Near-optimal, reliable and self-organizing hierarchical topology in VANET

In this paper we present the Distributed Construct Underlying Topology (D-CUT) algorithm, a self-organized algorithm aim to provide efficient, and reliable hierarchical topology by minimizing the interference between network participants. The D-CUT algorithm produces a geographically optimized clustering of the network, by grouping dense and consecutive nodes into clusters which are separated by maximally possible gaps. This type of clustering allows strong connections between cluster members and reduces the inter cluster interference. In addition, the awareness of each vehicle to the composition of its own cluster, and the location of its members enables a very intuitive and straightforward scheme for organizing and coordinating the vehicular network to achieve congestion control and efficient medium access performance. We show theoretically provable bounds for the algorithm performance as well as simulation study that supports our results.