Factors Inﬂuencing LoRa Communication in IoT Deployment: Overview

and Experience Analysis

Thierry Antoine-Santoni

1 a

, Bastien Poggi

1 b

, David Araujo

and Chabi Babatounde

UMR CNRS SPE 6134, Universit

a di Corsica, Corte, France

UMS CNRS STELLA MARE 3514, Universit

a di Corsica, Biguglia, France

Keywords:

Lora, RSSI, PCA, Performance.

Abstract:

LoRa communication offers wireless sensor networks deployment for system or environmental monitoring

over long distances and with low energy consumption. However, this radio communication technology is

subject to environmental disturbances. In this paper, we propose an overview of the studies carried out on

LoRa signal disturbances, taking the RSSI as a comparison parameter. Secondly, we extract the main inﬂu-

ences to compare them with the data collected on the experimental platform of the Smart Village of Cozzano

(Mediterranean area, Southern of Corsica island), a scientiﬁc program aiming to develop digital tools for the

monitoring and the preservation of the environment. We use one of the most popular techniques in multivariate

statistics, especially when analyzing large datasets, the principal component analysis (PCA). The results show

the impact of some environmental parameters on communication quality.

1 INTRODUCTION

Environmental monitoring is a crucial issue in the

scientiﬁc ﬁeld. The smart village, a scientiﬁc

project, proposes to develop a set of technologies

for the environment observation and preservation

and its inhabitants in a rural, mountainous and iso-

lated area. (Antoine-Santoni et al., 2019b),(Antoine-

Santoni et al., 2019a). In this context, it is essential to

think about a correct deployment of the devices to en-

sure the transmissions quality from the deployed sen-

sors.

With a view to sustainable development, the scien-

tiﬁc program has set up an information system based

on a wireless sensor network using LoRa communi-

cation technology with a LoRaWAN protocol. This

technological choice aligns with a desire for energy

efﬁciency to limit the impact of technologies on the

environment and human maintenance over time. In-

deed, we can ﬁnd many works using LoRa technol-

ogy for telemetry purposes (Haxhibeqiri et al., 2018):

smart cities, industry, transport, agriculture, etc. We

see many results presenting the various applications

of LoRa, but in (Shanmuga Sundaram et al., 2020),

the authors specify the axes of reﬂection around the

https://orcid.org/0000-0002-6645-9311

https://orcid.org/0000-0002-3253-6868

Figure 1: Smart Village of Cozzano (South of Corsica) -

Deployment of Lora devices.

scientiﬁc challenges around LoRa/LoraWAN. Within

the Smart Village, illustrated by Figure 1, in the vari-

ous applications deployed, we can ﬁnd weather sta-

tions on different points of the village and in alti-

tude and GPS trackers to locate animals raised in the

wild. However, despite the relative system stability,

we have noticed variations in the signal quality per-

formance, impacting the data transmission quality of

the information feedback. These failures create gaps

in the database. In this paper, we propose to make

a statistical analysis of the collected environmental

data to evaluate the impact of the environment on the

signal and determine the environmental factors inﬂu-

274

Antoine-Santoni, T., Poggi, B., Araujo, D. and Babatounde, C.

Factors Inﬂuencing LoRa Communication in IoT Deployment: Overview and Experience Analysis.

DOI: 10.5220/0011102600003194

In Proceedings of the 7th International Conference on Internet of Things, Big Data and Security (IoTBDS 2022), pages 274-280

ISBN: 978-989-758-564-7; ISSN: 2184-4976

encing long-range communication. In a ﬁrst step, we

will make a literature review on the studies conducted

around the inﬂuence of the environment on the LoRa

signal, taking the RSSI as a parameter. We confront

them with a statistical analysis on the devices de-

ployed in the Smart Village since 2018. We will then

conclude on the areas of improvement of these dif-

ferent parameters and improve communications and

network coverage.

