2024 International Conference on Computing, Sciences and Communications (ICCSC)

Performance Enhancement of IoT LoRa Network

for Mobile Devices based on Improved AMILCC
Technique
Venkatesan C Thamaraimanalan T Mohammed Ibrahim M
Department of Electronics and Department of Electronics and Department of Computer Science and
Communication Engineering Communication Engineering Engineering
HKBK College of Engineering Sri Eshwar College of Engineering Sona College of Technology
Bengaluru, Karnataka, India Coimbatore, Tamilnadu, India Salem, Tamilnadu, India
[email protected] [email protected] [email protected]
2024 International Conference on Computing, Sciences and Communications (ICCSC) | 979-8-3503-5364-8/24/$31.00 ©2024 IEEE | DOI: 10.1109/ICCSC62048.2024.10830343

Vinoly G Ramkumar M Saranya K

Department of Computer Science and Department of Computer Science and Department of Computer Science and
Engineering (Cyber Security) Business Systems Engineering
K S R College of Engineering Knowledge Institute of Technology Knowledge Institute of Technology
Thiruchengode, Tamilnadu, India Salem, Tamilnadu, India Salem, Tamilnadu, India
[email protected] [email protected] [email protected]

Abstract—In Internet of Things (IoT) networks, long-range Despite its advantages, LoRa was initially designed with
(LoRa) technology is well-known for its wide communication stationary devices in mind, posing significant challenges when
range, but using it with mobile devices presents unique applied to mobile contexts. For mobile IoT devices, issues
difficulties. Frequently occurring channel attenuation, which such as frequent channel attenuation, increased transmission
causes packet loss, increased energy consumption, and longer delays, packet loss, and higher energy consumption are
transmission times, is one of these difficulties. An improved prevalent [4].
Adaptive Mobility-based IoT LoRa Clustering Communication
(AMILCC) strategy addressed to solve these problems. Recent advancements in multi-hop LoRa network designs
AMILCC partitions the network into optimal Spreading Factor have partially mitigated these issues by introducing relay
(SF) areas by using a 2D random waypoint mobility model. The nodes and clustering techniques to extend network coverage
Hybrid Adaptive Data Rate (HADR) technique, which combines beyond traditional single-hop configurations. Multi-hop
intra- and inter-SF region modifications to improve resource systems enhance network reliability and range by forwarding
allocation for mobile IoT LoRa devices, is a key component of data through intermediate nodes, which is particularly
AMILCC. The simulation findings demonstrate the advantages beneficial in environments where direct communication with
of AMILCC over conventional stationary techniques, even a central gateway is impractical [5]. Relay-based approaches,
under low mobility conditions (up to 5 m/s). In particular, such as those proposed by [6], have demonstrated
AMILCC decreases energy usage by 56.5%, improves packet
improvements in coverage and network performance, but
success ratio (PSR) by 72%, and decreases end-to-end delay by
challenges remain in effectively managing mobility and
45.32%. These outcomes demonstrate how well AMILCC works
to enhance efficiency and performance for mobile LoRa
optimizing resource allocation.
applications. Current solutions for mobile LoRa networks often depends
on Adaptive Data Rate (ADR) algorithms that adjust
Keywords— Long-range, Internet of Things, AMILCC, Packet transmission parameters based on node position and status.
Success Ratio, Spreading Factor. These algorithms aim to address dynamic network conditions
I. INTRODUCTION by optimizing spreading factors and transmission power [7].
However, these methods face limitations related to channel
The Internet of Things (IoT) has increasingly influenced attenuation, increased collision rates, frequent network
both academic and industrial fields, propelled by the surge in reconfigurations, and elevated energy consumption. For
smart devices, sensors, and connected technologies. IoT's core instance, while ADR can enhance performance in static
capability is its facilitation of continuous information flow contexts, its effectiveness diminishes when nodes are in
between users and devices, which in turn supports intelligent motion, leading to suboptimal resource allocation and
decision-making and boosts productivity. The concept of IoT increased packet loss [8].
hinges on three essential components: self-identification of
devices, communication between them, and interaction with The challenge of optimizing resource allocation for mobile
end users and other networked systems [1]. IoT devices in LoRa networks necessitates the development
of more sophisticated approaches. Recent research has
LoRa technology, developed by Semtech [2], has emerged explored various clustering strategies to address these
as a prominent Low Power Wide Area Network solution due challenges. For example, the work of [9] and [10] introduced
to its ability to operate in unlicensed spectral bands. This SF-based clustering schemes to improve multi-hop capacity
feature makes LoRa a cost-effective option for a broad range by assigning unique spreading factors to clusters of sink
of applications, including industrial equipment tracking, smart nodes. However, these approaches often struggle with
city sensor networks, and emergency monitoring systems [3]. practical issues, such as maintaining network stability and

