ScienceDirect
Available online at www.sciencedirect.com
Procedia Computer Science 265 (2025) 49–56
1877-0509 © 2025 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer review under the responsibility of the scientific program committees
10.1016/j.procs.2025.07.155
Keywords: Energy Efficiency; Distributed Computing; Protocols; Edge Devices; Optimization; Swarm Robotics; Resource Management.
1. Introduction
The Internet of Things (IoT) has revolutionized various domains, including smart cities, healthcare, industrial au-
tomation, and environmental monitoring. However, the widespread deployment of IoT devices presents significant
challenges in terms of energy efficiency, particularly for battery-operated and energy-efficient nodes. Traditional IoT
devices rely on batteries or wired power sources, limiting their operational lifespan and scalability. Energy-efficient
IoT devices offer a sustainable solution to this challenge. Achieving reliable and efficient communication while main-
taining energy efficiency remains a substantial research challenge.
The IoT is set for substantial transformation owing to the advent of energy-efficient communication technologies.
The shift to edge computing signifies a major trend that will impact future interactions among IoT devices. Edge com-
∗ Corresponding author: Amir Ijaz
E-mail address: amir.ijaz@utu.fi
The 20th International Conference on Future Networks and Communications (FNC)
August 4-6, 2025, Leuven, Belgium
Towards Optimizing Communication Cost in Energy Efficient IoT
Devices for Swarm Robotics
Amir Ijaza,∗, Hashem Haghbayana, Abdul Malika, Ethiopia Nigussiea, Juha Plosilaa
aDepartment of Computing, University of Turku, Turku, 20014, Finland
Abstract
The rapid expansion of the Internet of Things (IoT) has led to an increasing demand for energy-efficient communication strategies,
particularly for battery-operated or energy-efficient devices. This paper explores optimization techniques for enabling energy-
efficient IoT device communication, ensuring uninterrupted operation by balancing energy consumption. We investigate adaptive
transmission strategies and various communication protocols to optimize energy usage while maintaining reliable data exchange.
Our approach leverages dynamic power control and energy-aware scheduling to enhance communication efficiency through the
use of DQL(Deep Q Learning) for parameter optimization. Experimental results and simulations demonstrate the effectiveness of
these optimization techniques in prolonging network lifespan and ensuring sustainable IoT deployments for swarm robotics. A
comparative evaluation with traditional IoT communication protocols showed that our optimization mechanism extended device
lifetime by up to 25 percent, while achieving a balance between energy efficiency and data reliability.
© 2025 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0)
Peer review under the responsibility of the scientific program committees
50 Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56
puting markedly improves privacy, reduces latency, and increases system reliability by processing data at its source
rather than relying on cloud services. This is advantageous for energy-efficient devices, facilitating effective data man-
agement and reducing power consumption. Moreover, the application of machine learning algorithms enhances data
analysis from connected devices, enabling adaptive responses to changing conditions and improving efficiency.
Optimization techniques are critical for allowing energy-efficient IoT devices to operate for an extended period of
time without depleting their energy resources. Dynamic power management, adaptive data transmission, and energy-
aware routing protocols are all examples of optimizations that seek to balance energy consumption and availability.
IoT devices can enhance network performance while operating continuously by strategically adjusting communication
parameters such as transmission power, duty cycles, and data rate. Improving energy efficiency in these devices is
critical for sustainability, as it reduces the costs associated with traditional battery replacements.
Complex communication protocols are required to ensure that data is sent reliably while using as little energy
as possible. Protocols such as MQTT, LoRaWAN, and Zigbee are critical for making communication more energy
efficient. Each has unique advantages that make it ideal for a specific type of IoT application. It is critical to understand
how reliable the energy used is, particularly for applications that require consistent power levels. Integrating these
systems into existing infrastructures is also difficult because they are complex, and it frequently costs more money and
time to research and develop these systems. IoT communication that uses less energy is still improving, demonstrating
a tension between new ideas and real-world applications. This is similar to larger conversations about sustainability
and the practicality of technology. IoT devices can now talk to each other better, which improves operational efficiency.
