=Paper=
{{Paper
|id=Vol-3092/p08
|storemode=property
|title=LoRa, Edge Computing and Blockchain Improving the IoT
World
|pdfUrl=https://ceur-ws.org/Vol-3092/p08.pdf
|volume=Vol-3092
|authors=Lorenzo Felli,Alessandro Vizzarri
|dblpUrl=https://dblp.org/rec/conf/system/FelliV21
}}
==LoRa, Edge Computing and Blockchain Improving the IoT
World==
https://ceur-ws.org/Vol-3092/p08.pdf
LoRa, Edge Computing and Blockchain Improving the IoT
World
Lorenzo Fellia , Alessandro Vizzarrib
a
Department of Engineering Science, Guglielmo Marconi University, Rome, Italy
b
Department of Enterprise Engineering, University of Rome Tor Vergata, Rome, Italy
Abstract
Long Range (LoRa) is a technology widely used to build Low-Power Wide-Area Networks (LPWANs) that enables bidirec-
tional transmission of data packets over long distances. Most suitable for Internet of Things (IoTs) applications, it enables the
broad-range communication and enables transmission of small data packets over long distances. This technology is perfect
for covering remote areas where other data networks are not available. The small amount of data exchanged can be limiting
if complex processing is required to understand specific events happening around the devices. An image elaboration or
video processing is an example. As of today this tasks are mostly available in the cloud. Recent developments in the field of
artificial intelligence hardware have led to the emergence of a new generation of low-power hardware capable of carrying
out inference on pre-trained models with good performance on the edge for a few watts. In this paper we investigate the
use of LoRa technology in combination with Edge Artificial Intelligence computing and Blockchain to build an architecture
enabling vehicle surveillance and control of large and remote areas.
Keywords
LoRa, AI, Surveillance
1. Introduction cult to solve. Imagine having to cover vast mountainous
or wooded areas, or in any case areas that are difficult
Over the past few years, interest in Low Power Wide to reach and are poorly or not at all covered by other
Area Area (LPWA) technologies has grown so much that types of networks [12, 13]. The aim of this work is to
it has gained unprecedented momentum and strong com- apply state-of-the-art computer and artificial intelligence
mercial interest especially in the field of the Internet of tools to optimise a remote surveillance system capable
Things (IoT) [1, 2, 3, 4]. Many candidates appeared on of understanding what is happening around the sensors
the LPWA scene (SigFox [5], LoRa [6], Wheighless [7], by processing the information on site, transmitting with
Ingenu [8]). In this paper we decided to use LoRa (Long a few data the occurrence of an event and allowing an
Range), one of the most promising wide-area LPWA tech- eventual supervisor to inspect the data once the alarm
nologies proposed by Semtech and subsequently pro- point is reached.
moted by the LoRa Alliance [9]. The ability to accommo-
date multiple users in the same channel through spread
spectrum multiple access techniques allows this technol- 2. Long Range (LoRa)
ogy to establish communication channels with low power
consumption and low cost design. The LoRa Alliance has Long Range (LoRa [6]) is a proprietary spread spectrum
defined the upper layers and network architecture above modulation technique by Semtech, derived from Chirp
the LoRa physical layers and called them LoRaWAN [9]. Spread Spectrum (CSS). Instead of modulating the mes-
Together, these features make LoRa attractive to devel- sage on a pseudorandom binary sequence, as is done in
opers who can build complete system solutions on top the well known Direct-Sequence Spread Spectrum (DSSS),
of it for both geographic [10] and residential/industrial LoRa uses a sweep tone that increases (upchirp) or de-
types of IoT networks [11], thus accelerating its market creases (downchirp) in frequency over time to encode
adoption. The ability to cover large areas with a low num- the message. Spreading the signal over a wide bandwidth
ber of devices compared to other technologies makes it makes it less susceptible to noise and interference. CSS
possible to approach problems that were previously diffi- in particular is resistant to Doppler effects [14] (com-
mon in mobile applications) and multipath fading [15].
