An Efficient Division Method of Traffic Cell Based on Improved K-means Clustering Algorithm for the Location of Infrastructure in Vehicular Networks

With the rapid development of urban construction, the number of motor vehicles has increased sharply. Reasonable division of traffic districts is the key to ensure traffic flexibility and network connectivity. Aiming at the problems of unreasonable number of clusters and random selection of initial clustering centres in existing clustering algorithms for dividing traffic cells. This paper propose a systematic clustering combined with Genetic Algorithm-K-means algorithm for traffic cell delineation method. Our proposed algorithm is validated through clustering experiments on real UCI datasets, and its clustering advantages are demonstrated by combining clustering evaluation indicators. This paper takes Shenzhen, China as an example and divides traffic districts based on Global Positioning System (GPS) data from 127979 taxis, buses and trucks. By analyzing the group trajectories of different types of vehicles, reference is provided for the location of infrastructure in vehicular networks and the establishment of regions of interest in swarm-aware networks.