PerIoT'18 - Second International Workshop on Mobile and Pervasive Internet of Things

Overview of Cellular LPWAN Technologies for IoT Deployment:

Sigfox, LoRaWAN, and NB-IoT
Kais Mekki1, Eddy Bajic1, Frederic Chaxel1, Fernand Meyer2
1
Research Centre for Automatic Control of Nancy, Campus Sciences, BP 70239, Vandoeuvre, 54506, France.
2
OKKO SAS, 34 Rue Nationale, Puttelange-aux-Lacs, 57510, France.
{kais.mekki, eddy.bajic, frederic.chaxel}@univ-lorraine.fr

Abstract— LPWAN are actually the most popular low to the partnership with various network operators [12].
cost, long battery lifetime, and long range communication LoRaWAN was firstly developed by the start-up Cycleo
technology for IoT applications. This paper presents a in 2009 (in Grenoble, France) and purchased 3 years
comprehensive and comparative study on three actually later by Semtech (USA). In 2015, LoRaWAN was
leading LPWAN technologies called Sigfox, LoRaWAN,
standardized by LoRa-Alliance and is actually deployed
and NB-IoT. We show that Sigfox and LoRaWAN excel on
network capacity, devices lifetime, and cost. Whereas, NB- in 42 countries and still under roll-out in others countries
IoT excels on quality of service and latency. In addition, we due to the investment of various mobile operators (e.g.
consider application scenarios and explain which Bouygues and Orange in France, KPN in Netherlands,
technology fits best to guide future researchers and Fastnet in South Africa) [13].
industrials. NB-IoT is a LPWAN technology based on narrowband
Keywords—Internet of Things, Low Power Wide Area radio technology. NB-IoT is standardized by the 3rd
Networks, LoRaWAN, Sigfox, NB-IoT. Generation Partnership Project (3GPP). Its specification
was published at Release 13 of 3GPP on June 2016.
I. INTRODUCTION Actually, NB-IoT is under test in Europe. In December
Nowadays, Internet of Things (IoT) technologies are 2016, Vodafone and Huawei integrated NB-IoT into the
used in various application fields including security, Spanish Vodafone network and send the first message
asset tracking, agriculture, smart metering, smart city, conforming to the NB-IoT standard to a device installed
and smart home [1]. IoT applications have specific in a water meter. Currently, Huawei is multiplying
requirements such as long communication range, very partnerships to deploy this technology throughout the
low energy consumption, and cost effectiveness. The world (it was announced to be deployed in many
widely used short communication range technologies countries on 2018). In May 2017, the Ministry of
(e.g. BLE and ZigBee) are not adapted for long Industry and Information Technology in China
transmission range requirement. Further, solutions based announced its decision to accelerate the commercial use
on mobile cellular communications (e.g. 2G, 3G, and 4G) of NB-IoT for utilities and smart city applications.
could ensure larger transmission range, however, it
deplete the device energy. Therefore, IoT applications
requirement leads to the emergence of Low Power Wide
Area Network (LPWAN). LPWAN technologies ensure
long transmission range, low energy consumption, and
low cost deployment solution. It allows up to 40 km as
communication range in rural zones and 10 km in urban
zones [2], up to 10 years of battery lifetime [3], less than
5$ of device cost, and less than 1$ per device per year of
operator subscription cost [4]. It was particularly
Fig. 1. Data rate vs. range: LPWAN positioning
designed for IoT applications that require transmitting
few tiny messages per day in long radio range as shown In this paper, the technical differences of Sigfox,
in Figure 1. These advantages have shoved various LoRaWAN, and NB-IoT are presented and compared in
performance studies of LPWAN in outdoor and indoor terms of physical/communication features. In addition,
environment [5-7]. As recently as early 2013, the term these technologies are compared in terms of IoT success
"LPWAN" did not even exist [8]. factors such as quality of service (QoS), coverage, range,
Many of the LPWAN technologies have arisen in both latency, battery life, scalability, payload length,
licensed and unlicensed frequency spectrum. Among deployment, and cost. Further, we consider application
them, Sigfox, LoRaWAN, and NB-IoT are today the scenarios and explain which technology fits best.
leading emergent technologies [8-11] which involve The rest of paper is organized as follows. Section 2
many technical differences. describes the network topology in Sigfox, LoRaWAN,
Sigfox technology was developed in 2010 by the start-up and NB-IoT. Section 3 details the technical features of
Sigfox (in Toulouse, France), from where Sigfox is both these technologies. Section 4 compares them in terms of
a company and a LPWAN network operator. Actually, IoT factors. Section 5 explains which technology fits best
Sigfox operates and commercializes its own IoT solution for different application scenarios. Finally, section 6
in 31 countries and still under roll-out over the world due discusses and concludes the paper.

