Suganthi Sundararaj, Umapathi Nagappan, Mahdal Miroslav, Ramachandran Manickam
Department of Computer and Communication, Sri Sairam Institute of Technology, Chennai 600044, India.
Department of Electronics and Communication Engineering, Jyothishmathi Institute of Technology and Science, Karimnagar 505481, India.
Sensors (Basel). 2022 Feb 23;22(5):1736. doi: 10.3390/s22051736.
Wireless Sensor Networks (WSNs) can be defined as a cluster of sensors with a restricted power supply deployed in a specific area to gather environmental data. One of the most challenging areas of research is to design energy-efficient data gathering algorithms in large-scale WSNs, as each sensor node, in general, has limited energy resources. Literature review shows that with regards to energy saving, clustering-based techniques for data gathering are quite effective. Moreover, cluster head (CH) optimization is a non-deterministic polynomial (NP) hard problem. Both the lifespan of the network and its energy efficiency are improved by choosing the optimal path in routing. The technique put forth in this paper is based on multi swarm optimization (MSO) (i.e., multi-PSO) together with Tabu search (TS) techniques. Efficient CHs are chosen by the proposed system, which increases the optimization of routing and life of the network. The obtained results show that the MSO-Tabu approach has a 14%, 5%, 11%, and 4% higher number of clusters and a 20%, 6%, 14%, and 6% lesser average packet loss rate as compared to a genetic algorithm (GA), differential evolution (DE), Tabu, and MSO based clustering, respectively. Moreover, the MSO-Tabu approach has 136%, 36%, 136%, and 38% higher lifetime computation, and 22%, 16%, 51%, and 12% higher average dissipated energy. Thus, the study's outcome shows that the proposed MSO-Tabu is efficient, as it enhances the number of clusters formed, average energy dissipated, lifetime computation, and there is a decrease in mean packet loss and end-to-end delay.
无线传感器网络（WSNs）可以定义为部署在特定区域以收集环境数据的、电源受限的传感器集群。最具挑战性的研究领域之一是在大规模无线传感器网络中设计节能数据收集算法，因为通常每个传感器节点的能量资源有限。文献综述表明，在节能方面，基于聚类的数据收集技术相当有效。此外，簇头（CH）优化是一个非确定性多项式（NP）难题。通过在路由中选择最优路径，可以提高网络的寿命及其能源效率。本文提出的技术基于多群体优化（MSO）（即多粒子群优化）和禁忌搜索（TS）技术。所提出的系统选择高效的簇头，这增加了路由优化和网络寿命。获得的结果表明，与遗传算法（GA）、差分进化（DE）、禁忌算法和基于MSO的聚类相比，MSO-禁忌算法的簇数量分别高出14%、5%、11%和4%，平均丢包率分别低20%、6%、14%和6%。此外，MSO-禁忌算法的寿命计算高出136%、36%、136%和38%，平均耗散能量高出22%、16%、51%和12%。因此，该研究结果表明，所提出的MSO-禁忌算法是高效的，因为它增加了形成的簇数量、平均能量耗散、寿命计算，并且平均丢包和端到端延迟有所减少。