2 STATE OF ART

LoRa, for “Long Range”, is a long-range wireless

communications system promoted by the LoRa Al-

liance (Alliance, 2015). This system aims at being us-

able in long-lived battery-powered devices, where the

energy consumption is of paramount importance (Au-

gustin et al., 2016). Lora distinguishes two main lay-

ers: a physical layer using the Chirp Spread Spectrum

(CSS) radio modulation technique and a MAC layer

protocol (LoRaWAN). LoRa is a technology that op-

erates in the ISM, Industrial, Scientiﬁc and Medical

(ISM) frequencies bands, 868 and 433 MHz in Europe

and 915 MHz in the USA. These different frequencies

can, however, have a performance impact on the net-

work as studied in (Alset et al., 2020) where the au-

thors determined with simulations that the lifespan of

the batteries can be affected by the carrier frequency

in use. With long-range and low power capacities,

LoRa belongs to the LPWAN category. Semtech, an

owner of the technology, claims an effective range

of thirty miles (sem, ), or approximately forty-eight

kilometres, in urban areas. However, with the ade-

quate material, it is possible with LoRa, to transmit

over a hundred kilometres (Jovalekic et al., 2018).

Thanks to the low consumption of LoRa devices, a

connected device can run for several years on a bat-

tery. LoRa technology is at the heart of many aca-

demic and industrial works, and many scientiﬁc ques-

tions are still open, as revealed by (Shanmuga Sun-

daram et al., 2020). The principal value in use when

studying a LoRaWAN, and generally signals, is prob-

ably the Received Signal Strength Indication (RSSI).

The RSSI is a measure of the power at the recep-

tion of a signal. It formulates in decibel-milliwatts

(dBm). In general, the RSSI usually spans between

-60 and -130 dBm; the closer value is to 0, the better

the RSSI and, on the contrary, a signal relative to -130

dBm means a weak reception. The network architec-

ture of LoRaWAN uses a star topology where the data

from all the devices gather to a gateway. LoRaWAN

is a data link layer protocol to provide a low power

connectivity system to battery-powered devices. The

current LoRaWAN speciﬁcation is 1.1 (the most used

speciﬁcation is 1.0.3). The gateway can then trans-

mit these data to a server. Through the Adaptive Data

Rate, a LoRaWAN gateway can change the data rate

by changing its spreading factor (SF). The CSS mod-

ulation uses a Spreading Factor (SF) to spread the in-

formation over the frequency (from 7 to 12), deter-

mining the number of bits necessary to transmit the

same amount of data. A higher number of bits per

symbol increases the capability of the receiver to de-

modulate the message. Higher SF means that more

bits are necessary to send the same information. By

increasing the SF, the range can be increased at the

cost of data throughput (Zhu et al., 2019), through-

put varying from 0.3 kilobits per second to 50 kilo-

bits per second. As with every means of commu-

nication, LoRa can suffer from external factors that

would diminish the signal quality, starting with the ur-

ban environment where IoT networks can use the LP-

WANs. This kind of impact has been studied in (Vil-

larim et al., 2019; Dambal et al., 2019; Inagaki et al.,

2019; Villarim et al., 2019; Yousuf et al., 2018). In

(Dambal et al., 2019) the authors study the impact of

a rural environment over a LoRa message. Thus, in an

urban situation, positioning the antennas is extremely

important. By increasing the antenna height, the cov-

erage will increase too, as the height of the buildings

can be an obstacle for the signal. As buildings oc-

cupy the Fresnel zone, the signal can be heavily im-

pacted. In a village like Cozzano, the buildings tend

to have fewer ﬂoors, thus less height, potentially re-

ducing the mentioned interference. However, in (Vil-

larim et al., 2019), the authors estimate that vegeta-

tion might be even more of an obstacle to the propa-

gation of Lora signal than buildings. It is important

to consider that Cozzano is located in the mountains

with much Mediterranean vegetation. This vegetation

needs to be taken into account when measuring the

values of the RSSI, according to (Iova et al., 2017).