Improvements in communication protocols, such as adaptive power control and advanced signaling techniques, let
devices use less energy while still sending data reliably, which leads to a more environmentally friendly technological
world. A lot of the important arguments about this subject center on the trade-offs between saving energy and keeping
data safe, as well as the difficulties of using these cutting-edge technologies in various places.
IoT communication that uses less energy could have benefits like lower costs and a smaller impact on the environ-
ment. This is why it is important to keep researching and making progress in this area. IoT is always getting better,
which means that communication methods need to get better at using less energy. This is important for promoting a
sustainable future. Combining advanced technologies with collaborative efforts between businesses, schools, and the
government will be key to creating new ideas that save energy and meet the needs of more and more devices that are
connected to each other.
This research looks at various optimization strategies that can help IoT networks communicate more efficiently
while using less energy. We look into energy models, communication protocols that use less power, and optimization
algorithms that make IoT implementations more environmentally friendly. We integrated Deep Q Learning(DQL)
technique to optimize the parameters. DQL utilizes the reinforcement methods to intelligently adjust trasmission
parameters. The primary objective is to develop intelligent mechanisms that enable devices to operate without the
need for frequent battery replacements or external power sources. Our research aims to contribute to the design of
robust and scalable energy-efficient IoT systems, fostering their adoption in real-world applications.
Energy optimization for swarms of robots operating within various environments is a critical challenge. Achiev-
ing energy-efficient swarm operation is essential for the future of robotic systems. Multiple components within a
robot—such as sensors, electronic systems, and mechanical engines—contribute to the overall energy drained from
the battery. The energy consumption of each component varies significantly based on its characteristics, including the
number of sensors, computational power, communication technology, and mechanical size. Thus, optimizing overall
energy consumption requires a comprehensive approach that addresses all components.
Research has previously focused on optimizing specific components separately, such as mechanical energy, compu-
tational energy efficiency, sensor gating, and low-power communication. However, in robotics, the interdependencies
among these components mean that decisions made for one part can influence the operations of others.
These interdependencies underscore the importance of co-optimizing interconnected components. Some studies
have proposed co-optimization strategies for the communication and computational components of robots. In swarm
robotics, these interdependencies become even more pronounced due to task sharing among nodes and the associated
costs. For example, Figure 1 illustrates the relationship between power costs for communication using four different
protocols and the distance between two nodes while transferring a specific data volume.
Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56 51
The main contributions of this paper include proposing a energy efficient approach to optimize power consumption
while considering communication costs between nodes as well as selection and analysis of communication protocols
based on overall efficiency by incorporating Deep Q-Learning (DQL) into our optimization approach.
Table 1. Protocol selection parameters.
Protocol Strength Energy Efficiency (Reliability)
BLE Short-range, low power High
LoRaWAN Long-range, scalable High
MQTT Lightweight, publish/subscribe Moderate to High
CoAP Low overhead High
2. Related work
The field of energy-efficient IoT device communication has garnered significant research attention due to the in-
creasing need for sustainable and autonomous networks for swarm robotics. A multitude of studies have examined
diverse optimization techniques, energy efficiency strategies, and communication protocols to enhance the perfor-
mance and longevity of IoT devices.
Methods for optimizing energy-efficient communication in IoT devices are crucial for extending device operational
lifespan and minimizing energy consumption. A variety of strategies and protocols have been developed to address
these challenges. The H-HEED protocol necessitates several iterations for clustering finalization and incurs consid-
erable energy consumption; however, it has demonstrated an enhancement in network longevity of about 63 percent
[22]. The energy-based rotated HEED (ER-HEED) protocol employs comparable clustering techniques; however, it
is insufficient in addressing interference from adjacent nodes, which may lead to heightened packet loss and reduced
network lifespan.
Several optimization strategies have been used to improve energy efficiency in IoT networks. A study [19] pre-
sented an adaptive duty cycling algorithm that reduces power by systematically scheduling device activation times.
Additionally, [20] used machine learning to optimize transmission power and data rates for better energy efficiency in
communication. To address interference, several protocols have been developed, most notably the Energy-Intelligent
Geographical Routing (EIGR) protocol, which focuses on adjusting transmission power to reduce interference during
data transmission to the next forwarding node. This modification is necessary because increased neighbor density
leads to more interference and energy depletion [21]. Recent studies have utilized ad hoc on-demand distance vec-
tor (AODV) protocols to identify the shortest paths that circumvent interference zones, thereby enhancing energy
efficiency in wireless networks[5].
