Game Theoretic Clustering Algorithm for Large Scale WSN

Reasonable clustering algorithm can prolong the Large Scale Wireless Sensor Network (LS-WSN) lifetime, so as to reduce the cost of using it. Up to now, most research on WSN clustering assumes that the nodes in WSN are uniformly distributed. But this assumption is different from practical LS-WSN. In this paper, a novel clustering algorithm is proposed, which is suitable for the non-uniformly distributed LS-WSNs. In the proposed algorithm, LS-WSN is clustered according to the location aware clustering method, which based on the hexagons virtual grid, then the clusters are adjusted following a game theoretic adjustment process. After the algorithm, the numbers of nodes in each cluster is nearly similar. Theoretical analysis and simulation results show that the proposed algorithm can effectively balance the number of nodes in each cluster, thus prolonging the effective lifetime of WSN.