By comparing the measures made in an airport and

a forest, the authors observed that while 95% of the

received data in an open ﬁeld, only 80% are received

when vegetation obscures the path.

Furthermore, it appears from their experimenta-

tions that vegetation could reduce the range at which

a message can be transmitted: from 500 meters in an

open ﬁeld, the signal reached about 90 meters in a for-

est, this kind of impact needs to be taken into account

as it could severely damage the reception of the data.

These results are comforted by other works such as

(Wiyadi et al., 2020; Ali et al., 2019; Ansah et al.,

2020; Hidayat et al., 2019; Elijah et al., 2019) where

we can observe an impact of the vegetation on signal

propagation. In (Ali et al., 2019) the authors men-

Factors Inﬂuencing LoRa Communication in IoT Deployment: Overview and Experience Analysis

275

Table 1: Comparison of works surrounding LoRa.

Publication Impacted parameters Distance Temperature Snow Vegetation Mobility Buildings

(Souza Bezerra et al.,

2019)

RSSI - - -

(Dambal et al., 2019) RSSI - -

(Boano et al., 2021) RSSI, PDR -

(Pet

arvi et al., 2017) RSSI, PDR - -

(Iova et al., 2017) RSSI - -

(Ali et al., 2019) RSSI, SNR, PDR -

(Alset et al., 2020) RSSI, PDR -

(Ansah et al., 2020) RSSI, SNR, PDR -

(Avila-Campos et al.,

2019)

RSSI, SNR -

(Qaraqe et al., 2020) RSSI - x

(Inagaki et al., 2019) RSSI - -

(Hidayat et al., 2019) RSSI, PDR - - -

(Wiyadi et al., 2020) RSSI, PDR - -

(Elijah et al., 2019) RSSI, PDR - - -

(Villarim et al., 2019) RSSI, PDR - - -

(Doroshkin et al., 2019) RSSI, SNR - x

(Yousuf et al., 2018) RSSI, PDR - x -

tion the vegetation occupying the Fresnel zone and

that this is the reason why the vegetation impact the

signal. Thanks to the equations surrounding the Fres-

nel zone, it is possible to estimate more precisely the

impact vegetation could have on a LoRaWAN. SF is

an adjustable parameter to increase the range of the

LoRa signal at the cost of throughput, representing

a solution inside dense vegetation. A signal emitted

with a higher SF tends to be more robust than one cast

with a lower SF, further increasing the signal capacity

to pass through vegetation.

Mobility is already been studied in (Pet

arvi

et al., 2017; Doroshkin et al., 2019; Qaraqe et al.,

2020). According to these studies, speed is a pa-

rameter that can negatively impact the RSSI. For

example, with an SF 12, an approximate speed of

40 km/h is enough to impact the RSSI value nega-

tively(Pet

arvi et al., 2017). However, according

to (Doroshkin et al., 2019), the impact of the speed

already observed in (Pet

arvi et al., 2017) cannot

be applied in Line of Sight (LOS) scenarios as the

multipath phenomenon can amplify the impact of the

Doppler effect.

Furthermore some works studied the impact of

temperature on a LoRa signal, such as (Souza Bezerra

et al., 2019) or (Boano et al., 2021). In (Souza Bezerra

et al., 2019) the authors studied the impact of tem-

perature on a LoRaWAN device in a Swedish town,

while in (Boano et al., 2021) the authors studied the

impact of temperature on a heated bed. From the re-

sults, relatively high temperature can hurt the qual-

ity of a LoRa network as it could decrease the RSSI.

Even if heat has a negative impact, too cold weather

associated with snow seems equally harmful. Indeed

in (Souza Bezerra et al., 2019), the authors precise

that snow hinders signal propagation.

The references analysing the impacts on the sig-

nal have been summarised in Table 1 where the sig-

niﬁcant effects mentioned have been noted as un-

favourable (-) or neutral (x). The LoRa parameters

that are studied are generally a combination of the

RSSI, the Signal to Noise Ratio (SNR) and the per-

centage of packet lost or received. To make the anal-

ysis more understandable, we consider that studying

the data portion of received packets (PDR) versus

transmitted packets is identical to the packet loss rate

(PL) because they are opposite.

