A Vehicular Ad Hoc Networks Clustering Algorithm based on Position-Competition

Qianjin He, Tao Yang
2017 Proceedings of The 7th International Conference on Computer Engineering and Networks — PoS(CENet2017)   unpublished
The vehicular ad hoc networks(VANETs) presents uneven node density, fast moving speed and dynamic topology change etc. In order to improve the reliability of VANETs in information broadcasting and reduce the redundancy of multi-hop broadcast, a protocol based on the clustering position-competition of vehicular ad hoc networks is hereby proposed. Furthermore, the concept of connectivity-stability is introduced. In the paper, the road is divided into the segment model and the intersection model.
more » ... he cluster head and the optimal nodes in the cluster are chosen as a radio relay node, which may reduce the broadcast redundancy and improve the broadcast efficiency. At the same time, for the isolated nodes in the network, the carryforwarding method is taken, which may improve the broadcast reliability in sparse areas and in "hole" areas. Finally, we use the NS2 simulation tool for analysis. The simulation results show that the proposed routing protocol features better comprehensive performance on successful delivery rate, average broadcast time delay and broadcast overhead. A Vehicular Ad Hoc Networks Clustering Algorithm based on Position-Competition Qianjin He 1.Introduction Vehicular ad hoc networks (VANETs), as a special kind of mobile ad-hoc network, will have a promising prospect in the future intelligent transportation system (ITS); however, the high nodes mobility, the frequently changed network topology and the unstable communication link in VANETs motivate the design of stable clustering algorithms. Much has been done on this topic and many methods have been explored. VANETs are subject to link dis-connectivity due to high mobility of vehicles and low density in rural areas [1] . At present, many clustering algorithms in VANETs stem from the MANETs. The typical one is the lowest-ID algorithm, in which every node has a unique ID and the node with lowest ID will be selected as the cluster head. According to the way how the nodes are grouped into clusters and how the cluster head is selected, the clustering algorithms can be classified into five varieties: the weight-based clustering[ 2], the mobility-aware clustering[3], the utility function based clustering[4], the traffic flow-based clustering [5] and the combine-metrics-based clustering [6] .
doi:10.22323/1.299.0027 fatcat:w3owu6brrzao7efimnnbyamv4u