A mobility based metric for clustering in mobile ad hoc networks

We present a novel relative mobility metric for mobile ad hoc networks (MANETs). It is based on the ratio of power levels due to successive receptions at each node from its neighbors. We propose a distributed clustering algorithm, MOBIC, based on the use of this mobility metric for selection of clusterheads, and demonstrate that it leads to more stable cluster formation than the "least clusterhead change" version of the well known Lowest-ID clustering algorithm (Chiang et al., 1997). We show reduction of as much as 33% in the rate of clusterhead changes owing to the use of the proposed technique. In a MANET that uses scalable cluster-based services, network performance metrics such as throughput and delay are tightly coupled with the frequency of cluster reorganization. Therefore, we believe that using MOBIC can result in a more stable configuration, and thus yield better performance.