978-1-5386-3227-7/18/$31.00 ©2018 IEEE 197

 PerIoT'18 - Second International Workshop on Mobile and Pervasive Internet of Things

II. NETWORK TOPOLOGY: SIGFOX, LORAWAN, NB-IOT noise levels, leading to very low power consumption,
high receiver sensitivity, and low cost antenna design.
Mesh topology has widely employed for expanding the
Sigfox initially upheld only uplink transmission, yet later
coverage of short range networks (e.g. BLE and ZigBee).
it was advanced to bidirectional communication. The
However, the major disadvantage is their high
downlink transmission occurs only following an uplink
deployment cost to connect huge number of devices
transmission. The number of messages over the uplink is
which are geographically dispersed in a wide area range.
limited to 140 messages per day. The maximum payload
Moreover, as data is transmitted through multi hops
length for each uplink message is 12 bytes. However, the
towards a gateway, a subset of devices are more
number of messages over the downlink is limited to 4
congested than others which reduce their batteries
messages per day, thus all uplink message could not be
lifetime (i.e. excessive energy consumption) and thus
acknowledged. The maximum payload length for each
limit the entire network lifetime [14][15]. This limitation
downlink message is 8 bytes.
is surmounted in Sigfox, LoRaWAN, and NB-IoT
Without acknowledgment capability of every uplink
technologies: the end-devices are directly connected to
message, time/frequency diversity and transmission
public base stations (i.e. star topology) as shown in
duplication are used for ensuring the uplink transmission
Figure 2. Unlike mesh topology, connected end-devices
reliability. Each end-device sends message three times
do not listen to radio channel before transmitting
through different frequency channels. Base stations could
message which ensures energy saving advantages. As
receive messages simultaneously over all channels, the
base stations are always-on, they ensure immediate
end-device can thus transmit the message in randomly
access to connected end-devices.
chosen channel which reduces the end-device complexity
Figure 2 shows two communication way: uplink and
and reduces its cost.
downlink. In uplink transmission, the end-device sends
messages towards base stations. Upon receiving a B. LoRaWAN
message, the base stations forward it to backend server
Since 2015, a LoRa based communication protocol
(i.e. network server) using IP-based network. The
called LoRaWAN was standardized by LoRa-Alliance.
network server checks authentication/authorization of
LoRa is a patented spread spectrum technology using the
messages and forward it to the application server.
unlicensed sub-GHZ band [16]. The LoRa chirp spread
Despite that LPWAN are firstly designed for only uplink
spectrums (CSS) modulations ensure full bidirectional
transmission, downlink transmission is also available in
communication, and the generated signal has low noise
Sigfox, LoRaWAN, and NB-IoT. If downlink
levels, enables high interference resilience, and is
transmission is required, the application server transmits
difficult to detect or jam [17].
a message towards the network server. The network
LoRaWAN provides six spreading factors (SF7 to SF12)
server forwards it to the suitable base station which sends
to adapt the data rate and range tradeoff. The higher
it then to the end-device.
spreading factor provides the longer transmission range
and lowest data rate [18]. The data rate is between 300
bps and 50 kbps and the maximum payload length for
each message is 243 bytes.
Using LoRaWAN, each message transmitted by an end-
device is received by all base stations in the range. By
exploiting this redundant reception, LoRaWAN improves
the communication reliability ratio. Nevertheless, this
feature requires many base stations deployment in each
Fig. 2. Network topology of IoT deployment using Sigfox,
region which would grow the network infrastructure cost.
LoRaWAN, or NB-IoT The backend system filters the redundant reception of
each message by verifying security and transmitting back
III. TECHNICAL DIFFERENCES acknowledgment to the end-device, and forwarding the
message to corresponding application servers. In addition,
In this section, we highlight emerging technologies and reception of each message by multiple base stations
the technical aspects of Sigfox, LoRaWAN, and NB-IoT avoid the handover in LoRaWAN network (i.e. if an end-
that are finally summarized in Table 1. device is mobile there is no handover needed between
A. Sigfox base stations). Further, this feature is exploited by
LoRaWAN for localizing end-devices using TDOA-
Sigfox uses its patented UNB technologies and deploys based technique.
its proprietary base stations deployed in various countries LoRaWAN defines multiple communication classes for
in the unlicensed sub-GHz ISM bands (e.g. 868MHz in addressing the different latency in IoT applications:
Europe, 915MHz in North America, and 433MHz in - Class A (bidirectional end-devices): end-devices allow
Asia). The end-devices connect to these base stations bidirectional communication where each end-device's
using BPSK modulation in an ultra narrow band of 100 uplink transmission is followed by two short windows
Hz at a maximum data rate of 100 bps. By employing for receiving downlink messages as shown in Figure 3.
ultra narrow band in sub-GHz spectrum, Sigfox The uplink transmission time is scheduled by the end-
efficiently uses the frequency band and has very low