979-8-3503-5364-8/24/$31.00 ©2024 IEEE

Authorized licensed use limited to: MEHRAN UNIV OF ENGINEERING AND TECHNOLOGY. Downloaded on January 27,2025 at 06:03:08 UTC from IEEE Xplore. Restrictions apply.
 managing the mobility of devices within higher spreading The literature highlights several key advancements and
factor clusters. persistent challenges in optimizing LoRa networks for both
static and mobile applications. While multi-hop
In summary, while LoRa technology offers a promising communication and clustering methods have improved
framework for IoT applications, its effectiveness is network performance, issues such as traffic redundancy,
constrained by challenges associated with mobile nodes. The device management, and adaptability to mobility remain
integration of relay nodes and clustering techniques has shown significant hurdles. Continued research is needed to develop
promise in enhancing network performance, but current more robust solutions that effectively address these challenges
methods often fall short in dynamic environments. To address and enhance the overall performance of LoRa networks in
these gaps, innovative approaches that leverage mobility- dynamic environments [19].
based clustering and optimize resource allocation are
essential. This paper proposes the improved AMILCC scheme LoRa devices are organized into discrete SF clusters for
as a novel solution to enhance the performance of mobile IoT LoRa network clustering, and each cluster has a Cluster
LoRa networks by addressing frequent channel attenuation, Head (CH) who oversees data flows between devices and the
packet loss, and increased energy consumption. gateway. Through efficient node management, this technique
extends the life of wireless networks and enhances network
II. RELATED WORKS connectivity. By eliminating redundant message delivery,
One of the early multi-hop LoRa network solutions, clustering reduces packet duplication, simplifies data routing,
LoRablink, utilized an IoT Time Division Multiple Access and boosts efficiency and scalability. As a result, several
(TDMA) protocol to enable reliable and energy-efficient cluster-based solutions for IoT LoRa networks have been
multi-hop communication. LoRablink's design demonstrated presented, with clustering emphasized as critical for efficient
a notable improvement in network efficiency by reducing communication among LoRa nodes. In [20] combined
collisions and optimizing data transmission, though it also clustering and forwarding relay nodes to increase LoRa
highlighted the complexity of managing multi-hop network coverage and energy efficiency at the cost of
communication in practical scenarios [11]. Similarly, the Fair increased traffic and redundancy. A different method
and Scalable Relay Control (FSRC) system was proposed to suggested assigning distinct SFs to groups of sink nodes in
enhance coverage and success probability across different order to increase multi-hop capacity. However, because it is
Spreading Factor (SF) zones. FSRC optimizes communication impractical to remove devices from higher SF clusters, this
settings to maximize coverage probability while balancing the technique has trouble maintaining the network. A different
success probability across SF zones, providing a more proposal, meanwhile, concentrated on SF Partition-based
scalable solution for multi-hop LoRa networks [12]. Clustering and Relaying (SFPCR), which limited its use in
dynamic scenarios by ignoring the mobility of IoT LoRa
In the context of network clustering, where LoRa devices devices while attempting to increase packet delivery success
are organized into SF-based clusters, various methods have across SF areas.
been proposed to improve network performance. Cluster-
based approaches, where a Cluster Head (CH) manages data III. PROPOSED HYBRID NEYWORK SYSTEM
flow between devices and gateways, have been shown to
The improved AMILCC scheme addresses the challenges
enhance connectivity and prolong network lifespan by of mobile IoT devices by utilizing a 2D random waypoint
reducing packet duplication and optimizing data routing [13]. mobility model to partition the network into strategically
Combining relay nodes with clustering strategies has further optimized SF regions, enhancing adaptability to dynamic
refined network performance by improving coverage and conditions. Mobile devices and sensors collect data—such as
energy efficiency, although these methods often face location, environmental readings, or user inputs—which is
challenges related to traffic redundancy and network transmitted to a LoRa transceiver. This transceiver relays the
sustainability [14, 15]. data to a LoRa gateway, which aggregates and forwards it to
One significant approach is the SF-based clustering the network server. The server processes and manages the data
method, which assigns unique SFs to clusters of sink nodes to flow, utilizing improved AMILCC techniques to enhance
enhance multi-hop capacity. This method aims to optimize the transmission reliability and efficiency. The system
performance of multi-hop communication by managing periodically evaluates channel conditions to adjust parameters
different SFs within designated clusters. However, practical like data rate and power, incorporating advanced methods
deployment issues, such as maintaining device management such as adaptive modulation, coding, and MIMO to optimize
and cluster stability in higher SF clusters, have been identified performance and reduce errors.
as major challenges [16]. Additionally, the SF Partition-based The 2D random waypoint mobility model is expressed as:
Clustering and Relaying (SFPCR) approach, which attempts
to improve packet delivery success across SF areas, falls short ( + 1) = ( ) + ( ) (1)
in dynamic contexts. The SFPCR method, while effective in
static environments, does not adequately address the mobility where: ( ) is the position of the device; ( ) is the
of IoT devices, which limits its applicability in dynamic velocity of the device and is the time step.
scenarios [17]. In Figure 1, the process initiates with the collection of data
Recent advancements have also explored more irrelevant from mobile devices or sensors. These devices, which can
methods for addressing these challenges. For instance, a study include anything from smartphones to specialized sensors,
by [18] introduced a hybrid adaptive data rate scheme that gather various forms of data such as location information,
dynamically adjusts transmission parameters based on environmental readings, or user inputs. This collected data is
network conditions. This approach enhances the adaptability crucial for further analysis and decision-making. Following
of LoRa networks to varying mobility scenarios but still faces the collection phase, the data is transmitted to a LoRa
limitations in addressing issues related to frequent channel transceiver. LoRa is a protocol designed to enable low-power,
attenuation and dynamic network reconfigurations. wide-area communication between devices and the internet. It
is particularly useful for applications where long-range

