Study of broadcasting flow structure in VANET

The paper provides analysis of message flow formation in a decentralized self-organizing vehicular network. An analytical description of the aggregated flow formation is carried out based on Burke's theorem. A variant of flow parameter description based on the sieved stream theory is presented. Next, the aggregated flow structure in the network and the principles of its formation are evaluated using the NS-2 tool. Using customized Perl scripts developed for analysis of output statistics, we have established that the aggregated flow has a Poisson distribution for low vehicle density and a Gamma distribution for high vehicle density. As a consequence of high intensity of messages in the system, the network resource becomes overloaded, which leads to degradation of the functionality of the road safety system as a whole. Based on estimations of the useful channel load, a novel message retransmission protocol was proposed to increase the efficiency of network resource utilization.