A Modified Clustering Algorithm Based on Fuzzy Clustering (FC) and Modified Grey Wolf Optimization (MGWO)

Abstract The goal of a Wireless Sensor Network (WSN) is to extend the life cycle of network and to control topology control. Particle Swarm Optimization is used to pick cluster heads based on Particle Swarm Optimization (PSO). In high-dimensional space, PSO can easily fall into local optimum, results in poor convergence rate when subjected to an iterative process. Proposed technique applies Fuzzy Clustering (FC) for preprocessing and Modified Grey Wolf Optimization to address this problem (MGWO). First, the FC algorithm is used to initialize the clustering process for sensor nodes based on their geographical locations, where each node belonging to a cluster with a given probability, and the count of initial clusters is studied and presented. In addition, the fitness function is created with WSN's energy consumption and distance aspects in mind. Finally, the MGWO is used to determine the CH nodes in hierarchical architecture. Experiments reveal that, when compared to standard methods, the proposed strategy was successful in lowering node mortality and prolonging the network life cycle.