Adaptive clustering in MANETs using graph theoretical algorithms

MANETs are distributed systems consisting of mobile wireless nodes that are self organized and perform operations to enable the communication in the network. An efficient way to exchange information between nodes in the network is to form virtual groups among nodes called clusters. This paper presents the performance analysis of a distributed maximal weighted independent set (MWIS) algorithm for clustering. The algorithm incurs low overhead in terms of processing convergence time and messages exchanged. It also achieves fairly uniform cluster head distribution across the network. Our simulation results demonstrate that MWIS outperforms weights adapted-based clustering protocols in terms of less signalling complexity effectiveness.