2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

Internet of Things (IoT) using LoRa technology

Alireza Zourmand Andrew Lai Kun Hing
School of Engineering Tanand Technology Sdn Bhd
KDU University College C-8-11, Centum@Oasis
Selangor, Malaysia Corporate Park, Malaysia
[email protected] [email protected]

Chan Wai Hung Mohammad AbdulRehman

School of Engineering Tanand Technology Sdn Bhd
KDU University College C-8-11, Centum@Oasis
Selangor, Malaysia Corporate Park, Malaysia
[email protected] [email protected]

Abstract—LoRa is the new communication technology under the conducted an indoor positioning test based on Wi-Fi and LoRa
Low Power Wide Area Network (LPWAN). It emphasizes on the technology respectively. The coverage area of the Wi-Fi based
long-range communication with the high receiving sensitivity ability
which allows it to work under the noise interference or noise floor
positioning system in [1] is only 2 meters while the work in [2] can
effectively. The range of communication has become the critical achieved a maximum coverage area of 200 meters and 28.8 meters
part on most of the IoT system, especially in Wi- Fi and Bluetooth- with high accuracy of positioning result. The result from both the
based IoT system. With the emergence of LoRa technology, further authors show the weakness of Wi-Fi technology in positioning
improvements to applications of the Internet of Things (IoT) can be system when comparing with LoRa technology. In addition, the
realized. By using a single receiver in the LoRa network, it is able to
handle many nodes at multiple locations within the area, unlike Wi- author in [3] stated that the transmission range of Bluetooth is 50
Fi-based system which needs to have many access points to increase feet (without range extender) while for Wi-Fi is 200 feet.
the coverage area. The combination of LoRa and Wi-Fi technology Implementing the solution with Wi-Fi or Bluetooth technology
reduces the cost of deployment of the IoT system. In this paper, the will need a higher cost in order to build a long-range
actual deployment of IoT system using LoRa technology with the communication network which requires additional range
combination of Wi-Fi technology. Besides, this paper presents the
performance and the actual coverage area of the LoRa network in extender. This motivates the development of many Low- Power
both the indoor and outdoor condition. For the indoor condition, a Wide Area Networks (LPWAN) technologies, such as LoRa, to
different set of LoRa parameter including Spreading Factor and fulfil these requirements.
Bandwidth is tested at different locations. As a result, the quality of LoRa (which stands for long range) is a spread spectrum
the LoRa network does not only depend on the distance from the modulation technique derived from chirp spread spectrum (CSS)
gateway but also the effect of path loss due to the structural element.
Moreover, a single LoRa gateway located at the middle floor of the technologies. The modulation technique used in LoRa makes it
building is able to handle the LoRa end-devices at any location robust to channel noise since the entire allocated bandwidth is
within the campus with Spreading Factor of 9. A better indoor used to broadcast a signal (information or data). Furthermore, the
coverage can be improved by adding an additional LoRa gateway security of the LoRa system can be guaranteed as the
in the campus. While for the outdoor condition, the maximum transmission is spread in a pseudo-random way which presents
coverage area in the LoRa network for Spreading Factor of 12 is
330 meters with an indoor gateway deployment. like a noise, hence the modulation technique had provided the
Index Terms—LoRa, Internet of Things, Low-cost LoRa node basic security for the LoRa system [4]. Aside from that, LoRa is
and gateway, monitoring system the best option for IoT solution which required a long range of
data communication while keeping very little power usage [5]. In
other words, the strong penetration of the LoRa signal makes it
I. INTRODUCTION
able to provide enough coverage in a difficult-to-reach indoor
Wi-Fi and Bluetooth are already established wireless tech- location. When compared to Wi-Fi or Bluetooth based IoT
nologies for Local Area Network (LAN). Both have well- solution which has a short range of data communication and
established standards such as Wi-Fi, Bluetooth 4.0, and Zig- bee. throughput (without adding range extender or repeater), this
Due to popularity and easy-to-use protocol, those wireless technology able to provide maximum efficiency in data
technology has dominated the IoT market. There is a downside communication while maintaining low development cost. Based
or challenge with that technology in IoT which is the trans- on the study [6], it stated the main targeted deployment of LoRa
mission range. For the IoT application domains such as smart technology is the smart devices which have limited energy and
positioning system, and smart farming requires high energy- does not require establishing frequent communication all the
efficient sensor nodes that can communicate across a long time. All these features make LoRa become an interesting
distance. In the recent studies [1], [2], both the authors had