Authorized licensed use limited to: MEHRAN UNIV OF ENGINEERING AND TECHNOLOGY. Downloaded on January 27,2025 at 06:03:08 UTC from IEEE Xplore. Restrictions apply.
 connectivity is essential, and the devices are often battery- ensuring that the overall network operates smoothly and
operated, making energy efficiency a key consideration. efficiently. The system then utilizes Improved AMILCC
Techniques, which could stand for Advanced Machine
For a battery-operated device using LoRa, the energy Learning for Channel Coding or similar advanced methods, to
consumption is a critical factor. The energy consumed can enhance the reliability and efficiency of data transmission.
be calculated by: These techniques are crucial for adapting to varying network
= ⋅ (2) conditions and improving overall performance.
In LoRa communication, power consumption can vary The flowchart includes a decision point where the system
depending on the operation mode (transmission, reception, evaluates channel conditions. This evaluation is essential for
idle, or sleep). The average power consumption can be assessing the quality of communication channels and
approximated by: determining the most effective transmission strategies. Based
on this assessment, the system might need to adjust
= ⋅ + ⋅ + ⋅ (3) transmission parameters such as the data rate, power level, or
frequency of transmission to maintain optimal performance.
Where is the power consumption during
transmission, is the power consumption during reception, Additionally, the system could employ advanced
techniques such as adaptive modulation, coding, and MIMO
is the power consumption during sleep, , ,
to further enhance data throughput and minimize errors. These
are the durations of transmission, reception, and sleep, techniques help in adapting to changing network conditions
respectively and !"!# is the total cycle time. and optimizing the efficiency of data transmission. The
Once the data reaches the LoRa transceiver, the process concludes with the aggregation of data, continuous
transceiver’s role is to act as a bridge between the mobile monitoring of system performance, and detailed reporting.
devices and the larger network infrastructure. The transceiver This final stage ensures that the communication system
captures the data transmitted from the devices and then operates effectively, any issues are promptly identified and
forwards it to the next stage in the communication system. addressed, and the system’s performance is continuously
This stage is crucial because the data is effectively handed off evaluated to maintain high standards of operation.
to the broader network without loss or corruption.
IV. RESULTS AND DISCUSSION
The data is then sent from the LoRa transceiver to a LoRa
gateway. The LoRa gateway serves as a crucial intermediary, The performance of various communication techniques is
collecting data from multiple transceivers within its range and rigorously evaluated to determine their effectiveness in
forwarding this information to the network server. This optimizing LoRa networks, particularly for mobile IoT
gateway plays a vital role in aggregating data streams from devices. The comparative analysis encompasses four key
various sources and ensuring that they are directed correctly techniques: Traditional Channel Coding, Adaptive
within the network. Once the data reaches the network server, Modulation and Coding (AMC), MIMO, and the proposed
it becomes the central node where the data is processed, Improved AMILCC Techniques. These methods were
stored, and analyzed. This server is important for managing assessed based on several critical parameters, including
the data flow and ensuring that the information gathered from transmission reliability, energy efficiency, adaptability to
various devices is used effectively. channel conditions, latency, data rate, packet success ratio,
energy consumption per packet, and end-to-end delay. The
aim is to identify the most effective technique for enhancing
network performance, addressing challenges such as packet
loss, energy consumption, and transmission delays that are
prevalent in mobile IoT applications.

