Dynamic PSO Based Fuzzy Clustering Algorithm for WSNs
Tanima Kole Bhowmik, Indrajit Banerjee · 2019
Clustering is a major methodology in any wireless sensor networks to attain energy efficiency. In clustering technique, cluster head is nominated in the network then the cluster is shaped by the sensor node by joining the adjacent cluster head. The major problem in this process, it forms the unbalanced cluster. The distribution of the cluster head in the network may be uneven. Here we propose a dynamic Particle swarm optimization based fuzzy clustering algorithm (PSO MF), to overcome the cited problem. In PSO MF, fuzzy C Means algorithm is cast-off entire nodes into a stable cluster. The suitable cluster heads are nominated via a Mamdani fuzzy inference system (MFIS). The fuzzy inputs of the MFIS contain the parameters like residual energy, node degree 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 effect on the attainment of the fuzzy system. We operate a particle swarm optimization algorithm to optimize the fuzzy instructions on PSO MF. The fitness function of the algorithm is precise to elongate the network lifespan. Simulation outcome shows that the proposed clustering algorithm is found to yield best results over other conventional algorithms.