requested

Scenarios asking for ultra-low cost IoT devices

Position and vision

Logistics, Retail, Manufacturing, Supply chain, Remote metering, Sensors,

Lower capabilities Higher capabilities

Figure 1.3-1 Technical position of Ambient IoT

NB-IoT MTC RedCap Ambient IoT

Figure 1.3-2 Comparison of IoT technologies

MTC / NB-IoT RedCap Ambient loT

2010 2015 2020 2025

Figure 1.3- 3 Roadmap of IoT technology development

Intelligent manufacturing and

Typical scenarios in intelligent manufacturing and industrial monitoring：

Typical requirements of intelligent manufacturing and industrial monitoring

Power consumption：lower than 1mW, battery-free, maintenance-free.

Requirements for network： Deployment：Flexible deployment based on network infrastructure. Network

2.2 Logistics and warehousing

Figure 2.2-2 Ambient IoT in code chain transportation

Typical requirements of smart logistics and smart warehousing

Power consumption：battery-free, maintenance-free.

Requirements Deployment：Flexible deployment based on network infrastructure.

2.3 Smart home

Figure 2.3-1 Ambient IoT in smart home

Typical scenarios in smart home：

Typical requirements of smart home：

Power consumption：battery-free, maintenance-free.

Requirements Deployment：Smart terminal can be used as gateway or Ambient IoT devices

2.4 Indoor positioning

Fig. 2.4-1 Application of Ambient IoT technology in indoor positioning

Fig. 2.4-2 Indoor positioning in smart manufacturing and warehousing

Typical requirements of Indoor positioning

Tag power consumption：Battery free, no battery replacement, charging

Network requirements： Coverage requirements ：10m to 30m.

2.5 Smart grid

Power Primary Distribution Transformation

Figure 2.5-1 Smart grid overview

Terminal Tag power consumption：Battery free, no battery replacement,

2.6 Smart agriculture

Figure 2.6-1 A picture for huge greenhouse

Terminal Tag power consumption：Battery free, no battery replacement, charging

Network Flexible deployment ：Use intelligent terminals as gateway devices or directly

[6]3GPP TR 22.840 V19.0.0. Study on Ambient power-enabled Internetof Things .

Ambient energy harvesting

3.1.1 Ambient energy harvesting

There are several challenges regarding RF power harvesting.

Table 3.1-1 Input power vs. RF energy conversion efficiency [1]

Efficiency（%） Input power（dBm） Center frequency（MHz） Reflector unit

1.2 -14 950 0.3-μm CMOS convertor

5.1 -14.1 920 0.18-μm CMOS convertor

10 -22.6 906 0.25-μm CMOS convertor

11 -14 915 90-μm CMOS convertor

12.8 -19.5 900 0.18-μm CMOS , CoSi2 - Si Schottky

13 -14.7 900 0.35-μm CMOS convertor

16.4 -9 963 0.35-μm CMOS convertor

18 -19 869 0.5-μm CMOS convertor

26.5 -11.1 900 0.18-μm CMOS convertor

36.6 -6 963 0.35-μm CMOS convertor

47 -8 915 0.18-μm CMOS convertor