3 COZZANO’S RSSI MAPPING

The data pass through an information system devel-

oped speciﬁcally for the Smart Village to collect the

information, as illustrated in Figure 2.

The central communication technology of the

Smart Village being LoRa, we felt a need for a bet-

ter understanding of how the signal spread around

the village of Cozzano. For this purpose, we devel-

oped LoRa devices allowing us to transmit a location

through a message to the LoRaWAN gateway on the

church bell tower. The used LoRa gateway is a ker-

link device (Kerlink, 2022). All the used devices on

IoTBDS 2022 - 7th International Conference on Internet of Things, Big Data and Security

276

Figure 2: Information system architecture.

the Smart Village are Class C of Lora.

We walked around the village with the devices and

regularly sent our position to the gateway to take these

measures. The transmissions used a spreading factor

of 7 and a bandwidth of 125 kHz. Using the data col-

lected in the Smart Village, we were able to produce

the map presented in Figure 3. We can see coloured

circles representing the RSSI on a location. In Fig-

ure 3, the gradient from green to red represents an in-

crease in the quality of the RSSI. However, these cir-

cles are numerous and difﬁcult to observe separately.

The rings regroup 30 square meters wide. By calcu-

lating the circle’s mean value, we attributed a colour

to each one, following the already used colour scale.

Furthermore, we can see there is more than one square

Figure 3: Map of RSSI distribution in Cozzano.

type. The standard squares are measured from our re-

sults, while the barred ones are estimated from the

neighbouring measured squares. For simplicity, we

assumed that the propagation of the signal would be

regular on really short distances such as 30 meters

and estimated the value of the barred squares as the

mean of the surrounding squares (in the case where

there are at least ﬁve neighbours). We attributed a

colour to these estimated squares according to the

previous colours gradient. As presented on 3, there

is a substantial evolution of the RSSI within the vil-

lage. Thus, RSSI varies from strong values (around

-80 dBm) to fragile ones (down to -131 dBm). The

village’s topography can explain this range of val-

ues; granite stone buildings strongly impact a signal

and cause rebound through the alleys, deteriorating

the signal. Thus, the more open ﬁelded areas can

transmit a message with high RSSI. While taking our

measures, we could also retrieve the RSSI on a trail

above the village. It cannot be seen, but the trail alti-

tude gives a clear viewpoint gateway’s position on the

bell tower. Without any obstacles, the message can

be transmitted to the gateway in the best condition;

thus, the high values of the RSSI we can ﬁnd on this

trail despite the gateway distance. From these results,

we can see that buildings signiﬁcantly impact mes-

sage transmitting. The presence of buildings can lead

to RSSI lower than -100 dBm for a distance approx-

imating 150 meters, while in an open ﬁeld situation,

messages are received with higher RSSI for greater

distances.

4 STATISTICAL ANALYSIS

Figure 4: PCA of the trackers datas.

From the analysis of the RSSI values, the ques-

tion is to know which parameters inﬂuence the LoRa

signal strength. We, therefore, relied on the pop-

ular statistical method, Principal components analy-

sis (PCA) (Jolliffe, 2013). The principal component

analysis allows extraction and visualising informa-

tion from a multivariate data table. PCA synthesises

this information into just a few new variables called

principal components. These new variables are a lin-

ear combination of the original variables. The num-

ber of principal components is less than or equal to

the number of original variables. The information

Factors Inﬂuencing LoRa Communication in IoT Deployment: Overview and Experience Analysis

277

Figure 5: PCA of the datas from the ﬁreﬁghters station.

Figure 6: PCA of the datas from the saffron ﬁeld.