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 324

 2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

candidate for the current Internet of Thing (IoT) market and III. LORA FREQUENCY BAND IN MALAYSIA
make it able to compete with other IoT technology such as Wi-
Selection of LoRa frequency bands used is the very first step
Fi and Bluetooth.
before designing a LoRa-based IoT network because each country
In this paper, the focusing point is to conduct several
has its own frequency regulation. According to the lat- est regional
experiments on the LoRa coverage performance and also to
parameters provided by the LoRa AllianceT M [8], it listed two
implement LoRa-based real-time monitoring which is located at
frequency bands/channels for Malaysia which is 433 – 435 MHz
KDU University College.
and 919 – 924 MHz (represented by a channel of EU433 and
AS923 respectively). LoRa communication de- vice is categorized
II. OVERVIEW OF LORAWAN AND LO R A TECHNOLOGY
as a short-range device under the Malaysian Communications and
The LoRaWAN is Low Power Wide Area Network (LP- WAN) Multimedia Commission (MCMC). It is governed by the
protocol and system architecture developed by LoRa Alliance Communications and Multimedia ACT 1998 (CMA). In 2015,
while LoRa (developed by Semtech) defines the physical layer of MCMC allocated 19 of frequency bands for a short-range radio
the system. In addition, LoRaWAN is a media access control communication device (SRD) in the class assignment (CA) No.2
(MAC) layer protocol which built on top of the LoRa physical of 2015 [9]. Although 868 MHz frequency bands are placed under
layer. It defines the network architecture which operates in a non- SRD, it has a limited period of usage as stated in the same class
licensed band below 1000 MHz. LoRaWAN uses AES-128 assignment.
encryption for security which is the AppsKEY and NwkKey. In 17th April 2015, MCMC had announced that the frequency
There are three frequency bands that are most commonly used band of 868 MHz will be taken out for SRD starting on the 1st
which are 433 MHz for Asia, 868 MHz for Europe and 915 MHz January of 2018 [10]. In December 2015, MCMC had allocated
for North America [6]. Machine-to-Machine (M2M) and Internet the new frequency band for the use of SRD as addition to bands
of Thing (IoT) has become the main target of LoRa deployment 433 MHz, 2.4 GHz and 5.8 GHz that can be used for IoT
due to its specification of long-range and low-power applications which is 4 MHz of spectrum in the frequency band
communication. Besides, the adaptive data rate algorithm of 919 to 923 MHz as stated in class assignment No.3 of 2015 [11].
LoRa technology helps to maximize the node’s battery life and
network capacity. A. Equivalent Isotropic Radiated Power limitation
LoRa modulation occurs at the physical layer which uses Apart from the frequency band used in Malaysia, another
proprietary Chirp Spread Spectrum (CSS) modulation derived important parameter will be discussed. EIRP refers to the
from the spread spectrum modulation scheme. The concept of Equivalent Isotropic Radiated Power, which is the radiated output
chirp spread spectrum modulation is transforming a single bit of power refer to an isotopic antenna radiating power equally in all
information into another series of the bit and spread it into the directions and whose gain is expressed in dBi.
entire spectrum. Besides, this type of modulation operates below
the noise level which makes it more robust to noise interference TABLE I
FREQUENCY BANDS THAT ARE AVAILABLE FOR SRD AND ITS OPERATING
and jamming. In other words, the signal or infor- mation is CONDITION (SOURCE FROM CLASS ASSIGNMENT NO. 1 OF 2017 MCMC)
spreading across the wideband. The spread spectrum modulation
is an old modulation technique developed in 1940 which was Frequency Bands Maximum Power
originally used for military communication [7]. The word “chirp” 433 MHz to 435 MHz 100 mW EIRP
919 MHz to 923 MHz 500 mW EIRP
stands for ‘Compressed High-Intensity Radar Pulse’. It defines
that the increasing and decreasing of the signal frequency with Referring to Table I, the two frequency bands have their own
time during the modulation process. There are five important limitation on the maximum power rating listed by MCMC. The
parameters in LoRa physical which are the carrier frequency, maximum of EIRP should be followed when designing the LoRa
transmit power, spreading factor (SF), bandwidth (BW) and code network. There are three parameters involve in the EIRP which is
rate (CR). As mention early, there is three main carrier frequency the output power of the transmitter, cable gain, and the transmitter
which is 433 MHz, 868 MHz and 915 MHz where the LoRa antenna gain shown in Equation (2).
operate at.
EIRP = Pout − Ct + Gt (2)
Overall, LoRa is beneficial for IoT system required long-
distance communication comparing to short-range protocols like The power and the gain of the antenna must be select
Wi-Fi and Bluetooth though there are some disadvantages on the accordingly to avoid exceeding the EIRP limitation set by the
speed of transmission and limitation on the size of the payload. country.