198
 PerIoT'18 - Second International Workshop on Mobile and Pervasive Internet of Things

device based on its own communication needs. This class NB-IoT communication protocol is based on LTE
is the lowest power end-device system for IoT protocol (i.e. NB-IoT reuses various principles and
applications that only require short downlink building blocks of LTE physical and higher protocol
communication after the end-device has sent an uplink layers). In fact, NB-IoT reduces LTE protocol
message. Downlink transmission at any other time will functionalities to the minimum and enhances them as
have to wait until the next uplink message of the end- needed for IoT applications. As example, LTE backend
device. system is used to broadcast information which is valid
for all end-devices within a cell. As backend system
broadcasting takes resources and causes battery
consumption for each end-device, it is kept to a
minimum as well in its size as in its occurrence. It was
Fig. 3. Bidirectional communication between end-device and base
optimized to small and infrequent data messages and
station for LoRaWAN class A abstain the features not required for IoT purpose (e.g.
radio quality measurements, aggregation, etc). This way,
- Class B (bidirectional end-devices with scheduled the end-devices can be kept in an efficient cost and needs
receive slots): in addition to the random receive windows only a small amount of battery power.
of class A, the class B devices open extra receive NB-IoT allows connectivity of more than 100K devices
windows at scheduled times. In order to open receive per cell and it could be increased by exploiting multiple
window at the scheduled time, end-devices receive a NB-IoT carriers. It employs QPSK modulation,
time synchronized Beacon from the base station. This Frequency Division Multiple Access (FDMA) in uplink,
allows the server to know when the end-device is and Orthogonal FDMA (OFDMA) in downlink [19]. The
listening. maximum throughput rate is 200 kbps and 20 kbps in
- Class C (bidirectional end-devices with maximal downlink and uplink respectively, with 1600 bytes of
receive slots): end-devices have nearly continuously payload size in each message. As discussed in [20], NB-
open receive windows. This class consumes excessive IoT and LTE-M technologies can achieve 10 years of
energy and is defined for IoT applications with battery lifetime by transmitting on average 200 bytes per
continuous energy power resources. day.
The next version of LoRaWAN is actually under Actually, the improvement of NB-IoT continues with
specification by LoRa-Alliance [13]. The new expected Release 15 of 3GPP. According to the 3GPP current plan,
features are: class B clarification and the temporary NB-IoT will be extended to include multicast services
switching between class A and class C. (e.g. end-device software update and messages
concerning a whole group of end-devices), mobility, as
C. NB-IoT well as further technical details to enhance the
NB-IoT is a narrowband LPWAN technology which can applications field of NB-IoT technology.
coexist in LTE or GSM under licensed frequency bands. IV. COMPARISON REGARDING IOT FACTORS
NB-IoT occupies a frequency bandwidth of 200 KHz,
which corresponds to one resource block in GSM and To choose an appropriate LPWAN technology for an IoT
LTE transmission [19]. With this frequency bands application, various factors have to be considered: range,
selection, the following operation modes are possible as coverage, device lifetime, latency, scalability, payload
shown in Figure 4: length, deployment, quality of service, and cost. Sigfox,
- Stand alone operation: a possible scenario is the NB-IoT, and LoRaWAN are compared in the following
utilization of currently used GSM frequencies bands. regarding these factors.
- Guard band operation: using the unused resource A. QoS
blocks in LTE carrier's guard-band.
- In-band operation: using resource blocks in LTE carrier. Sigfox and LoRaWAN employ license-free sub-GHz
bands and asynchronous communication based on
For the stand alone operation, the GSM carriers in the ALOHA protocol. They can efficiently bounce
right part of Figure 4 are shown as an example in order interference and fading/multi-path, however, they do not
to indicate that this is a possible NB-IoT deployment. In provide quality of service. NB-IoT employs licensed
fact, 3GPP recommends the integration of NB-IoT in spectrum and LTE-based synchronous protocol which are
conjunction with the LTE. optimal for QoS at the detriment of cost (i.e. LTE
frequency band auctions are over 500 million euro per
MHz [8]). Due to QoS and cost tradeoff, NB-IoT is
preferred for applications that need guaranteed quality of
service, while the applications that do not have this
constraint should choose LoRaWAN or Sigfox.
Fig. 4. Operation modes for NB-IoT

