Wang Chuhang
College of Computer Science and Technology, Changchun Normal University, Changchun 130032, China.
Sensors (Basel). 2023 Jul 26;23(15):6699. doi: 10.3390/s23156699.
Clustering is considered to be one of the most effective ways for energy preservation and lifetime maximization in wireless sensor networks (WSNs) because the sensor nodes are equipped with limited energy. Thus, energy efficiency and energy balance have always been the main challenges faced by clustering approaches. To overcome these, a distributed particle swarm optimization-based fuzzy clustering protocol called DPFCP is proposed in this paper to reduce and balance energy consumption, to thereby extend the network lifetime as long as possible. To this end, in DPFCP cluster heads (CHs) are nominated by a Mamdani fuzzy logic system with descriptors' residual energy, node degree, distance to the base station (BS), and distance to the centroid. Moreover, a particle swarm optimization (PSO) algorithm is applied to optimize the fuzzy rules, instead of conventional manual design. Thus, the best nodes are ensured to be selected as CHs for energy reduction. Once the CHs are selected, distance to the CH, residual energy, and deviation in the CH's number of members are considered for the non-CH joining cluster in order to form energy-balanced clusters. Finally, an on-demand mechanism, instead of periodic re-clustering, is utilized to maintain clusters locally and globally based on local information, so as to further reduce computation and message overheads, thereby saving energy consumption. Compared with the existing relevant protocols, the performance of DPFCP was verified by extensive simulation experiments. The results show that, on average, DPFCP improves energy consumption by 38.20%, 15.85%, 21.15%, and 13.06% compared to LEACH, LEACH-SF, FLS-PSO, and KM-PSO, and increases network lifetime by 46.19%, 20.69%, 20.44%, and 10.99% compared to LEACH, LEACH-SF, FLS-PSO, and KM-PSO, respectively. Moreover, the standard deviation of the residual network was reduced by 61.88%, 55.36%, 54.02%, and 19.39% compared to LEACH, LEACH-SF, FLS-PSO, and KM-PSO. It is thus clear that the proposed DPFCP protocol efficiently balances energy consumption to improve the overall network performance and maximize the network lifetime.
聚类被认为是无线传感器网络（WSN）中实现能量保存和寿命最大化的最有效方法之一，因为传感器节点的能量有限。因此，能量效率和能量平衡一直是聚类方法面临的主要挑战。为了克服这些问题，本文提出了一种基于分布式粒子群优化的模糊聚类协议DPFCP，以减少和平衡能量消耗，从而尽可能延长网络寿命。为此,在DPFCP中，簇头（CH）由一个Mamdani模糊逻辑系统指定，该系统考虑描述符的剩余能量、节点度、到基站（BS）的距离以及到质心的距离。此外，应用粒子群优化（PSO）算法来优化模糊规则，而不是传统的手动设计。因此，确保选择最佳节点作为簇头以减少能量消耗。一旦选择了簇头，在非簇头加入簇时考虑到该簇头的距离、剩余能量以及簇头成员数量的偏差，以形成能量平衡的簇。最后，利用按需机制而非周期性重新聚类，基于本地信息在本地和全局维护簇，从而进一步减少计算和消息开销，进而节省能量消耗。与现有相关协议相比，通过广泛的模拟实验验证了DPFCP的性能。结果表明，平均而言，与LEACH、LEACH-SF、FLS-PSO和KM-PSO相比，DPFCP的能量消耗分别降低了38.20%、15.85%、21.15%和13.06%，网络寿命分别增加了46.19%、20.69%、20.44%和10.99%。此外，与LEACH、LEACH-SF、FLS-PSO和KM-PSO相比，剩余网络的标准差分别降低了61.88%、55.36%、54.02%和19.39%。因此，很明显，所提出的DPFCP协议有效地平衡了能量消耗，以提高整体网络性能并最大化网络寿命。