SYSTEM 2021 @ Scholar’s Yearly Symposium of Technology, A LoRa receiver can decode transmissions 20 dB below
Engineering and Mathematics. July 27–29, 2021, Catania, IT the noise floor [16], making very long communication
" lorenzo.felli@isprambiente.it (L. Felli); distances possible, while operating at a very low power.
alessandro.vizzarri@uniroma2.it (A. Vizzarri) LoRa transceivers available today can operate between
0000-0001-8338-7957 (L. Felli); 0000-0002-6274-991X7
(A. Vizzarri) 137MHz to 1020MHz, and therefore can also operate in
© 2021 Copyright for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
licensed bands. However, they are often deployed in
CEUR
CEUR Workshop Proceedings (CEUR-WS.org)
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073 ISM bands (EU: 868MHz and 433MHz, USA: 915MHz
48
Lorenzo Felli et al. CEUR Workshop Proceedings 48–52
and 433MHz). The LoRa physical layer may be used cal sensors (such as temperature, lights, speakers), and
with any MAC layer; however, Long Range Wide Area moving computing power closer to these sensors in the
Network (LoRaWAN) is the currently proposed MAC. physical world makes sense. With collection and pro-
LoRaWAN operates in a simple star topology. A LoRa cessing power now available on the edge, companies can
transceiver has five runtime-adjustable transmission pa- significantly reduce the volumes of data that must be
rameters: Transmission Power (TP), Carrier Frequency moved and stored in the cloud, saving themselves time
(CF), Spreading Factor (SF), Bandwidth (BW), and Coding and money in the process. Image recognition and video
Rate (CR). These parameters have an influence on the streaming are just the tip of the iceberg. Security camera
transmission duration, energy consumption, robustness companies, for example, struggle to use cloud-based solu-
and range [17]. Transmission Power (TP). TP on a LoRa tions because real-time data and video streaming to the
receiver can be adjusted between -4 dBm and 20 dBm in cloud is prohibitively expensive [19]. Autonomous cars
1 dB steps. Because of regulatory and hardware limita- need offline functionality on the road [20], while AR/VR
tions, however, this is often limited between 2 dBm and gaming companies maintain their brand credibility by
14 dBm. TP has a direct influence on energy consumption keeping their products resilient to lag using 5G networks
and the range of the signal. Carrier Frequency (CF). CF or other technologies [21].
is the centre frequency, which can be programmed in
steps of 61Hz between 137MHz to 1020MHz. Spreading
Factor (SF). SF determines how many bits are encoded 4. Blockchain
in each symbol, and can be set between 6 and 12. A
Blockchain technology was introduced by a single en-
higher spreading factor increases the Signal to Noise Ra-
tity or group under the name of Satoshi Nakamoto in
tio (SNR) and therefore receiver sensitivity and range
2008 and the code of its implementation was published
of the signal. However, it lowers the transmission rate
a year later in 2009 in the document ‘Bitcoin: A Peer-
and thus increases the transmission duration and energy
to-Peer Electronic Cash System’ [22]. The Blockchain
consumption. The SFs in LoRa are orthogonal. Conse-
is essentially a distributed and transactional database
quently, concurrent transmissions with different SF do
shared by the various nodes of the network. The validity
not interfere with each other, and can be successfully de-
and integrity of the data is maintained by chaining the
coded (assuming a receiver with multiple receive paths).
transactions contained in the blocks using hash functions
Bandwidth (BW). BW can be set from (a fairly narrow)
that prevent them from being modified without consent.
7.8 kHz up to 500 kHz. In a typical LoRa deployment,
Bitcoin uses the public key infrastructure (PKI) mech-
only 125 kHz, 250 kHz and 500 kHz are considered. A
anism [23]. In PKI, the user has a couple formed by a
wider bandwidth means a more spread-out and therefore
public and a private key. The public key is used as the
more interference-resilient link. In addition, it increases
address of the user’s wallet, while the private key is used
the data rate, as the chips are sent out at a rate equivalent
to sign transactions. A block is accepted by the network
to the bandwidth. The downside of a higher bandwidth
on average every 10 minutes through a consensus mech-
is a less sensitive reception, caused by the integration
anism. The new chain with the new block on top will
of additional noise. Coding Rate (CR). CR is the amount
spread quickly in all the nodes of the network.
