Research On The Applications Of Game Theory To VANET System Construction

As an application of IoT (Internet of Things) in term of intelligent transportation, VANET (vehicle ad-hoc network) is a mobile self-organizing multi-hop wireless networks. Integral performance of VANET system is closely related to the communication performance between RSU and vehicle node. In the development of VANET, the system construction must be a gradual process and at the beginning of the construction we need to consider the following two game problems, also are the main research works of this thesis.1. Game between vehicle node and VANETFor saving their own network sources, selfish nodes in VANET will refuse to forward packets that received from other nodes, which will reduce the network throughput and power. To solve this problem, people usually add reputation-based or pricing-based incentives to making nodes cooperate. As the roadside infrastructure of VANET, RSU is with enough storage space and strong computation power. Combining the advantages of existing achievements, this thesis propose a dual-incentive mechanism, which can punish selfish behavior by reduce their reputation permissions and paying credits to cooperative nodes as reward. In VANET, the reputation permissions of vehicle nodes are divided into good and bad, which determined whether a node could participate in data communication of VANET. Except punishing selfish nodes by reduce their reputation value, this scheme will reward cooperative nodes credits which will more deeply stimulate nodes to behavior cooperation from the aspects of economic. Moreover, in order to solve the defect in most of reputation mechanism that it is excessively strict and may reduce the effectiveness of the network nodes, this thesis also introduces a reputation rebuilding model to give bad nodes the opportunity of re-acquire benign permissions. To prove the effectiveness of the mechanism, we analyze the behavior of nodes in a single data packet forwarding game from the economic perspective. The results of the theoretical analysis demonstrate that the proposed mechanism is efficient in promoting cooperation while improving the network throughput. In addition, after the vehicle nodes enter into the service area of a single RSU, the packet loss ratio will reduce close to 0 in a limited time and the packet forwarding rate of node will quickly rise close to 100% in the simulation model. The experimental results show that the proposed mechanism makes the packet-loss rate converge quickly, and the vehicle nodes exhibit extremely high forwarding rate.2. Game between RSU and VANET systemIn the scenario of multiple communication service providers to build VANET infrastructure together only with their own income as the precondition, service providers ignore the optimization of overall network environment in VANET that often generate redundant coverage of RSU. Because of the RSU have some characteristics such as higher construction cost, less communication demand for the initial, so it cannot be achieved that full coverage of RSU in road environments in the short term. For the application scenarios that it needs to reconfigure a RSU or construct a new RSU when the road surrounding environment changes, we propose a RSU optimal configuration scheme for VANET based on game theory. First, we introduce traffic fluctuation factor to adjust traffic flow function model fitting by statistical data based on location information in city road. And through introducing the effective area dividing coefficient to exact divides road area within RSU communication coverage based on economic benefits. Finally, by calculating revenue weight and revenue coefficient of the service provider to establish a single RSU configuration game, through payoff matrix in the game model analysis of the Nash equilibrium strategy set, and establish revenue of VANET systems to find the conditions and directions of system that RSU single configuration game could achieve equilibrium.