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 325

 2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

IV. RELATED WORK

When the LoRa gateway identified the sync word, it will forward
According to (LoRa Alliance, 2018), the latest update on the the received packet from the end-devices to the local MQTT
ISM band for Malaysia is 920 to 923 MHz. But there is no broker through the ESP8266. The ESP8266 is a Wi-Fi module
research paper or deployment of LoRa gateway using that which will be used to provide an internet interface for the LoRa
frequency band in Malaysia. In [12], [13], [14], the authors tested gateway. The connection between the LoRa gateway and
the performance LoRa using ISM band of 433 MHz (old ISM ESP8266 is through SPI communication which is the same as the
band) in Malaysia. The authors in [12] presented the performance connection with Arduino. For the web interface part, message
of LoRa modulation in the outdoor scenario at Johor, Malaysia. query telemetry transport (MQTT) is used. MQTT is a famous
The authors focus on the Line-of-sight (LOS) and Non-Line-of- machine-to-machine (M2M) communication protocol on the
Sight (NLOS) test on the LoRa network. The transmit power used Internet of Thing (IoT). The ESP8266 on the LoRa gateway will
by the authors was 23 dBm with the antenna gain of 3 dBi. publish the data received from those end-devices to the local
Received signal strength indication (RSSI) was used as the MQTT broker, while Node-RED will be used to show the
measurement parameter in the testing. receiving packet from the end-devices by subscribing to the same
MQTT broker.
Based on the study, the authors concluded that SX1272 is the
most power efficiency (used 0.2423W for each transmission) B. Hardware setup
during transmission while SX1278 has the longest coverage of 1 In this paper, the LoRa end-devices is the combination of
km. While in [14], the authors discussed the performance of the sensor, sx1278 (433 MHz), and Arduino Nano while the LoRa
LoRa network in the forested tropical environment. Same as gateway is the combination of sx1278 (433 MHz) and ESP8266
previous works done by other authors, different spreading factor as mention earlier. All the LoRa end-devices will be powered by
and bandwidth are used to measure the signal strength (RSSI). two 3.7 V lithium rechargeable battery, and the LoRa gateway
The authors conducted two tests on a different location (open will be powered by the 5 V DC power source. Three different
environment and foliage environment) and compared the types of sensor are connected to each of the LoRa end-devices.
obtained result in the graphical method. Based on the result
C. Network interface
obtained, the authors found out that the LoRa signal experienced
much more attenuation in the foliage environment compared Figure 2 illustrated the web interface of the LoRa network.
with the open environment as the distance between the receiver This is the live dashboard from the Node-Red application which
and transmitter increases. displayed the information obtained from the LoRa end- devices
V. I NTEGRATION OF IOT MONITORING SYSTEM IN LO R A such as SNR, RSSI, and sensor data.
NETWORK
A. System Architecture

Fig. 1. Connection of LoRa network

Figure 1 shows the basic connection of the LoRa network in

the first part of the project. Based on the figure, the LoRa end-
devices are communicated to the LoRa single channel gateway in
a star topology using LoRa modulation. The spreading factor and
bandwidth of the LoRa network will be configured in the coding
part before the communication began. Fig. 2. Graphical User Interface (GUI) of the LoRa system using Node-Red

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 326

 2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

VI. PARAMETER TESTING FOR INDOOR SCENARIO

The indoor testing is conducted at seven different locations in
KDU University College (Table II). During the testing, the
LoRa gateway is placed at the Engineering Communication Lab
at the second floor while the LoRa end-device is moved from
one location to another for each test. The distance between the
LoRa transmitter and receiver is measured using the laser
distance meter and measuring tape. In this part, only one LoRa
end-device and LoRa gateway are used for the testing. There is
a total of six modes are used to investigate the indoor
performance of the LoRa network. Based on Table III, each
mode indicated a different combination of spreading factor and
bandwidth while the code rate is constant for all the mode. In
the following experiment, each location is tested with six
different modes and a total of 20 samples is taken for data
analysis.