199
 PerIoT'18 - Second International Workshop on Mobile and Pervasive Internet of Things

Table 1. Overview of LPWAN technologies: Sigfox, LoRaWAN, and NB-IoT

B. Battery life & Latency D. Coverage & Range

In Sigfox, LoRaWAN, and NB-IoT, end-devices are in The coverage of an entire city using only one base
sleep mode the most of time as long as the application station is the most Sigfox benefit (i.e. range >40 km). As
needs which reduce as much as possible the amount of example, Belgium (i.e. total surface ≈ 45000 km2) is
consumed energy (i.e. long end-devices lifetime). entirely covered by only seven Sigfox base stations [8].
However, the NB-IoT end-device consumes additional In contrary, LoRaWAN has lower range (i.e. range <20
energy due to synchronous communication and QoS Km) that allow covering an entire city like Barcelona
handling, and its OFDM/FDMA access modes require with three base stations.
more peak current [9]. This additional energy NB-IoT has the lowest range and coverage capability (i.e.
consumption reduces the NB-IoT end-device lifetime range <10 Km). It provides mainly services for end-
compared to Sigfox and LoRaWAN. devices which are out of reach of cellular mobile
Further, NB-IoT provides low IoT latency connectivity. networks (e.g. indoor, deep indoor). In addition, NB-IoT
Unlike Sigfox, LoRaWAN provides class C to handle is deployed only within LTE infrastructure, thus it is not
also low bidirectional latency at the expense of increased adapted for regions without LTE coverage (e.g. rural
energy consumption. Thus, Sigfox and LoRaWAN- zones).
Class-A are the best solution for latency insensitive
E. Deployment model
applications. For IoT applications with low latency
connectivity, NB-IoT and LoRaWAN-Class-C are the The NB-IoT standardization was published on 2016, and
better choice. is actually under rollout to establish its network over the
world. In contrary, LoRaWAN and Sigfox ecosystems
C. Scalability & Payload length
are mature and under commercialization in various
Supporting thousands of end-devices is one of the key countries and cities. LoRaWAN has the advantage that it
features for Sigfox, LoRaWAN and NB-IoT. These is currently deployed in 42 countries versus 31 countries
cellular LPWAN networks provide high scalability for Sigfox [12][13]. Although, LoRaWAN and Sigfox
feature. However, NB-IoT offers the advantage of very worldwide deployments are still in the rollout phase.
high scalability than Sigfox and LoRaWAN. NB-IoT Unlike Sigfox and NB-IoT, LoRaWAN offers local
allows connectivity of more than 100K devices per base network deployment (i.e. LAN using LoRa gateway) as
station compared to 50K per cell for Sigfox and well as public network operation via base stations. In
LoRaWAN [18]. industrial field, a hybrid operating model could be used
On the other hand, NB-IoT offers also the advantage of which deploys a local LoRaWAN network in factory
maximum payload length. NB-IoT allows transmission areas and uses the public base stations to cover the
of data up to 1600 bytes. LoRaWAN allows sending in outside areas.
maximum 243 bytes. In contrary, Sigfox proposes the
F. Cost
lowest payload length of 12 bytes which limit its
utilization on various IoT applications that need sending Various cost factors should be considered including
large data size. spectrum (license), deployment, and end-device costs.
Table 2 presents Sigfox, NB-IoT, and LoRaWAN costs.

