On the higher order statistics of car clustering in vehicle communications networks on a road

Abstract

This paper is devoted to the investigation of different properties of dynamic vehicular ad hoc network on a road. In our model each moving vehicle on the road communicates with several neighbouring cars. We derive important characteristics of such network including distributions of number of clusters, cluster size, biggest cluster size distribution and distribution of cars not being able to communicate with any other car in the network as well as probability of graph being fully connected. Understanding of clustering is an important issue in management of virtual cells organization, distributed data collection and processing. One of advantages of the considered model is that it can be used for arbitrary intervehicle distribution model.