The choice of communication protocol has a significant impact on energy consumption in IoT networks. BLE,
LoRaWAN, MQTT, and CoAP protocols have all been tested for their effectiveness in energy-limited environments.
The evaluation of these protocols shows that LoRaWAN is best suited for long-range applications, whereas BLE
is best for short-range, low-power scenarios. The selection of an appropriate communication protocol is critical for
increasing energy efficiency and protecting IoT devices. Each protocol has distinct advantages and is tailored to
specific applications, necessitating that developers and engineers assess their individual needs when implementing IoT
solutions. Multi-hop communication strategies outperform single-hop transmission in terms of energy consumption,
as shorter transmission distances can increase network lifespan [18].
Recent studies have demonstrated the effectiveness of Adaptive Power Control and Network Topology Optimiza-
tion in increasing energy efficiency. These strategies have exhibited reductions in energy consumption between 15
and 25 percent, highlighting the significance of customized optimization techniques in the implementation of smart
city IoT networks[12]. The incorporation of Narrowband IoT (NB-IoT) technology and LTE-M within organized
frameworks enhances energy management and resource utilization in smart cities[9].
Attaining energy efficiency necessitates the use of low-power hardware and the optimization of software. Energy-
efficient microcontrollers and sensors can markedly diminish power consumption. Moreover, adopting efficient coding
52 Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56
practices reduces computational demands and encourages low-power modes during periods of inactivity. Routine
software updates can rectify inefficiencies and improve energy performance over time[5].
Recent advancements have integrated machine learning for energy-efficient IoT communication. Research such as
[8] has shown that deep learning models can forecast energy consumption trends and enhance transmission schedules.
Another approach [7] proposed federated learning for distributed IoT nodes to reduce overhead and optimize power.
Despite the advancements, there are still gaps in the integration of multiple energy optimization techniques. Ex-
isting works mainly focus on either energy efficiency or optimization algorithms, with limited studies addressing a
holistic mechanism combining both. Moreover, the trade-offs between latency, reliability, and energy efficiency need
further exploration. This work builds upon these previous studies by presenting an integrated optimization mechanism
that leverages energy-aware transmission protocols, dynamic power management, and adaptive scheduling to achieve
sustainable IoT communication. The study contributes to the development of self-sustaining IoT systems, reducing
dependency on external power sources and enhancing the viability of large-scale IoT networks. We present results for
power analysis, packet delivery ratio and latency in various scenarios of distance and data transmission for different
protocols.
3. Power optimization strategies
Conventional energy management techniques, like static duty cycling, frequently do not adjust to fluctuating en-
ergy conditions, resulting in inefficient performance or system malfunctions. This study focuses on developing a
communication mechanism that enhances energy efficiency by optimizing power consumption and communication
effectiveness for the smooth functioning of IoT devices. Our objective is to develop a sustainable IoT communication
system for swarm robotics, that utilizes models, real-time optimization algorithms, and adaptive control mechanisms
to effectively respond to energy fluctuations while ensuring reliable data transmission.
In addition to tackling energy supply challenges, enhancing communication protocols among IoT devices is essen-
tial for realizing energy efficiency. Proposed solutions focused on improving communication efficiency and reducing
energy consumption incorporate cross-layer strategies that integrate the Physical, Medium Access Control (MAC),
and Network layers. This investigation aims to connect theoretical energy-efficient operations with practical applica-
tions by integrating optimization techniques and real-world hardware implementations.
The combination of BLE, LoRaWAN [2], MQTT, and CoAP encompasses a broad range of IoT communication
scenarios, from localized device interaction to extensive data transmission and cloud integration. Each protocol, as
outlined in Table 1, presents distinct benefits that, when optimized, markedly decrease overall communication ex-
penses while maintaining the device’s energy efficiency. This research integrates BLE, LoRaWAN, MQTT, and CoAP
to address diverse IoT communication requirements, improving coverage, reliability, and power efficiency for a sus-
tainable communication system.
