A centrality based random walk approach for topology formation in networks

Abstract

Wireless network topology is highly dynamic and effective computation of topology plays a major role in improving network performance. Random walk based approaches are very useful to find packet delivery subgraphs in such networks. This paper proposes a novel random walk strategy based on centrality of the nodes to compute the topology. To improve the speed of computation, proposed a parallel random walk in which agents computes partial paths and stitches it to obtain the complete topology of communication. The results shows that the proposed approaches can find the topology with lesser path lengths compared to other random walk approaches.