Investigating Optimization Techniques for Cluster Head Election in WSN

Wireless sensor network (WSN) is a briskly augmenting hightech platform with remarkable and neoteric applications. Many new protocols specifically designed for the requirement of energy awareness are provided as per consequence of newfangled advances in WSN. In Actu, optimization of the network operation is vital to prolong network’s lifetime. For energy-efficiency in WSNs, one of the most accepted solutions is to cluster the networks. The regular nodes sensing the field and sending their data to the cluster-head, and then, transmitting to the base station is a process usually followed in a typical clustered WSN. Furthermore, cluster formation done inaptly, can make some CHs burdened with high number of sensor nodes. This overwork may lead to abrupt death of the CHs thereby deteriorating the overall performance of the WSN. Network Lifetime can be increased by preventing faster death of the highly loaded CHs. Three evolutionary algorithms namely Flower Pollination Algorithm (FPA), Harmony Search Algorithm (HSA) and Particle Swarm Optimization (PSO) with appropriate fitness functions are compared with the intrinsic properties of clustering in mind. The main idea is the embodiment of criteria of compactness (i.e. cohesion) and separation error in the fitness function to direct the search into promising solutions. The property of heterogeneity of nodes, in terms of their energy; in hierarchically clustered wireless sensor networks has also been involved. Simulation over 20 random heterogeneous WSNs shows that our FPA always prolongs the network lifetime, sustain more energy in comparison to the results obtained using the PSO and HSA protocols.