200
 PerIoT'18 - Second International Workshop on Mobile and Pervasive Internet of Things

It shows that Sigfox and LoRaWAN have an advantage alkalinity sensors could significantly reduce the water
in relation to cost compared to NB-IoT. consumption and improve the yield as discussed in [22].
Devices update sensed data few times per hour as the
Table 2. Costs of Sigfox, LoRaWAN, and NB-IoT environment conditions have not radically changed. Thus,
Sigfox and LoRaWAN are ideal for this application.
Spectrum End-device
cost
Deployment cost
cost Moreover, various farms do not have today LTE cellular
coverage, thus NB-IoT cloud not be an agriculture
Sigfox Free >4000€/base station <2€ solution in the future.
>100€/gateway
LoRaWAN Free
>1000€/base station
3-5€ C. Manufacturing automation

>500 M€ Real time machinery monitoring prevents industrial

NB-IoT >15000€/base station >20€ production line down and allows remote control for
/MHz
efficiency improving. In factory automation, there are
In summary, each LPWAN technology has advantages in various types of sensors and communication
terms of different IoT factors as shown in Figure 5. requirements. Some applications require frequent
communication and high quality of service, thus NB-IoT
is a better solution than Sigfox and LoRaWAN. Other
applications require low cost sensors and long battery
lifetime for asset tracking and status monitoring, in this
case Sigfox and LoRaWAN are a better solution. Due to
this requirements variety, hybrid solutions could also be
used.
D. Smart building
Temperature, humidity, security, water flow, and electric
plugs sensors alert property managers to prevent damage
Fig. 5. Respective advantages of Sigfox, NB-IoT, and LoRaWAN and instantly responding to requests without having
manually building monitor. The buildings cleaning and
V. APPLICATION EXAMPLES: WHICH TECHNOLOGY FITS usage could also be done more efficiently. These sensors
BEST? require low cost and long battery lifetime as shown in
[23]. They do not require quality of service or frequent
IoT factors and technical differences of Sigfox, communication, so Sigfox and LoRaWAN are a better fit
LoRaWAN, and NB-IoT will impose which technology for this class of applications.
is suitable for each specific application. As discussed
before in this paper, one technology cannot serve equally E. Retail point of sale terminals
all IoT applications. In this section, various application
Sale point systems require guaranteed quality of service
use cases are discussed with a summary of which
since they handle frequent communication [24]. These
technology fits the best.
systems have continuous electrical power source, thus
A. Electric metering there is no constraint on battery lifetime. There is also a
strong requirement of low latency (i.e. long latency times
Electric metering was tested in [21] using LPWAN limit the number of transactions that a store can make).
technology. In this application field, companies typically Thus, NB-IoT is a better fit for this application.
require frequent communication, low latency, and high
data rate. Mostly, they do not require low energy F. Pallet tracking for logistics
consumption neither long battery lifetime as electric
Nowadays, pallets tracking to determine the location and
meters have continuous power source. Moreover,
goods condition are highly desirable in logistics. In the
companies need real time grid monitoring to take
literature, pallet tracking was tested in [25][26] using
immediate decisions (e.g. load, outages, and
LPWAN technology. In this application, the most
interruptions). Thus, Sigfox is inappropriate for this
requirements are device cost and battery lifetime. Pallet
application since it does not handle low latency. On the
tracking is a good example of hybrid deployment
contrary, electric meters can be set up using LoRaWAN-
solution. Logistics companies can have their own
Class-C to ensure very low latency. However, NB-IoT is
network so they have a guaranteed coverage in their
a better fit for this application due to the required high
facilities. Low cost IoT devices could be easily deployed
data rate and frequent communication. Moreover, electric
on vehicles. Sigfox or LoRaWAN public base stations
meters are typically in stationary locations in densely
can be then used when vehicles are outside the facilities
populated areas. It is then easy to ensure NB-IoT
or when goods arrive at customer locations. However,
coverage by cellular operators (LTE).
LoRaWAN allows more reliable communication than
B. Smart farming Sigfox when moving at high speeds [3]. For NB-IoT,
LTE network might not be available for all logistic
In agriculture field, long battery lifetime of sensors locations, which are typically in rural areas. Due to the
devices are required. Using temperature, humidity, and