of Forward Error Correction (FEC) that is applied to the
Inside each node there is a key-value database in which
message to protect it against burst interference. Higher
the blocks containing the transactions that have reached
CR makes the message longer and therefore increases the
consensus will be written. Each node validates the new
time on air. LoRa transceivers with different CR, and op-
blocks. Although the search for the hash that satisfies the
erating in ‘explicit header mode’, can still communicate
consensus called Proof of Work takes on average 10 min-
with each other, as the CR is encoded in the header.
utes regardless of the network’s computational capacity,
checking the correctness of these hashes is extremely fast.
3. Edge computing This method creates a linear chain of blocks on which all
nodes agree (Figure 1). This chain of blocks is the public
Edge computing refers to applications, services, and pro- ledger technique of Bitcoin, called Blockchain.
cessing performed outside of a central data center and
closer to end users [18]. The definition of “closer” falls
along a spectrum and depends highly on networking 5. Application Scenarios
technologies used, the application characteristics, and
In this section, we discuss several recent IoT applications
the desired end user experience. While edge applications
based on LoRa and LPWAN and analyze possible areas
do not need to communicate with the cloud, they may
for improvement. The studies will be broken down by
still interact with servers and internet based applications.
application type.
Many of the most common edge devices feature physi-
49
Lorenzo Felli et al. CEUR Workshop Proceedings 48–52
Figure 1: Blockchain structure. (Zheng et al. 2016 [24])
5.1. Environmental Monitoring
The study and monitoring of the environment are essen-
tial for the preservation and prevention of all natural
environments. Within precision agriculture (PA) prac-
tices, proximity data collection is often accomplished by
IoT devices that act as collectors of environmental infor-
mation. Many studies are present about environmental
monitoring platforms realized by exploiting LORA tech-
nology. Ali et al. [25] built an environmental monitoring
platform by combining the features of LORA and ZigBee.
Leveraging the best wireless features of ZigBee and LoRa, Figure 2: Blockchain, IoT and LoRA main use cases
they proposed a diversified communication system with
smart features in a single loT system. Wang et al. [26]
presented a wireless sensor network system dedicated for
meteorological monitoring based on Long Range (LoRa) with a low-bandwith but long-range technology such
for real-time services such as agricultural production de- as LoRa. The results show that correct parameters al-
mand. Very interesting the study conducted by Chen et low large urban areas to be covered while keeping the
al. to monitor air quality by using a UAV (Unmanned transmission time low enough to keep packet losses at
Aerial Vehicle) and LoRa communication system [27]. satisfactory levels using LoRa. Interesting continous real-
Authors in [28] proposed a system for automated oil spill time testbed was conducted by Loriot et al. [36] analyzing
detection by remote sensing. Nordin et al. implemented a the different receptions within an urban context by mak-
narrowband IoT-based hydrological monitoring system in ing connectivity maps and following the measurements
a lagoon environment [29]. The authors studied network through a custom made Web application for realtime data
performance predictability, limitations, and reliability by visualization.
comparing 2G and LoRa systems. They concluded that
GSM-based data communication is unreliable in rural 5.3. Health monitoring
areas due to uneven terrain and non-line-of-sight of view
A new emerging technology called WHDs (Wearable
and concluded that LoRa is a better alternative in terms
Health Devices) has gained traction. This technology
of RSSI as long as the antenna is placed at high altitude.
enables continuous health monitoring of human vital
signs in everyday life, even 24h per day. Clinical environ-
5.2. Smart City ments can also benefit from the innovative advantages
The smart city concept has become definitively consol- of this technology minimizing interference and discom-
idated with the emergence of IoT devices. These two fort with patients’ daily lives. A low-cost LoRa health
areas have now become closely related concepts. Many monitoring system was proposed by Lousado et al. [37].
