Optimal Cluster Number Determination for Clustered Wireless Sensor Networks

A method to determine the optimal clusterhead number in a clustered wireless sensor network is proposed to improve the energy efficiency of clustering schemes. The wireless sensor network is modeled as points locating randomly on the plane according to a homogeneous spatial Poisson process. This leads to derive the means of some typical cluster parameters accordingly. After that, the total energy cost consumed in the system is calculated based on a practical radio model. As a result, the optimal probability of becoming a clusterhead for each node is deduced via minimizing the system energy cost. Ultimately, the total energy costs are compared in the scenarios that the networks employ the same clustering algorithm with a range of values for the probability of becoming a clusterhead. Simulation results show that the optimal probability we obtained has indeed minimized energy dissipation in the system and is applicable to other clustering algorithms directly.