Vehicular Ad hoc Networks (VANET) used to build wireless networks between vehicles (V2V), and between vehicles to infrastructure (V2I) are a subclass of Mobile Ad hoc Networks (MANET). VANETs introduce a variety of applications to enhance the safety and create a comfortable driving environment for drivers. Connectivity in VANET is a fundamental operation that allows a vehicle to construct a source-todestination but is problematic due to the rapidly changing topologies and high speed mobility of the vehicles involved. This study presents a Predictive Geographic Routing Protocol (PGRP) to improve connectivity with predict greedy forwarding which could handle a recovery strategy for local maximum environment. In PGRP, every vehicle gives a weight to its neighbors according to the direction and location of the vehicle. Besides that, PGRP is capable of predicting the location of every vehicle in the interval time of hello packet based on the acceleration of the vehicle before forwarding the packets based on location after a short interval. An improved PGRP known as Predictive Geographic Routing Protocol with Delay-Tolerant Network (PGRP+DTN) would be used whenever a network disconnection occurs. A buffer can be used for every vehicle and each packet could be stored when a network is disconnected. To simulate the PGRP and PGRP+DTN, an integrated simulation NS2.35 based model and Simulator of Urban Mobility (SUMO) were used to generate a realistic traffic situation. The simulation results show that PGRP outperformed GPSR and GPCR in terms of packet delivery ratio, end-to-end delay, and average hops for improve connectivity between every two nodes. PGRP+DTN also outperformed Geopps, GeoDTN+NAV, and VADD in terms of packet delivery ratio, end to end delay, average hops for improve connectivity between every two nodes whenever they want to exit from DTN. The research has proven that the problems of connectivity in VANET can be resolved by using the proposed protocols.
展开▼
