Improving on HEED protocol of wireless sensor networks using non probabilistic approach and fuzzy logic (HEED-NPF)

Clustering sensor nodes is an effective topology control for increasing network lifetime and scalability. It also, balances the load on the sensor nodes. HEED is a well known distributed clustering protocol that uses both energy and communication cost to elect Cluster Heads (CHs) in a probabilistic way. This paper improves HEED protocol using fuzzy logic and a non probabilistic approach for CH election. Fuzzy logic control is capable of making wisely decisions from blending different environmental parameters. The protocol uses nodes remaining energy and two fuzzy descriptors, node degree and node centrality, to elect CHs. On the other hand, the probabilistic model of HEED for electing tentative CHs is eliminated by introducing delay, inversely proportional to the nodes residual energy, for each node. Therefore, tentative CHs are selected according to their priority. Simulation results demonstrate that our approach performs better than HEED in terms of network lifetime and also cluster formation.