Figure 7: PCA of the datas from the Casteddu.

contained in a dataset is the total variance or iner-

tia. The objective of PCA is to identify the directions

(i.e., principal axes or principal components) along

which the variation in the data is maximum. We tried

to deﬁne the parameters above by running the PCA

through multiple emitters. The ﬁrst emitters chosen

for this analysis were GPS trackers (Rf-track, 2022)

on the animals. We can retrieve a certain amount of

data through these trackers, such as position, RSSI,

spreading factor, frequency, movement speed, and di-

rection. We can calculate the distance between the

emitter and the gateway at each transmission with a

known location. We used this distance instead of the

location in the PCA. For the experiment, tracker de-

vices were placed on different herd animals. These

animals move freely on a vast mountainous territory.

We retrieved around fourteen thousand transmissions

and the PCA results presented in Figure 4. Many

models such as Free Space Path Loss or Okumura-

Hata already consider distance when estimating RSSI

values. Thus, it seems coherent to see distance nega-

tively correlated to the RSSI. According to this PCA,

the SF appears also negatively correlated to the RSSI.

The SF could have changed to assure the transmis-

sion of the message in harsh conditions. It could ex-

plain the aforementioned negative impact. Further-

more, the animals usually travel at low speed, which,

according to literature, isn’t enough to impact LoRa

signal propagation. This observation can explain the

lack of correlation between the speed and the RSSI

represented in Figure 4. The minor variations in fre-

quency operated in LoRaWAN don’t seem to impact

the RSSI either as the RSSI and frequency vectors are

at an angle close to 90°. Thus it is unlikely that the

differences in frequency provoked by the channels in

use by LoRaWAN are detrimental to the signals.

Another exciting piece of information is the im-

pact of atmospheric pressure, as the weather could

affect signal propagation. It seems that with higher

pressure, we obtain a better RSSI.

5 WAY TO ENHANCE LoRa

DEVICES DEPLOYMENT

We evoke the axis in which we will contribute to im-

proving the deployment of wireless sensor networks.

We imagine coupling a Machine Learning algorithm

to predict positions and optimize these positions. Our

idea base itself on the coupling of two algorithms for

placement optimization:

• The k-nearest neighbours algorithm (KNN),

which allows an accurate estimation of the signals

on the deployment area from localized measure-

ments

IoTBDS 2022 - 7th International Conference on Internet of Things, Big Data and Security

278

• Hitchcock bird-inspired algorithm (HBIA) opti-

mizes the devices’ positioning according to dif-

ferent parameters: environmental parameters and

signal quality according to the deﬁned areas of in-

terest.

In the example in Figure 8, we can see different

signal qualities collected by ﬁeld measurements. It

can see that the signal from the antenna is worse to the

northeast of the antenna than to the southeast, proba-

bly due to obstacles (vegetation, buildings) or envi-

ronmental conditions. If we want to place the antenna

in an optimal way (RSSI, gateway link) in an area of

interest, we use an optimisation process; the yellow

diamond represents the result in the Figure 8. This

process can reproduce in different regions of the map.

Figure 8: Example of best position found by signal estima-

tion.

6 CONCLUSION AND

PERSPECTIVES

This paper ﬁrst presents a literature review on the fac-

tors inﬂuencing LoRa signals. In a second step, we

have analysed the RSSI results in a village in Corsica

with its own LoRA network. We see that environ-

mental factors, as well as granite buildings strongly,

inﬂuence the signal quality. These exploratory works

lead us to think about optimisation of the deployment

of the devices by using machine learning and optimi-

sation algorithms.

REFERENCES

What is LoRa? | Semtech LoRa Technology | Semtech.

https://www.semtech.com/lora/what-is-lora. Ac-

cessed on 2020-01-22.

Ali, N. A. A., Latiff, N. A. A., and Ismail, I. S. (2019).

Performance of LoRa Network for Environmental

Monitoring System in Bidong Island Terengganu,

Malaysia. International Journal of Advanced Com-

puter Science and Applications (IJACSA), 10(11).

Number: 11 Publisher: The Science and Information

(SAI) Organization Limited.

Alliance, L. (2015). White Paper: A Technical Overview of

Lora and Lorawan. Technical report, CA, USA.