TABLE I. PERFORMANCE COMPARISON OF IMPROVED AMILCC

WITH OTHER CONVENTIONAL TECHNIQUES

Traditional Improved
Parameter Channel AMC MIMO AMILCC
Coding Technique
Transmission
Reliability 88 93 95 99
(%)
Energy
Efficiency 0.7 0.5 0.5 0.3
(mW/bit)
Adaptibility
to Channel 3/10 6/10 7/10 8/10
Conditions
Latency (ms) 15 23 25 28
Data Rate
1 1.5 1.8 2
(Mbps)

Fig. 1. Process flow of the improved AMILCC Technique. A comparative overview of various communication
techniques based on key performance parameters is provided
With the data now at the network server, the system in Table 1. Traditional Channel Coding offers reliable
proceeds to process the data and manage network resources. transmission at 88% but has moderate energy efficiency (0.7
This involves several tasks, including analyzing the incoming mW/bit) and adaptability (3/10), with a latency of 15
data, optimizing the allocation of network resources, and milliseconds and a data rate of 1 Mbps. The AMC shows

Authorized licensed use limited to: MEHRAN UNIV OF ENGINEERING AND TECHNOLOGY. Downloaded on January 27,2025 at 06:03:08 UTC from IEEE Xplore. Restrictions apply.
 improved transmission reliability at 93%, better energy
efficiency (0.5 mW/bit), and greater adaptability (6/10),
although it has a higher latency of 23 milliseconds and a data
rate of 1.5 Mbps. The MIMO enhances reliability to 95%,
maintains energy efficiency at 0.5 mW/bit, and offers better
adaptability (7/10), but with a slightly higher latency of 25
milliseconds and a data rate of 1.8 Mbps. Improved AMILCC
Techniques achieve the highest transmission reliability at
99%, the best energy efficiency (0.3 mW/bit), and top
adaptability (8/10), although it has the highest latency of 28
milliseconds and the highest data rate of 2 Mbps.

TABLE II. COMPARISON OF COMMUNICATION TECHNIQUES WITH THE

PROPOSED AMILCC TECHNIQUE

End-to-
Packet Energy
End
Technique Success Consumption (per
Delay
Ratio packet)
(ms)
Traditional
(b)
87% 1.2 mJ 75
Channel Coding
AMC 93% 0.85 mJ 35

MIMO 97% 1.8 mJ 20

Improved
AMILCC 99% 0.7 mJ 10
Techniques

Table 2 presents a comparative analysis of four

communication techniques—Traditional Channel Coding,
Adaptive Modulation and Coding (AMC), MIMO, and
Improved AMILCC Techniques—across three critical
performance metrics: Packet Success Ratio, Energy
Consumption, and End-to-End Delay. Traditional Channel
Coding achieves a packet success ratio of 87% with an energy
consumption of 1.2 mJ per packet and an end-to-end delay of
75 milliseconds, indicating moderate performance in
reliability and efficiency. The AMC improves the packet
success ratio to 93%, reduces energy consumption to 0.85 mJ (c)
per packet, and decreases end-to-end delay to 35 milliseconds,
demonstrating better adaptability and efficiency. Figure 2
gives the comparison of different techniques with the
improved AMILCC method.