201
 PerIoT'18 - Second International Workshop on Mobile and Pervasive Internet of Things

low cost, long battery lifetime, and reliable mobile [10] L. Mads, N.H. Cong, V. Benny, K. Istvan, M.P. Elgaard,
communication, LoRaWAN is a better fit for this S. Mads, "Coverage comparison of GPRS, NB-IoT,
application. LoRa, and SigFox in a 7800 km2 area", IEEE Vehicular
Technology Conference, 4-7 Jun 2017, Sydney, Australia.
[11] G.B. Mermer, E. Zeydan, "A comparison of LP-WAN
VI. CONCLUSION technologies: An overview from a mobile operators'
This paper has detailed the technical differences of perspective", 25th IEEE Signal Processing and
Sigfox, LoRaWAN and NB-IoT, and discussed their Communications Applications Conference, 15-18 May
2017, Antalya, Turkey.
advantages regarding IoT factors and major issues. Each
[12] Sigfox world coverage: sigfox.com/en/coverage
technology will have its place in the IoT market. Sigfox [13] LoRa world coverage: lora-alliance.org
and LoRaWAN will serve the lower device cost, very [14] G. Strazdins, A. Elsts, K. Nesenbergs, and L. Selavo,
long range (high coverage), infrequent communication "Wireless sensor network operating system design rules
rate, and very long battery lifetime. Unlike Sigfox, based on real-world deployment survey", Sensor and
LoRaWAN will serves also the local network Actuator Networks, Vol. 2, No. 3, 2013, pp. 509-556.
deployment and the reliable communication when [15] F.J. Oppermann, C.A. Boano, K. Romer, "A Decade of
devices moving at high speeds. In contrast, NB-IoT will Wireless Sensing Applications: Survey and Taxonomy",
serve the higher value IoT applications that are willing to The Art of Wireless Sensor Networks, Vol. 1, No. 2,
2014, pp.11-50.
pay for very low latency and high quality of service.
[16] F. Sforza, Communications system, 26 March 2013, US
Finally, it is expected that the 5th generation (5G) of Patent US8406275 B2.
mobile cellular communication will allow an all- [17] B. Reynders, W. Meert, "Range and coexistence analysis
connected world of humans and devices by the year 2020 of long range unlicensed communication", 23rd
[8], which would lead to a global LPWAN solution for Conference on Telecommunications, 16-18 May 2016,
IoT applications. Thessaloniki, Greece.
[18] K. Mikhaylov, T. Haenninen, "Analysis of Capacity and
REFERENCES Scalability of the LoRa Low Power Wide Area Network
Technology", 22th IEEE European Wireless Conference,
[1] R. Ratasuk, N. Mangalvedhe, A. Ghosh, "Overview of 18-20 May 2016, Oulu, Finland.
LTE enhancements for cellular IoT", 26th Symposium on [19] Y. Wang, X. Lin, A. Grovlen, Y. Sui, "A Primer on 3GPP
Personal, Indoor, and Mobile Radio Communications, 30 Narrowband Internet of Things", IEEE Communication
August-2 September 2015, Hong Kong, China. Magazine, Vol. 55, No. 3, 2016, pp. 117-123.