Average RSSI
7 Mode 6 (SF12,
BW 250)

6 Mode 5 (SF12,
BW 125)
5 Mode 4 (SF9, BW
Location

A. Result 250)
4
At each test location (test no.), the average received signal Mode 3 (SF9, BW
125)
strength indicator (RSSI), average signal-to-noise ratio (SNR), 3
Mode 2 (SF7, BW
and packet loss rate of different mode is tabulated in Tables IV, 2 250)
V and VI. The result in the table was converted into different Mode 1 (SF7, BW
graph format to have better visualization on the result (Figure 3, 1 125)

4 and 5).
-150 -100 -50 0
RSSI, dBm

Fig. 3. Average RSSI at each test location for different mode

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 327

 2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

SNR of the signal. Based on the observation, the testing location

Average SNR at each test location which is closest to the gateway location and lesser obstacle in the
12
transmission path (lower overall obstacle percentage) such as
10
location 2 and 3 shows different result than others when
8
6 Mode 1 (SF7, BW 125) increasing the spreading factor. A significant drop of SNR of the
4 Mode 2 (SF7, BW 250)
signal is shown when the spreading factor is increased from 9 to
12 at location 2 and 3.
SNR, dB

2
0 Mode 3 (SF9, BW 125)
Based on Figure 5, location 4 is the only test site that has packet
-2 1 2 3 4 5 6 7 Mode 4 (SF9, BW 250)
loss during the testing. In mode 1, location 4 has the highest packet
-4
-6 Mode 5 (SF12, BW 125) loss rate (45 %), while in mode 2, location 4 shows lesser packet
-8 Mode 6 (SF12, BW 250)
loss (10 %). This shows that increasing the bandwidth can reduce
-10 the packet loss rate. In addition, further increasing the spreading
Location
factor (Mode 3 to Mode 6) eliminates the chance of the packet loss
Fig. 4. Average SNR at each test location for different mode as a higher spreading factor give better sensitivity in the
communication.
Based on Figure 3, the average RSSI of the entire test result is in
between -90 dB and -120 dB. Besides, location 2 which has the B. Discussion
shortest distance between the receiver shows the lowest average
In indoor testing which is non-line-of-sight (NLOS), the distance
RSSI value while location 4 and 7 show the highest average RSSI
between the LoRa node and the LoRa gateway does affect the
value in each mode. Besides, the average RSSI value decreases
quality of the communication. This is because of the path loss due
when the spreading factor increases at the same bandwidth
to the structural element in the campus such as glass, metal, brick
(Mode 1, Mode 3, and Mode 5 for bandwidth 125 kHz and Mode
wall and more. The different element has a different effect on the
2, Mode 4, and Mode 6 for bandwidth 250 kHz).
level of signal loss as stated in [15]. Under the same environment,
Figure 4 shows the average SNR collected at each test location. further increasing the distance from the LoRa gateway will reduce
Based on the figure, location 1, 4, and 7 are the area which has the quality of the packet (in term of RSSI and SNR). Apart from
higher noise interference compared to location 2, 3, 5 and 6 as that, there is another factor that affects the quality of the
lower SNR indicated that more noise was presented in the communication which is the interference of other electronic
communication link than signal. Among all the location, location components or devices around the campus. This effect was present
2 shows the highest average SNR in all the mode while location at location 4 which is the only location that is crowded with people
7 shows the lowest average SNR in all the mode. Furthermore, during testing and also near to the radio studio in the campus.
increasing the bandwidth from 125 kHz to 250 kHz with the same Even location 4 is located on the same floor with location 5 (but
spreading factor is also reduced the SNR reading (decreasing in different distance), all the LoRa packet received from location 4
receiving sensitivity) except for spreading factor 7 (Mode 1, and was below the noise floor (negative SNR is shown in Figure 4)
Packet Lost rate, % especially for mode 1 which reaches the maximum LoRa
demodulator sensitivity at that particular spreading factor (-7.5
50 dB). In addition, this is also affecting on the communication link
Mode 1 (SF7, BW 125) whereby the packet is not able to transmit in a good condition and
40
more packet loss during the transmission due to the increase of
Packet Lost rate

Mode 2 (SF7, BW 250)

30 noise floor level at the environment. In general, 2.5 dB of extra
Mode 3 (SF9, BW 125)
link budget is provided in each increment of spreading factor. By
20 Mode 4 (SF9, BW 250)
increasing the spreading factor can improve the sensitivity of the
10
Mode 5 (SF12, BW 125)
communication path, therefore reduce the packet loss.
Mode 6 (SF12, BW 250)
In summary, the deployment of LoRa gateway at the middle of
0
1 2 3 4 5 6 7
the campus was able to cover the entire building with a minimum
Location spreading factor of SF 9 and above without any packet loss during
transmission (minimum SNR of -5 dB for spreading factor 9). The
Fig. 5. Overall packet loss rate at each test location
average RSSI value for the entire testing location was in the range
between -90 dBm to -120 dBm. Even the SNR of the received
2). In addition, an increase of spreading factor with the same packet sometimes reached negative reading, it does not affect the
bandwidth (for example from mode 1 to mode 3) increases the quality of the transmission whereby the packet is able to reach at
the receiver side without any losses.