Alset, U., Kulkarni, A., and Mehta, H. (2020). Performance

Analysis of Various LoRaWAN Frequencies For Op-

timal Data Transmission Of Water Quality Parame-

ter Measurement. In 2020 11th International Confer-

ence on Computing, Communication and Networking

Technologies (ICCCNT), pages 1–6, Kharagpur, India.

IEEE.

Ansah, M. R., Sowah, R. A., Meli

a-Segu

ı, J., Katsriku,

F. A., Vilajosana, X., and Banahene, W. O. (2020).

Characterising foliage inﬂuence on LoRaWAN

pathloss in a tropical vegetative environment. IET

Wireless Sensor Systems, 10(5):198–207.

eprint:

https://onlinelibrary.wiley.com/doi/pdf/10.1049/iet-

wss.2019.0201.

Antoine-Santoni, T., Poggi, B., Vittori, E., Hieux, H. V.,

Delhom, M., and Aiello, A. (2019a). Smart entity –

how to build DEVS models from large amount of data

and small amount of knowledge ? In Proceedings

of 11th EAI International Conference on Simulation

Tools and Techniques, pages 615–626. EAI.

Antoine-Santoni, T., Poggi, B., Vittori, E., Manicacci, F.-

M., Gualtieri, J.-S., and Aiello, A. (2019b). Proposi-

tion of a smart environment architecture for resources

monitoring and rural activities management. In Pro-

ceedings of SENSORCOMM 2019, pages 62–68.

Augustin, A., Yi, J., Clausen, T., and Townsley, W. (2016).

A Study of LoRa: Long Range & Low Power Net-

works for the Internet of Things. Sensors, 16(9):1466.

Avila-Campos, P., Astudillo-Salinas, F., Vazquez-Rodas,

A., and Araujo, A. (2019). Evaluation of LoRaWAN

Transmission Range for Wireless Sensor Networks in

Riparian Forests. In Proceedings of the 22nd Inter-

national ACM Conference on Modeling, Analysis and

Simulation of Wireless and Mobile Systems - MSWIM

’19, pages 199–206, Miami Beach, FL, USA. ACM

Press.

Boano, C. A., Cattani, M., and R

omer, K. (2021). Impact

of Temperature Variations on the Reliability of LoRa

- An Experimental Evaluation. pages 39–50.

Dambal, V. A., Mohadikar, S., Kumbhar, A., and Gu-

venc, I. (2019). Improving LoRa Signal Coverage

in Urban and Sub-Urban Environments with UAVs.

arXiv:1902.11243 [eess]. arXiv: 1902.11243.

Doroshkin, A. A., Zadorozhny, A. M., Kus, O. N.,

Prokopyev, V. Y., and Prokopyev, Y. M. (2019). Ex-

perimental Study of LoRa Modulation Immunity to

Doppler Effect in CubeSat Radio Communications.

IEEE Access, 7:75721–75731. Conference Name:

IEEE Access.

Elijah, O., Rahman, T. A., Saharuddin, H., and Khairodin,

F. (2019). Factors that Impact LoRa IoT Communi-

cation Technology. In 2019 IEEE 14th Malaysia In-

ternational Conference on Communication (MICC),

pages 112–117, Selangor, Malaysia. IEEE.

Haxhibeqiri, J., De Poorter, E., Moerman, I., and Hoebeke,

J. (2018). A Survey of LoRaWAN for IoT: From Tech-

nology to Application. Sensors, 18(11):3995.

Factors Inﬂuencing LoRa Communication in IoT Deployment: Overview and Experience Analysis

279

Hidayat, M. S., Nugroho, A. P., Sutiarso, L., and Okayasu,

T. (2019). Development of environmental monitor-

ing systems based on LoRa with cloud integration for

rural area. IOP Conference Series: Earth and Envi-

ronmental Science, 355:012010.