fields of urban monitoring can be covered by IoT and The system is able to track different environment data to
LORA technology. Del Campo et al. [30] proposed the monitoring the health conditions of the elderly in their
LoRa technology to monitor the power distribution on homes using LoRa technology and The Things Network
suburban area. Many studies address public lighting in cloud framework. The LoRa node was developed using
a smart city context [31, 32, 33, 34]. One of the most in- an ESP32/LoRa microcontroller and collect various envi-
teresting is presented by Pasolini et al [35]. They studied ronmental sensors data on temperature, humidity, carbon
how to implement smart street lighting by comparing monoxide, gas, and smoke. The results seem encouraging,
IEEE 802.15.4 short-range communication technology succeeding in closed places to reach a communication
range of 1.2 km. Dimitrievski et al. [38] address the dig-
50
Lorenzo Felli et al. CEUR Workshop Proceedings 48–52
ital device issue in rural areas that could create further geometry approach, in: 2018 IEEE Global Commu-
disparity in the use of smart helth devices using a LoRa- nications Conference (GLOBECOM), IEEE, 2018, pp.
IoT oriented architecture. However the constraints of 206–212.
low network bandwidth and the need to reduce the active [5] J. C. Zuniga, B. Ponsard, Sigfox system description,
time of the IoT nodes limit the capabilities of the overall LPWAN@ IETF97, Nov. 14th 25 (2016).
system. [6] J. M. Marais, R. Malekian, A. M. Abu-Mahfouz, Lora
and lorawan testbeds: A review, in: 2017 Ieee
5.4. Securing LoRa comunications Africon, IEEE, 2017, pp. 1496–1501.
[7] R. A. Abbas, A. Al-Sherbaz, A. Bennecer, P. Pic-
Security and resistance to data tampering became critical ton, A new channel selection algorithm for the
issues during the development of the LoRa system for weightless-n frequency hopping with lower colli-
the Internet of Things. The LoRa system is a centralized sion probability, in: 2017 8th International Con-
system, where data is stored in the central cloud. This ference on the Network of the Future (NOF), IEEE,
approach makes the system vulnerable to security risks 2017, pp. 171–175.
such as data forgery or loss. To solve this problem, many [8] U. Raza, P. Kulkarni, M. Sooriyabandara, Low power
researchers are using various types of Blockchain to in- wide area networks: A survey, IEEE Commun. Surv.
crease the level of data consistency and achieve a more se- Tutorials 19 (2017).
cure long-range communication system. LU et al. [39] in- [9] N. Sornin, M. Luis, T. Eirich, T. Kramp, O. Hersent,
troduced HyperLoRa, a blockchain-enabled LoRa system Lorawan specification, LoRa alliance (2015).
by using the open-source Hyperledge Fabric blockchain. [10] G. M. Bianco, R. Giuliano, G. Marrocco, F. Mazzenga,
Ozyılmaz et al. [40] have investigated the possibility of A. Mejia-Aguilar, Lora system for search and res-
using a decentralized storage called SWARM for secure cue: Path-loss models and procedures in mountain
data storage, using Ethereum’s blockchain infrastruc- scenarios, IEEE Internet of Things Journal 8 (2020)
ture. Unfortunately, the system throughput is low due 1985–1999.
to the choice of consensus type. Using in the Ethereum [11] S. Sağır, İ. Kaya, C. Şişman, Y. Baltacı, S. Ünal, Eval-
private network the PoS instead of the PoW the num- uation of low-power long distance radio communi-
ber of transactions per second would have been much cation in urban areas: Lora and impact of spreading
greater [41, 42]. factor, in: 2019 Seventh International Conference
on Digital Information Processing and Communi-
5.5. Conclusion cations (ICDIPC), IEEE, 2019, pp. 68–71.
[12] R. Giuliano, F. Mazzenga, A. Vizzarri, Satellite-
This paper provides a brief introduction to Blockchain, based capillary 5G-mMTC networks for environ-
Edge Computing and LoRa technologies and how they mental applications, IEEE Aerospace and Electronic
can together be used to improve both performance and Systems Magazine 34 (2019) 40–48.
security of IoT devices. As a result of this combination of [13] F. Bonanno, G. Capizzi, G. Lo Sciuto, C. Napoli,
technologies, the number of use cases and real-world ap- Wavelet recurrent neural network with semi-
plications will undoubtedly tend to increase, bringing an parametric input data preprocessing for micro-wind
ever-increasing level of maturity to the IoT environment. power forecasting in integrated generation sys-
tems, 2015, pp. 602–609. doi:10.1109/ICCEP.