[2] M. Centenaro, L. Vangelista, A. Zanella, "Long-range [20] A. Adhikary, X. Lin, Y. Wang, "Performance Evaluation
communications in unlicensed bands: The rising stars in of NB-IoT Coverage", IEEE 84th Vehicular Technology
the IoT and smart city scenarios", IEEE Wireless Conference, 18-21 September 2016, Montreal, Canada.
Communications, Vol. 23, No. 5, 2016, pp. 60-67. [21] N. Andreadou, M.O. Guardiola, G. Fulli,
[3] D. Patel, M. Won, "Experimental Study on Low Power "Telecommunication Technologies for Smart Grid
Wide Area Networks (LPWAN) for Mobile Internet of Projects with Focus on Smart Metering Applications",
Things", 85th IEEE Vehicular Technology Conference, Energies Journal, Vol. 9, No. 5, May 2016, pp. 1-35.
4-7 June 2017, Sydney, Australia. [22] R.P. Pratim, "IoT for smart agriculture: Technologies,
[4] U. Raza, P. Kulkarni, "Low Power Wide Area Networks: practices and future direction", Ambient Intelligence and
An Overview", IEEE Communications Surveys & Smart Environments, Vol. 9, No. 4, 2017, pp. 395-420.
Tutorials, Vol. 19, No. 2, 2017, pp. 855-873. [23] S. Nanni, E. Benetti, G. Mazzini, "Indoor monitoring in
[5] A.M. Baharudin, W. Yan, "Long-range wireless sensor Public Buildings: workplace wellbeing and energy
networks for geo-location tracking: Design and consumptions. An example of IoT for smart cities
evaluation", IEEE International Electronics Symposium, application", Advances in Science, Technology and
29-30 September 2016, Denpasar, Indonesia. Engineering Systems, Vol. 2, No. 3, 2017, pp. 884-890.
[6] W. Guibene, J. Nowack, N. Chalikias, M. Kelly, [24] T. Saarikko, U.H. Westergren, T. Blomquist, "The
"Evaluation of LPWAN Technologies for Smart Cities: Internet of Things: Are you ready for what’s coming?",
River Monitoring Use-Case", IEEE Wireless Business Horizons, Vol. 60, No. 5, 2017, pp. 667-676.
Communications and Networking Conference, 19-22 [25] D.H. Kim, J.B. Park, J.H. Shin, J.D. Kim, "Design and
March 2017, San Francisco, CA, USA. implementation of object tracking system based on
[7] O. Vondrous, Z. Kocur, T. Hegr, O. Slavicek, LoRa", IEEE Conference on Information Networking,
"Performance evaluation of IoT mesh networking 11-13 january 2017, Da Nang, Vietnam.
technology in ISM frequency band", 17th IEEE [26] R.F. Garcia, I. Gil, "An Alternative Wearable Tracking
Conference on Mechatronics - Mechatronika, 7-9 System Based on a Low-Power Wide-Area Network",
December 2016, Prague, Czech Republic. Sensors, Vol. 17, No. 3, 2017.
[8] R.S. Sinha, Y. Wei, S.-H. Hwang, "A survey on LPWA
technology: LoRa and NB-IoT", ICT Express, Vol. 3,
2017, pp.14-21.
[9] S. Oh, J. Shin, "Efficient Small Data Transmission
Scheme in 3GPP NB-IoT System", IEEE
Communications Letters, Vol. 21, 2017, pp. 660- 663.

202