| VANET (Vehicular Ad Hoc Networks),which is composed of cars equipped with a wireless network interface devices,is a special MANET(Mobile Ad hoc Networks). Nowdays, VANET becomes hotspots of ITS field. The high mobility of vehicle will result in the frequent changes of network topology, it is especially true in certain VANET scenarios in which buildings and distances discontinue communication channels frequently, and where the available time for connecting to vehicles could be really short. There exists huge challenge in the VANET research all the time, especially in reliable file transfer application. It has always been the focus of research on how to make information deliver timely and reliably in the vehicular environments.We firstly introduce the origin, structure and characters of VANET. According to poor communications environment and low efficiency of P2P information transmission in VANET, We propose the optimization problem of VDTP, and realize VDTP in NS-2.After give a detailed analysis of its implementation. We have defined the OFPC (Optimal File Transfer Protocol Configuration) problem, which lies in searching efficient parameters setting of VDTP that maximizes the amount of data being transferred and minimizes the transmission time as the number of lost packets.Then, we establish the optimization model of VDTP protocol based on the strategy of heuristic optimization algorithm combined with VANET simulation software, and define the search space and objective function.In this paper, Particle Swarm Optimization (PSO) and Simulated Annealing (SA) are used to find the best possible VDTP configuration. And aiming at that the PSO is easy to fall into local extremum, evolution convergence slow, and poor precision, we put forward an improved particle swarm optimization algorithm based on simulated annealing (SAPSO). The proposed algorithm has paid attention to both the advantages of PSO and SA, it has global search ability,the simple realization and has a stronger ability to jump out of local optimal solution, thus it avoid the problem that PSO is easy to fall into loacal extremum.At last, two typical car-to-car environment instances have been defined:urban scenarios and highway scenarios. We rely on a flexible simulation structure using MOVE/Ns-2. These scenarios simulate the vehicular node’s mobility behavior seemingly. In turn, we have added a new functionality to ns-2, which it is used to return the metrics needed for evaluating the fitness function of a given configuration. Then, we import the generated mobility trajectory into NS-2.Through simulation The SAPSO is proved to be more stable than other algorithms and can help improving the transmission efficiency of P2P information in VANET. |