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 328

 2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

occur at inside and outside of the campus. In general, when the

LoRa node sends a message to the LoRa gateway, the power of the
signal that is radiated in the air will reduce greatly as the distance
increases. This is because of the nature of the antenna
(omnidirectional) used in the testing. The sender antenna converts
the signal into a form of power and sends it through the air in all
direction. In other words, the power spread into a different
VII. LORA PERFORMANCE TESTING IN direction and most of the power is lost because it does not flow to
the receiver antenna. Furthermore, the penetration of signal power
OUTDOOR SCENARIO (Partial-NLOS)
through the obstacle in the sending path will also reduce the power
Table VII shows the LoRa parameter setup in the outdoor of the sending signal as mention in the indoor test. When the
coverage test. In this part, the spreading factor and bandwidth are power of the signal from the sender is lower than the maximum
set to 12 and 125 kHz respectively which can provide maximum receiver sensitivity, the LoRa gateway is not able to detect and
coverage distance and increase the sensitivity of the pick up the signal. Hence, better coverage of LoRa
communication. The transmit power and antenna gain is chosen communication can be achieved when it is conducted in a line-of-
based on the EIRP limitation stated in Section 3.1, Table I. sight situation, for example, placed the LoRa gateway near to the
The LoRa gateway was placed inside the communication window instead of a room.
laboratory at 2nd floor and the LoRa node (sender) moved away
from the LoRa gateway while capturing the data. During the VIII. CONCLUSION
testing, the LoRa node sends a message to the LoRa gateway
every 10 seconds and the LoRa gateway will send back a LoRa technology has the advantages of working under noise
confirmation message to the LoRa node which indicated that the interference and allow long-distance communication thanks to its
communication is successful (a LED will blink on the LoRa modulation technique. In this paper, the performance and
node). coverage result of the LoRa indoor and outdoor deployment is
presented using a selected combination of Spreading Factor and
Bandwidth setting. The LoRa network consists of a low-cost
A. Result single channel gateway and a private network application. All the
Based on Table VIII, the maximum detectable distance of the
measurement was recorded in a real environment at KDU
LoRa communication network is 330 meters which has a signal
University College campus. In summary, the environment of the
SNR of -16.75 dB and RSSI of -130 dBm. Furthermore, as the
deployment of the LoRa network does affect the signal quality
distance from the gateway increases, the quality of the signal
especially the high dense of obstruction level in between the
reduced. From the result of the testing, at a distance above 120
transmission path has the most impact on the degradation of the
meters from the gateway, the communication system starts to
signal quality as seen in the experiment. Spreading factor and
work below the noise floor (negative SNR value).
Bandwidth were the core factors that influence the performance
B. Discussion of the LoRa network. Higher Spreading Factor and lower
Bandwidth setting allow a longer range of communication and
In this part, the maximum communication coverage for an
higher noise immunity. This shows the advantages of LoRa
indoor LoRa gateway setup is limited to 330 meters which do not
network comparing to others wireless protocol. The LoRa
meet the minimum distance of 1 km stated in the datasheet. This
network is able to cover the entire building with one gateway or
is mainly caused by the higher noise interference and path loss
receiver deployment. At the meantime, the LoRa parameter
should be tuned to the optimum setting for each location as
different LoRa parameter setting will have different power
consumption which will be analyzed in the future. By utilizing
the optimum setup, the devices in the LoRa network will be able
to work for a few years with the off-grid supply such as a battery.

ACKNOWLEDGMENT
This project is funded by Tanand Technology Sdn. Bhd and
supported by KDU University College.