2015.7177554.
References [14] J. Petäjäjärvi, K. Mikhaylov, M. Pettissalo, J. Jan-
[1] R. S. Sinha, Y. Wei, S.-H. Hwang, A survey on lpwa hunen, J. Iinatti, Performance of a low-power wide-
technology: Lora and nb-iot, Ict Express 3 (2017) area network based on lora technology: Doppler
14–21. robustness, scalability, and coverage, International
[2] F. Muteba, K. Djouani, T. Olwal, A comparative sur- Journal of Distributed Sensor Networks 13 (2017)
vey study on lpwa iot technologies: Design, consid- 1550147717699412.
erations, challenges and solutions, Procedia Com- [15] U. Noreen, A. Bounceur, L. Clavier, A study of lora
puter Science 155 (2019) 636–641. low power and wide area network technology, in:
[3] R. Avanzato, F. Beritelli, M. Russo, S. Russo, M. Vac- 2017 International Conference on Advanced Tech-
caro, Yolov3-based mask and face recognition al- nologies for Signal and Image Processing (ATSIP),
gorithm for individual protection applications, in: IEEE, 2017, pp. 1–6.
CEUR Workshop Proc., 2020, pp. 41–45. [16] R. Ghanaatian, O. Afisiadis, M. Cotting, A. Burg,
[4] A. Azari, C. Cavdar, Performance evaluation and Lora digital receiver analysis and implementation,
optimization of lpwa iot networks: A stochastic in: ICASSP 2019-2019 IEEE International Confer-
ence on Acoustics, Speech and Signal Processing
51
Lorenzo Felli et al. CEUR Workshop Proceedings 48–52
(ICASSP), IEEE, 2019, pp. 1498–1502. A. Santamaria, Power distribution monitoring us-
[17] G. Capizzi, F. Bonanno, C. Napoli, Hybrid neural ing lora: Coverage analysis in suburban areas., in:
networks architectures for soc and voltage predic- EWSN, 2018, pp. 233–238.
tion of new generation batteries storage, in: 2011 [31] N. Khatavkar, A. Naik, B. Kadam, Energy efficient
International Conference on Clean Electrical Power street light controller for smart cities, in: 2017 Int.
(ICCEP), IEEE, 2011, pp. 341–344. conf. Microelectronic Devices, Circuits and Systems
[18] W. Shi, J. Cao, Q. Zhang, Y. Li, L. Xu, Edge comput- (ICMDCS), IEEE, 2017, pp. 1–6.
ing: Vision and challenges, IEEE internet of things [32] F. Bonanno, G. Capizzi, A. Gagliano, C. Napoli, Op-
journal 3 (2016) 637–646. timal management of various renewable energy
[19] G. Ananthanarayanan, et al., Real-time video ana- sources by a new forecasting method, in: Inter-
lytics: The killer app for edge computing, computer national Symposium on Power Electronics Power
50 (2017) 58–67. Electronics, Electrical Drives, Automation and Mo-
[20] S. Liu, L. Liu, J. Tang, B. Yu, Y. Wang, W. Shi, Edge tion, IEEE, 2012, pp. 934–940.
computing for autonomous driving: Opportunities [33] A. Murthy, D. Han, D. Jiang, T. Oliveira, Lighting-
and challenges, Proceedings of the IEEE 107 (2019) enabled smart city applications and ecosystems
1697–1716. based on the iot, in: IEEE 2nd World Forum on
[21] B. Krogfoss, J. Duran, P. Perez, J. Bouwen, Quan- Internet of Things (WF-IoT), 2015, pp. 757–763.
tifying the value of 5g and edge cloud on qoe for [34] R. Fujdiak, P. Mlynek, J. Misurec, J. Slacik, Sim-
ar/vr, in: 2020 Twelfth International Conference on ulation of intelligent public light system in smart
Quality of Multimedia Experience (QoMEX), IEEE, city, in: 2017 Progress In Electromagnetics Research
2020, pp. 1–4. Symposium-Spring (PIERS), 2017, pp. 2515–2519.