The primary challenge in attaining energy efficiency in IoT involves optimizing energy consumption to correspond
with the variable availability of energy, while simultaneously ensuring reliable data transmission and reducing la-
tency. Conventional communication protocols and power management strategies, tailored for static energy supplies,
are ineffective in energy-intensive situations. This necessitates innovative optimization techniques that dynamically
modify transmission power, duty cycling, and data scheduling to guarantee sustainable and efficient performance. We
can improve the performance and longevity of IoT networks by creating more energy-efficient algorithms, models,
and hardware. This study seeks to balance energy availability with communication needs, ensuring the continuous and
efficient operation of IoT devices while reducing reliance on nonrenewable energy sources.
Optimization techniques for energy efficient IoT device communication are crucial in reducing energy consumption
while maintaining operational efficiency for swarm robotics. Various strategies have been identified and implemented
to achieve these goals, significantly enhancing the sustainability of IoT deployments. Developing devices with low-
power chips and sensors is one key strategy for minimizing energy consumption without sacrificing functionality.
For example, GAO Tek employs advanced NB-IoT modules that integrate low-power components to optimize energy
management in IoT systems.
To further enhance the energy efficiency of IoT communication, we integrate Deep Q-Learning (DQL) into our
optimization. DQL enables intelligent adaptation of transmission parameters by utilizing reinforcement learning to
dynamically adjust power levels, duty cycles, and routing decisions based on real-time network conditions and en-
Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56 53
Fig. 1. (a) Power analysis for transmitting 500Kb, (b) Power analysis for transmitting 1MB, (c) Power analysis for transmitting 1.5MB.
ergy availability. The DQL model is trained using experience replay and a deep neural network (DNN) for Q-value
approximation. The IoT device logs power consumption, transmission success rate, and environmental factors over
multiple episodes, storing past experiences in a replay buffer to improve learning stability. The Q-network is updated
using a target network to prevent instability in learning. Once trained, the optimized policy allows the device to adap-
tively switch between transmission power levels and optimize wake/sleep schedules based on real-time conditions,
significantly reducing unnecessary energy expenditures. By integrating DQL into our Energy-Aware Transmission
Power Control (EATPC) algorithm, we enhance dynamic transmission power control, duty cycle optimization, and
energy-aware routing.
4. Proposed approach
This section introduces the potential methodology scenario and discusses several results from proprietary protocols
like MQTT, BLE, CoAP and LoRaWAN for the proposed energy efficient IoT communication system that leverages
efficient energy consumption.
4.1. Setup
To achieve energy-efficient communication for IoT devices, we propose a multi-layered optimization approach by
integrating energy adaptive communication protocols and machine learning-based optimization techniques. In energy-
constrained IoT networks, efficient path planning and balanced computational workload distribution are critical to
minimizing overall communication costs while extending device lifetime.
This study proposes an approach to optimizing communication costs for energy-efficient IoT devices that takes
into account computational power and protocol selection. The methodology is intended to reduce the overall energy
consumption of the network while maintaining dependable and efficient data transmission. The first step is to represent
the IoT network as a collection of interconnected nodes, each with its own set of attributes such as remaining energy,
processing power, distance to data sinks, and protocol adherence.
The methodology employs several communication protocols, including BLE, LoRaWAN, MQTT, and CoAP. A
Protocol Selection Matrix is used to determine the best protocol for each communication link, taking into account
factors such as transmission range, data payload size, and the energy characteristics of the protocols used. For ex-
ample, BLE is preferred for short-range, low-energy communication, whereas LoRaWAN is chosen for long-range
transmissions when applicable.
An Energy-Aware Transmission Power Control (EATPC) algorithm dynamically adjusts the transmission power Pt
based on available energy and distance d:
Pt(d) = Pmin + αdn (1)
where the greek letter ”alpha” is a scaling factor, and n is the path-loss exponent (typically 2-4).
54 Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56
Fig. 2. Power consumption, Packet delivery ratio and Latency at rest.
The energy-aware communication protocol was implemented to optimize data transmission based on available
energy. It’s key features include three key aspects of (i) dynamic transmission power control which adjusts the trans-
mission power based on residual energy levels and network conditions. (ii) Duty Cycle Optimization which adapts
the sleep/wake schedules to balance energy consumption and data throughput. (iii) Energy-Aware Routing which im-
plements an energy-efficient routing algorithm to minimize communication overhead. A heuristic-based optimization
algorithm is applied to dynamically select the best communication protocol based on the channel conditions (RSSI,
SNR), network latency as well as throughput and energy consumption constraints.
