A Fuzzy Inference based Vehicle to Vehicle Network Connectivity Model is proposed to support Optimization Routing Protocol for Vehicular Ad-hoc Network (VANET), where the real-time vehicle to vehicle network connectivity situation of road segments is expressed using fuzzy inference according to the vehicle distribution situation, and the optimized routing protocol modifies the transmission path dynamically and optimizes packet forwarding. The proposed model expresses the real-time vehicle to vehicle network connectivity of each road segment that cannot be easily expressed directly by a mathematical model and decreases the end-to-end delay and the overall network control overhead. The computation time of the proposed protocol is analyzed and shown asO(IlgI+R+V) whereI,R, andVrepresent the number of intersections on a map, the number of road segments on a map, and the number of vehicles within communication range of the vehicle that wants to transfer a data packet, respectively. The simulation tools NS2 and TraNS are used to perform experiments that include wireless data packet transmission and vehicle mobility traces. The results show that the proposed method decreases end-to-end delay and decreases the control overhead by 20% compared with other routing protocols, e.g. GyTAR and RTRP. This proposal implements an intelligent transportation system application and a traffic-monitoring system in NS2 using the optimization routing protocol. This protocol will be implemented to develop a real vehicle telematics system using the embedded system to improve the user-driving experience.
Connected model for opportunistic sensor network based on Katz centrality
