An Improved PSO Based Fuzzy Clustering Algorithm in WSNs

In modern wireless technology, wireless sensor networks with Internet of Thing is poised to be one of the most disruptive technology over the next degrade. Clustering is a major methodology in any wireless sensor networks to minimize the consumption of energy. So overall network life span is increased. In distributed approach, sensor nodes are grouped inefficient manner to create cluster. Then cluster head is nominated in the network. The problem is that it forms hot spot and energy hole problems. The distribution of the cluster head in the network is also not properly distributed. Here we propose an improved Particle swarm optimization based fuzzy clustering algorithm (IPSOFC), to overawe the cited problem. By using improved Particle swam optimization, a new fitness function is proposed which nominated a suitable cluster head. Fuzzy c-means clustering (FCM) can overcome the energy hole and hot spot problem by choosing exclusive boundary for individual cluster head, so un-identical clustering is established. The fuzzy inputs of the FCM contain the parameters like residual energy, node density and distance to sink. In literature, the table based on the fuzzy rule is manually elaborated. So when we tune the fuzzy instructions, it will affect the attainment of the fuzzy system. The fitness function is precise to elongate the network lifespan. Simulation result is verified with its proposed algorithm with some existing algorithm. Outcome indicates that the proposed clustering algorithm is found to yield best results over other conventional algorithms.