The total energy consumed for communication is the sum of the individual components:
EComm = Etx + Erx + Elisten + Eoverhead (2)
A real-world testbed was deployed using NVIDIA Jetson Nano devices equipped with energy efficient communi-
cation modules (LoRa modules, BLE sensors). Performance metrics such as energy consumption, data transmission
reliability, and network lifetime were analyzed under varying environmental conditions. We use an Intel Core i5 based
laptop which acts as the edge gateway, processing the optimization protocols and data analysis. A NVIDIA Jetson
Nano serves as an IoT end node with an external current consumption sensor attached to it.
The total energy consumption of the IoT device is modeled by considering the energy usage of each module:
EConsumed = ESense + EProcess + EComm + EIdle, (3)
The primary data extracted from the onboard sensors are the power consumption of the Nano along with the
voltage as well as the current readings while the application is running in a loop. The values are used to establish
a relationship between the transmission and the reception sequences while the Nano is operating. We started the
data logging at an initial distance and kept on repeating the experiment after moving the node every 10 meters. We
carried on with recording the measurements until a distance of 70 meters was achieved and after this significant
distortion in the communication capabilities was observed. We used the obtained data to form a relationship between
the communication power and the distance covered by the node while our application was running on the Nano.
Our experimental setup consists of an Intel Core i5 based laptop, NVIDIA Jetson Nano as the IoT nodes and a
wireless communication infrastructure. The key hardware/software components are wireless communication modules,
INA219 current/voltage sensor for real-time power measurement, FreeRTOS as the operating system for efficient task
scheduling,communication protocols MQTT, CoAP, BLE, and LoRaWAN for comparative evaluation. After setting
up the basic infrastructure, we calibrate the INA219 current/voltage sensor for baseline power measurements. Each
protocol (MQTT, CoAP, BLE, and LoRaWAN) is tested individually with fixed data transmission rates. Network per-
formance metrics are recorded on the laptop communicating with the Jetson Nano which include power consumption,
packet delivery ratio and latency. Data is collected as well as analysed over experiments in controlled and real-world
environments suited for swarm robotics.
Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56 55
Fig. 3. (a) Power analysis; (b) Packet delivery ratio analysis ; (c) Latency analysis.
4.2. Results
As presented in Figure 3, our experiments demonstrated that energy adaptive communication protocols and ma-
chine learning-based optimization techniques improved energy efficiency by 10-15 percent. Devices were able to
sustain longer operation even in fluctuating environmental conditions. Moreover, utilizing sophisticated signaling
protocols, such as SNR (Signal-to-Noise Ratio) optimization, can improve connection quality, especially in areas with
interference or low signal quality. This approach reduces the need for re-transmissions, thereby conserving energy.
By analysing results from the node at rest, which are presented in Figure 2, we also concluded, that to manage the
vast amounts of data generated by IoT devices efficiently, data aggregation and pre-processing can be implemented at
the edge of the network. This reduces unnecessary transmission, effectively cutting down on power usage. Moreover,
allowing local devices to perform basic computations before sending data to centralized systems improves reliability.
The adaptive power control mechanism reduced unnecessary energy consumption, resulting in an average power
savings of 20 percent while maintaining reliable connectivity. Moreover, the optimized communication mechanism
ensured a consistent data transmission rate with minimal packet loss (below 5 percent), even under low-energy con-
ditions. Furthermore, the IoT node dynamically adjusted their sleep/wake schedules, leading to an improvement in
network lifespan by 20 percent. Therefore, implementing adaptive power control strategies can significantly enhance
energy efficiency in IoT networks. For example, adjusting transmission power based on real-time network conditions
helps minimize energy wastage for swarm robotics, while ensuring reliable communication.
In our analysis, a fixed data volume of 500KB, 1MB and 1.5MB is transmitted to reflect typical IoT sensor data
aggregation over a period of time. This allows for a fair comparison across BLE, LoRaWAN, MQTT, and CoAP,
focusing on how efficiently each protocol handles data transmissions both before and after optimization strategies
such as Deep Q-Learning (DQL) are applied.