(d)
Fig. 2. Comparison of Different Techniques with the improved AMILCC
Method: (a) Transmission Reliability, (b) Energy Efficiency, (c)
Adaptability to Channel Conditions, and (d) Latency
(a)
The MIMO further enhances performance with a packet
success ratio of 97%, higher energy consumption of 1.8 mJ
per packet, but a significantly reduced end-to-end delay of 20
milliseconds, reflecting its effectiveness in high-throughput
scenarios. The Improved AMILCC Techniques achieve the
highest packet success ratio of 99%, the lowest energy
consumption of 0.7 mJ per packet, and an end-to-end delay of

Authorized licensed use limited to: MEHRAN UNIV OF ENGINEERING AND TECHNOLOGY. Downloaded on January 27,2025 at 06:03:08 UTC from IEEE Xplore. Restrictions apply.
 just 10 milliseconds, showcasing superior performance in all routing using ASFO and a cross-layer-based expedient routing protocol
aspects due to advanced optimization and adaptation methods. for wireless sensor networks," Sensors, vol. 23, no. 5, p. 2788, 2023
[7] C. Venkatesan, T. Thamaraimanalan, M. Ramkumar, A.
The results reveal that the Improved AMILCC Techniques Sivaramakrishnan, and M. Marimuthu, "ECG signal feature extraction
outperform the other methods across most performance and SVM classifier based cardiac arrhythmia detection," in Proc. 2023
metrics. It achieved the highest transmission reliability (99%) 2nd Int. Conf. Electr., Electron., Inf. Commun. Technol. (ICEEICT),
and the best energy efficiency (0.3 mW/bit), despite having Apr. 2023, pp. 1–4.
the highest latency and data rate among the evaluated [8] M. Niazi and A. Hussain, "Adaptive data rate mechanisms in
techniques. Notably, the Improved AMILCC Techniques also LoRaWAN: A survey and future directions," IEEE Internet Things J.,
exhibited superior performance in packet success ratio (99%), vol. 8, no. 7, pp. 5329–5343, 2021. doi: 10.1109/JIOT.2021.3067411.
minimal energy consumption (0.7 mJ per packet), and the [9] V. V. Teresa, T. Thamaraimanalan, S. Jeevanantham, A. Balaji, H.
shortest end-to-end delay (10 milliseconds). These Gokulanand, and N. Jeevanprasath, "LoRa enhanced real-time
improvements highlight the effectiveness of AMILCC in earthquake monitoring system for constructions using IoT and cloud,"
addressing the unique challenges associated with mobile in Proc. 2023 7th Int. Conf. Electron., Commun. Aerosp. Technol.
LoRa devices, thereby demonstrating its potential for (ICECA), Coimbatore, India, 2023, pp. 1410–1415. doi:
10.1109/ICECA58529.2023.10395485.
optimizing connectivity and data transmission in dynamic IoT
environments. [10] M. Ali, H. Zhang, and Q. Wang, "Enhancing LoRa network performance
with SF-based clustering and multi-hop communication," IEEE Trans.
V. CONCLUSION Wireless Commun., vol. 21, no. 5, pp. 2905–2918, 2022. doi:
10.1109/TWC.2022.3144876.
The distribution of LoRa technology for mobile IoT
devices has been made easier with the introduction of the [11] E. Hossain, Z. M. Fadlullah, and M. S. Hossain, "A comprehensive
survey on the integration of LoRa and edge computing for smart city
improved AMILCC scheme. By dividing the network into applications," IEEE Access, vol. 10, pp. 54523–54541, 2022. doi:
optimal SF areas using a 2D random waypoint mobility 10.1109/ACCESS.2022.3194537.
model, AMILCC solves problems such frequent channel
attenuation, packet loss, higher energy consumption, and [12] M. Kang, Y. Liu, and X. Zhang, "Fair and scalable relay control (FSRC)
in multi-hop LoRa networks," IEEE Trans. Wireless Commun., vol. 20,
