Investigation of Clustering Algorithms in a Transportation System Employing Vehicular Ad Hoc Network Technology

Abstract

The current issue of improving the quality of road traffic in VANET transport networks is considered. It is shown that one effective approach to increas-ing network scalability and connection reliability in VANET networks is the use of clustering algorithms. Clustering improves the network’s overall stability and performance by combining transport nodes into clusters under certain conditions, effectively reducing network load. The study of ABCA and MOCA algorithms aims to obtain results that confirm the general goal of clustering: grouping transport units into long-lasting clusters. The study focuses on identifying key performance indicators for the VANET transport network: cluster service life, cluster member lifetime, and the number of clusters. To assess changes in network load, statistical data on the number of clusters and transport nodes at different points in time when the transport flow density changed were collected. The results of the study were evaluated by analyzing specific metrics characterizing the operation of transport nodes in the network. The intensity of service message exchange between network nodes was analyzed to compare and assess the effectiveness of clustering algorithms. It was noted that an increase in the number of transport nodes affects the total traffic volume and the possibility of forming clusters with acceptable parameters. It was shown that the choice of algorithm depends on the project’s priorities and network characteristics at any given time in the transport situation. The practical significance of the results lies in the successful formation of stable clusters and the possibility of inter-cluster service message exchanges between transport nodes to maintain communication, thereby contributing to the effective real-time organization of traffic. Recommendations are given for the use of the MOSA and ABCA clustering algorithms in different traffic flow conditions. The MOSA algorithm enables the formation of a stable group of clusters on highways, whereas the ABCA algorithm demonstrates more stable operation in an urban environment.