A comparative evaluation with traditional IoT communication protocols showed that our optimization mechanism
extended device lifetime by up to 25 percent, while achieving a balance between energy efficiency and data reliability.
The testbed results demonstrated that the proposed method kept IoT devices running efficiently while they were
outdoors and over long distances. This demonstrated that long-term, scalable IoT communication is feasible.
5. Conclusion
This study looked into strategies for improving energy efficiency in communication between IoT devices, with a
focus on effective energy management and adaptive transmission techniques. We demonstrated that IoT devices can
achieve long-term operation while maintaining reliable data transmission by utilizing modeling, parameter optimiza-
tion, and real-time energy-efficient algorithms.
The results indicate that integrating adaptive power control and duty-cycling strategies significantly enhances en-
ergy efficiency without compromising communication quality. The adaptive power control mechanism reduced unnec-
essary energy consumption, resulting in an average power savings of 20 percent while maintaining reliable connec-
56 Amir Ijaz  et al. / Procedia Computer Science 265 (2025) 49–56
tivity. Moreover, the optimized communication mechanism ensured a consistent data transmission rate with minimal
packet loss (below 5 percent), even under low-energy conditions. Furthermore, the IoT node dynamically adjusted
their sleep/wake schedules, leading to an improvement in network lifespan by 20 percent. The suggested optimiza-
tion mechanism guarantees a balance between energy efficiency and consumption, thus improving device autonomy.
Moreover, the tangible hardware execution on platforms like the Jetson TX2 and Jetson Nano validates our approach.
Future efforts will focus on improving machine learning-based predictive models for energy availability and opti-
mizing network-level coordination among diverse IoT nodes. By enhancing these techniques, we can approach gen-
uinely autonomous, self-sustaining IoT systems for swarm robotics, that function without external maintenance.
References
[1] Kim, Sungryul, and Younghwan Yoo. (2018) “Contention-aware adaptive data rate for throughput optimization in LoRaWAN.” Sensors, 18(6),
1716. MDPI.
[2] Reynders, Brecht, Qing Wang, Pere Tuset-Peiro, Xavier Vilajosana, and Sofie Pollin. (2018) “Improving reliability and scalability of Lo-
RaWANs through lightweight scheduling.” IEEE Internet of Things Journal, 5(3), 1830–1842. IEEE.
[3] Triantafyllou, Anna, Dimitrios Zorbas, and Panagiotis Sarigiannidis. (2022) “Time-slotted LoRa MAC with variable payload support.” Com-
puter Communications, 193, 146–154. Elsevier.
[4] Benkhelifa, Fatma, Zhijin Qin, and Julie A. McCann. (2021) “User fairness in energy harvesting-based LoRa networks with imperfect SF
orthogonality.” IEEE Transactions on Communications, 69(7), 4319–4334. IEEE.
[5] BenSaleh, Mohammed Sulaiman, Raoudha Saida, Yessine Hadj Kacem, and Mohamed Abid. (2020) “Wireless sensor network design method-
ologies: A survey.” Journal of Sensors, 2020(1), 9592836. Wiley Online Library.
[6] Ketshabetswe, Lucia Keleadile, Adamu Murtala Zungeru, Mmoloki Mangwala, Joseph M. Chuma, and Boyce Sigweni. (2019) “Communica-
tion protocols for wireless sensor networks: A survey and comparison.” Heliyon, 5(5). Elsevier.
[7] Modieginyane, Kgotlaetsile Mathews, Babedi Betty Letswamotse, Reza Malekian, and Adnan M. Abu-Mahfouz. (2018) “Software defined
wireless sensor networks application opportunities for efficient network management: A survey.” Computers & Electrical Engineering, 66,
274–287. Elsevier.
[8] Kori, Gururaj S., Mahabaleshwar S. Kakkasageri, and Sunil Kumar S. Manvi. (2021) “Computational intelligent techniques for resource
management schemes in wireless sensor networks.” In Recent Trends in Computational Intelligence Enabled Research, 41–59. Elsevier.