extended transmission times. Performance is greatly improved no. 1, pp. 341–354, 2021. doi: 10.1109/TWC.2020.3034352.
by its HADR system, which modifies both inside and between
SF zones. According to simulation studies, AMILCC reduces [13] Y. Li, S. Zhang, and Y. Yang, "Efficient clustering and routing for LoRa
networks with dynamic node deployment," Ad Hoc Netw., vol. 112, p.
energy usage by 56.5%, increases Packet Success Ratio (PSR) 102337, 2022. doi: 10.1016/j.adhoc.2020.102337.
by over 72%, and reduces end-to-end latency by 45.32%.
Compared to traditional methods, as well as AMC and MIMO [14] C. Venkatesan, Y. D. Zhang, and Q. Xin, "Preface to Special Issue on
techniques, AMILCC offers superior performance with the Digital Twin Empowered Internet of Intelligent Things for Engineering
Cyber-Physical Human Systems," Comput. Assist. Methods Eng. Sci.,
highest reliability (99%), best energy efficiency (0.3 mW/bit), vol. 30, no. 3, pp. 259–262, 2023.
and lowest latency (10 ms). This demonstrates AMILCC's
effectiveness in optimizing LoRa networks for mobile [15] J. Suri, B. Chandrasekhar, and S. Narasimhan, "Enhancing LoRa
network performance through relay node clustering," Comput. Netw.,
scenarios and its potential to enhance connectivity and data vol. 211, p. 109191, 2023. doi: 10.1016/j.comnet.2022.109191
transmission in dynamic IoT environment.
[16] L. Xie, L. Zhang, and Z. Jiang, "Spreading factor-based clustering for
REFERENCES multi-hop LoRa networks," IEEE Internet Things J., vol. 9, no. 4, pp.
2867–2880, 2022. doi: 10.1109/JIOT.2021.3078992.
[1] L. Atzori, A. Iera, and G. Morabito, "The Internet of Things: A survey,"
Comput. Netw., vol. 54, no. 15, pp. 2787–2805, 2010. doi: [17] J. Zhang, X. Liu, and M. Hu, "SF partition-based clustering and relaying
10.1016/j.comnet.2010.05.010. for enhanced packet delivery in LoRa networks," Sensors, vol. 21, no.
15, p. 5106, 2021. doi: 10.3390/s21155106
[2] Semtech, "LoRa® Technology Overview," 2023. [Online]. Available:
https://www.semtech.com/products/wireless-rf/lora. [18] L. Qiao, C. Lin, and H. Wang, "Hybrid adaptive data rate for mobile IoT
LoRa networks," J. Netw. Comput. Appl., vol. 191, p. 103241, 2023. doi:
[3] LoRa Alliance, "LoRaWAN® 1.1 Specification," 2023. [Online]. 10.1016/j.jnca.2021.103241.
Available: https://lora-alliance.org/lorawan-specification/
[19] T. Thamaraimanalan, C. Venkatesan, M. Ramkumar, A.
[4] S. Kumar, A. Mishra, and G. Dhiman, "Performance analysis of Sivaramakrishnan, and M. Marimuthu, "ANFIS-based multilayered
LoRaWAN for IoT applications: A review," Wireless Netw., vol. 26, no. algorithm for botnet detection," in Proc. 2023 Int. Conf. Recent Adv.
4, pp. 2871–2888, 2020. doi: 10.1007/s11276-019-02109-0 Electr., Electron., Ubiquitous Commun., Comput. Intell. (RAEEUCCI),
[5] M. S. Sivagamasundari, T. Thamaraimanalan, S. Ramalingam, and K. Apr. 2023, pp. 1–5.
Balachander, "Improved particle swarm optimization based distributed [20] C. Venkatesan, A. T. Fadi, and D. Pelusi, "Guest Editorial: Smart
energy-efficient opportunistic algorithm for clustering and routing in measurement in machine vision for challenging applications," IEEE
WSNs," J. Inf. Technol. Manag., pp. 5–20, 2023. Instrum. Meas. Mag., vol. 26, no. 8, pp. 3–3, 2023.
[6] V. Cherappa, T. Thangarajan, S. S. Meenakshi Sundaram, F. Hajjej, A.
K. Munusamy, and R. Shanmugam, "Energy-efficient clustering and

Authorized licensed use limited to: MEHRAN UNIV OF ENGINEERING AND TECHNOLOGY. Downloaded on January 27,2025 at 06:03:08 UTC from IEEE Xplore. Restrictions apply.