Inagaki, K., Narieda, S., Fujii, T., Umebayashi, K., and

Naruse, H. (2019). Measurements of LoRa Propaga-

tion in Harsh Environment: Numerous NLOS Areas

and Ill-Conditioned LoRa Gateway. In 2019 IEEE

90th Vehicular Technology Conference (VTC2019-

Fall), pages 1–5, Honolulu, HI, USA. IEEE.

Iova, O., Murphy, A. L., Picco, G. P., Ghiro, L., Molteni, D.,

Ossi, F., and Cagnacci, F. (2017). LoRa from the City

to the Mountains: Exploration of Hardware and Envi-

ronmental Factors. In Proceedings of the 2017 Inter-

national Conference on Embedded Wireless Systems

and Networks, EWSN ’17, pages 317–322,

USA. Junction Publishing.

Jolliffe, I. T. (2013). Principal Component Analysis.

Springer Science & Business Media.

Jovalekic, N., Drndarevic, V., Darby, I., Zennaro, M.,

Pietrosemoli, E., and Ricciato, F. (2018). LoRa

Transceiver With Improved Characteristics. IEEE

Wireless Communications Letters, 7(6):1058–1061.

Kerlink (2022). Web page of kerlink.

https://www.kerlink.com/product/wirnet-station/.

Pet

arvi, J., Mikhaylov, K., Pettissalo, M., Janhunen,

J., and Iinatti, J. (2017). Performance of a low-

power wide-area network based on LoRa technology:

Doppler robustness, scalability, and coverage. In-

ternational Journal of Distributed Sensor Networks,

13(3):155014771769941.

Qaraqe, M., Althunibat, S., Badarneh, O. S., and Mesleh,

R. (2020). Performance analysis of chirp spread spec-

trum system under mobility scenario. Physical Com-

munication, 43:101233.

Rf-track (2022). Rf-track, tracker 7. https://www.rf-

track.com/balises-de-geolocalisation/.

Shanmuga Sundaram, J. P., Du, W., and Zhao, Z. (2020).

A Survey on LoRa Networking: Research Problems,

Current Solutions, and Open Issues. IEEE Communi-

cations Surveys & Tutorials, 22(1):371–388.

Souza Bezerra, N.,

Ahlund, C., Saguna, S., and

de Sousa, V. A. (2019). Temperature Impact in Lo-

RaWAN—A Case Study in Northern Sweden. Sen-

sors, 19(20):4414. Number: 20 Publisher: Multidis-

ciplinary Digital Publishing Institute.

Villarim, M. R., de Luna, J. V. H., de Farias Medeiros, D.,

Pereira, R. I. S., and de Souza, C. P. (2019). LoRa

Performance Assessment in Dense Urban and Forest

Areas for Environmental Monitoring. In 2019 4th In-

ternational Symposium on Instrumentation Systems,

Circuits and Transducers (INSCIT), pages 1–5, Sao

Paulo, Brazil. IEEE.

Wiyadi, E., Setiadi, R. N., and Umar, L. (2020). Effect

of Vegetation Proﬁle and Air Data Rate on Packet

Loss Performance of LoRa E32-30dBm 433 MHz as a

Wireless Data Transmission. Journal of Physics: Con-

ference Series, 1655:012015. Publisher: IOP Publish-

ing.

Yousuf, A. M., Rochester, E. M., Ousat, B., and Ghaderi,

M. (2018). Throughput, Coverage and Scalability

of LoRa LPWAN for Internet of Things. In 2018

IEEE/ACM 26th International Symposium on Quality

of Service (IWQoS), pages 1–10, Banff, AB, Canada.

IEEE.

Zhu, G., Liao, C., Sakdejayont, T., Lai, I., Narusue, Y., and

Morikawa, H. (2019). Improving the Capacity of a

Mesh LoRa Network by Spreading-Factor-Based Net-

work Clustering. IEEE Access, 7:21584–21596. Con-

ference Name: IEEE Access.

IoTBDS 2022 - 7th International Conference on Internet of Things, Big Data and Security

280