[9] Deng, Xiaoheng, Peiyuan Guan, Cong Hei, Feng Li, Jianqing Liu, and Naixue Xiong. (2021) “An intelligent resource allocation scheme in
energy harvesting cognitive wireless sensor networks.” IEEE Transactions on Network Science and Engineering, 8(2), 1900–1912. IEEE.
[10] Mukherjee, Amrit, Pratik Goswami, Ziwei Yan, Lixia Yang, and Joel JPC Rodrigues. (2019) “ADAI and adaptive PSO-based resource alloca-
tion for wireless sensor networks.” IEEE Access, 7, 131163–131171. IEEE.
[11] Tyagi, Sumarga Kumar Sah, Amrit Mukherjee, Shiva Raj Pokhrel, and Kamal Kant Hiran. (2020) “An intelligent and optimal resource alloca-
tion approach in sensor networks for smart agri-IoT.” IEEE Sensors Journal, 21(16), 17439–17446. IEEE.
[12] Farhan, Laith, Rasha Subhi Hameed, Asraa Safaa Ahmed, Ali Hussein Fadel, Waled Gheth, Laith Alzubaidi, Mohammed A. Fadhel, and
Muthana Al-Amidie. (2021) “Energy efficiency for green internet of things (IoT) networks: A survey.” Network, 1(3), 279–314. MDPI.
[13] Moore, Samuel J., Chris D. Nugent, Shuai Zhang, and Ian Cleland. (2020) “IoT reliability: a review leading to 5 key research directions.” CCF
Transactions on Pervasive Computing and Interaction, 2, 147–163. Springer.
[14] Adil, Muhammad, Rahim Khan, Mohammed Amin Almaiah, Muhammad Binsawad, Jehad Ali, Adeeb Al Saaidah, and Qui Thanh Hoai Ta.
(2020) “An efficient load balancing scheme of energy gauge nodes to maximize the lifespan of constraint oriented networks.” IEEE Access, 8,
148510–148527. IEEE.
[15] Adil, Muhammad, Rahim Khan, Jehad Ali, Byeong-Hee Roh, Qui Thanh Hoai Ta, and Mohammed Amin Almaiah. (2020) “An energy profi-
cient load balancing routing scheme for wireless sensor networks to maximize their lifespan in an operational environment.” IEEE Access, 8,
163209–163224. IEEE.
[16] Khan, Muhammad Nawaz, Haseeb Ur Rahman, Mohammed Amin Almaiah, Muhammad Zahid Khan, Ajab Khan, Mushtaq Raza, Mohammed
Al-Zahrani, Omar Almomani, and Rahim Khan. (2020) “Improving energy efficiency with content-based adaptive and dynamic scheduling in
wireless sensor networks.” IEEE Access, 8, 176495–176520. IEEE.
[17] Liu, Xuesong, and Jie Wu. (2019) “A method for energy balance and data transmission optimal routing in wireless sensor networks.” Sensors,
19(13), 3017. MDPI.
[18] Wang, Kun, Chih-Min Yu, and Li-Chun Wang. (2020) “DORA: A destination-oriented routing algorithm for energy-balanced wireless sensor
networks.” IEEE Internet of Things Journal, 8(3), 2080–2081. IEEE.
[19] Patel, Nileshkumar R., Shishir Kumar, and Sanjay Kumar Singh. (2021) “Energy and collision aware WSN routing protocol for sustainable
and intelligent IoT applications.” IEEE Sensors Journal, 21(22), 25282–25292. IEEE.
[20] Kumar, Mohit, Xingzhou Zhang, Liangkai Liu, Yifan Wang, and Weisong Shi. (2020) “Energy-efficient machine learning on the edges.” In
2020 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), 912–921. IEEE.
[21] Farhan, Laith, Omprakash Kaiwartya, Laith Alzubaidi, Waled Gheth, Eric Dimla, and Rupak Kharel. (2019) “Toward interference aware IoT
framework: energy and geo-location-based-modeling.” IEEE Access, 7, 56617–56630. IEEE.
[22] Kour, Harneet, and Ajay K. Sharma. (2015) “Performance evaluation of HEED and H-HEED protocol for realistic models in WSN.” In 2015
International Conference on Computer, Communication and Control (IC4), 1–5. IEEE.