A new distributed intelligent clustering protocol (ICP) in wireless sensor networks

Constraint of energy is one of the main research challenges in wireless sensor network protocol development. High energy efficiency can be achieved if sensor nodes are organized in to groups to form clusters. In this paper, we propose a new distributed clustering approach in which the set of appropriate cluster heads are elected in each data gathering period upon real time decision made by fuzzy expert system. Fuzzy based cluster head election based on neighbor density and centrality requires either local topology information exchange among nodes or the base station can centrally compute upon reception of information from sensor nodes which naturally consumes significant amount of energy in cluster set up phase. In the proposed protocol, we have defined new approach to estimate level of neighbor density based on mathematical formulations to reduce packet overhead in cluster formation. The fuzzy parameters, residual energy of node and newly defined level of neighbor density put a limit on qualification of nodes to act as cluster heads in the proposed protocol. We have developed fuzzy logic tool on the top of OMNeT++ simulation platform and the performance of the proposed protocol is compared with LEACH and CHEF protocols. A simulation result underpins the superiority of the proposed protocol.