[22] S. Nakamoto, Bitcoin: A peer-to-peer electronic [35] G. Pasolini, et al., Smart city pilot projects using lora
cash system, Decentralized Business Review (2008) and ieee802. 15.4 technologies, Sensors 18 (2018)
21260. 1118.
[23] R. Housley, Public key infrastructure (PKI), vol- [36] M. Loriot, A. Aljer, I. Shahrour, Analysis of the use
ume 3, John Wiley & Sons, Inc.: Hoboken, NJ, USA, of lorawan technology in a large-scale smart city
2004. demonstrator, in: 2017 Sensors Networks Smart
[24] Z. Zheng, S. Xie, H.-N. Dai, X. Chen, H. Wang, and Emerging Technologies (SENSET), IEEE, 2017,
Blockchain challenges and opportunities: A sur- pp. 1–4.
vey, International Journal of Web and Grid Services [37] J. P. Lousado, S. Antunes, e-health monitoring sys-
14 (2018) 352–375. tem for senior citizens based on lora technology,
[25] A. I. Ali, S. Z. Partal, S. Kepke, H. P. Partal, Zig- in: 2020 5th International Conference on Smart and
bee and lora based wireless sensors for smart envi- Sustainable Technologies (SpliTech), IEEE, 2020, pp.
ronment and iot applications, in: 2019 1st Global 1–5.
Power, Energy and Communication Conference [38] A. Dimitrievski, et al., Rural healthcare iot architec-
(GPECOM), IEEE, 2019, pp. 19–23. ture based on low-energy lora, International Jour-
[26] D. Wang, H. Hu, H. Yang, K. Zheng, A real-time nal of Environmental Research and Public Health
lora-based iot system for intelligent meteorological 18 (2021) 7660.
monitoring, in: 2020 IEEE 6th International Con- [39] L. Hou, K. Zheng, Z. Liu, X. Xu, T. Wu, Design and
ference on Computer and Communications (ICCC), prototype implementation of a blockchain-enabled
IEEE, 2020, pp. 921–925. lora system with edge computing, IEEE Internet of
[27] L.-Y. Chen, H.-S. Huang, C.-J. Wu, Y.-T. Tsai, Y.-S. Things Journal 8 (2020) 2419–2430.
Chang, A lora-based air quality monitor on un- [40] K. R. Ozyilmaz, A. Yurdakul, Designing a
manned aerial vehicle for smart city, in: 2018 Inter- blockchain-based iot with ethereum, swarm, and
national Conference on System Science and Engi- lora: the software solution to create high availabil-
neering (ICSSE), IEEE, 2018, pp. 1–5. ity with minimal security risks, IEEE Consumer
[28] G. Capizzi, G. Lo Sciuto, M. Woźniak, R. Damaše- Electronics Magazine 8 (2019) 28–34.
vicius, A clustering based system for automated [41] B. Cao, Z. Zhang, D. Feng, S. Zhang, L. Zhang,
oil spill detection by satellite remote sensing, in: M. Peng, Y. Li, Performance analysis and compar-
International Conference on Artificial Intelligence ison of pow, pos and dag based blockchains, Dig.
and Soft Computing, Springer, 2016, pp. 613–623. Com. Netw. 6 (2020) 480–485.
[29] H. M. Jawad, et al., Energy-efficient wireless sen- [42] B. Sriman, S. G. Kumar, P. Shamili, Blockchain
sor networks for precision agriculture: A review, technology: Consensus protocol proof of work and
Sensors 17 (2017). proof of stake, in: Int. Comp. App., Springer, 2021,
[30] G. Del Campo, I. Gomez, S. C. Sierra, R. Martinez, pp. 395–406.
52