Shortest Average Routing Path-Based d-hop Clustering in wireless sensor networks

In the paper, we propose a new d-hop Clustering method for a clustering-based multi-hop routing scheme in large-scale wireless sensor network. d-hop clustering means that each cluster contains all nodes that are at distance at most d-hops from the clusterhead, so that the number of clusters can be getting smaller to make it possible to guarantee the combined system performances including end-to-end delay and throughput. To determine the clusterheads, we give the constraint of minimizing the expected average hop distance from a node to its neatest sink in order to minimize the energy consumption for data forwarding in the clustering-based routing scheme. To deal with the high complexity of the problem, we relax it to a special form of Minimum Weighted d-hop Dominating Set Problem(MWdDSP), Then we present two approximation algorithms and corresponding analyses for it. One is a greedy algorithm with approximation ratio O(lnn), another achieves a constant factor performance ratio.