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 329

 2019 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS 2019), 29 June 2019, Selangor, Malaysia

REFERENCES

[1] A. Zourmand, N. W. Sheng, A. L. K. Hing, and M. AbdulRehman, “Human [9] M. Communications and M. Commision, “Class Assignment No. 2 of
Counting and Indoor Positioning System Using WiFi Technology,” in 2015,” 2015. [Online]. Available: https://www.mcmc.gov.my/resources/
Automatic Control and Intelligent Systems, Shah Alam. IEEE, October registered-assignments/register-of-class-assignments.aspx
2018, pp. 142–147. [Online]. Avail- able: [10] M. Communications and M. C. (MCMC), “Usage of Short Range Device
https://www.researchgate.net/publication/330253846_Human_ (SRD) and Radio Frequency Identification Device (RFID),” 2015. [Online].
Counting_and_Indoor_Positioning_System_Using_WiFi_Technology Available: https://www.mcmc.gov.my/skmmgovmy/
[2] C. Wongeun, C. Yoon-Seop, J. Yeonuk, and S. Junkuen, “Low-Power LoRa media/General/pdf/Public-Notice-RFID-SRD.pdf
Signal-Based Outdoor Positioning Using Fingerprint Algorithm,” Goe- [11] M. Communications and M. Commision, “Class Assignmnet No. 1 of
Information, vol. 7, no. 11, pp. 440–455, 2018. [Online]. Available: 2017,” 2017. [Online]. Available: https://www.mcmc.gov.my/resources/
10.3390/ijgi7110440;https://www.mdpi.com/2220-9964/7/11/440/pdf registered-assignments/register-of-class-assignments.aspx
[3] J. Teel, “Bluetooth or WiFi – Which is Best for Your New Wireless [12] K. A. Ahmad, J. D. Segaran, F. R. Hashim, and M. T. Jusoh, “LoRa
Product?” 2018. [Online]. Available: https://predictabledesigns.com/ what- Propagration at 433 MHz in Tropical Climate Environment,” Journal of
type-of-wireless-is-right-for-your-product-bluetooth-wifi/ Fundamental and Applied Sciences, vol. 9, no. 3S, pp. 384–394,
[4] E. Notes, “LoRa Physical Layer & RF Interface,” 2018. [Online]. 2017. [Online]. Available: 10.4314/jfas.v9i3s.31;https:
Available: https://www.electronics-notes.com/articles/connectivity/lora/ //www.researchgate.net/publication/322809453_LoRa_propagation_at_
radio-rf-interface-physical-layer.php 433_MHz_in_tropical_climate_environment
[5] “Why LoRa?” [Online]. Available: https://www.semtech.com/lora/why- [13] S. Daud, T. S. Yang, M. A. Romli, Z. A. Ahmad, N. Mahrom, and
lora R. A. A. Raof, “Performance Evaluation of Low Cost LoRa Modules
[6] L. Alliance, “White Paper: A Technical Overview of LoRa in IoT Applications ,” IOP Conference Series: Materials Science and
and Lorawan,” 2015. [Online]. Available: https: Engineering, vol. 318, no. 1, 2018. [Online]. Available: 10.1088/1757-
//www.tuv.com/media/corporate/products_1/electronic_components_ 899X/318/1/012053;https:
and_lasers/TUeV_Rheinland_Overview_LoRa_and_LoRaWANtmp.pdf //www.researchgate.net/publication/323860006_Performance_
[7] S. R. Sharan, W. Y. Qiao, and H. Seung-Hoon, “A survey on LPWA Evaluation_of_Low_Cost_LoRa_Modules_in_IoT_Applications
technology: LoRa and NB-IoT,” ICT Express, vol. 3, no. 1, 2017. [Online]. [14] K. A. Ahmad, M. S. Salleh, J. D. Segaran, and F. R. Hashim, “Impact
Available: 10.1016/j.icte.2017.03.004 of foliage on LoRa 433MHz propagation in tropical environment,” ser. 1,
[8] L. A. T. C. R. P. Workgroup, “LoRaWANTM 1.1 Regional Parameters,” vol. 1930. AIP, February 2018. [Online]. Available:
January 2018. [Online]. Available: https://lora-alliance.org/resource- https://www.researchgate.net/publication/322809453_LoRa_
hub/lorawantm-regional-parameters-v11rb propagation_at_433_MHz_in_tropical_climate_environment
[15] R. Schaupp, “LoRaWAN range, part 1: The most important factors
for a good LoRaWAN signal range,” October 2018. [Online].
Available: https://smartmakers.io/en/lorawan-range-part-1- the-most-
important-factors-for-a-good-lorawan-signal-range/

978-1-7281-0784-4/19/$31.00 ©2019 IEEE 330