Research On Performance Optimization Technology Of Connected Vehicle Network Based On Game Theory

As an important component of smart transportation and smart cities,the Internet of Vehicles(Io V)faces challenges such as dynamic changes in network topology,rapid channel fading and limited airtime and frequency resources.Vehicle-to-Infrastructure(V2I)and Vehicle-to-Vehicle(V2V)communication methods impose demanding requirements on the timeliness of Io V data and the energy efficiency of mobile communication devices connected to the Io V.To this end,this thesis introduces a new performance metric,Age of Information(Ao I),in order to achieve an optimal compromise between the timeliness of data transmission and the energy efficiency of mobile devices in the Telematics network.In order to solve the problem that it is difficult to predict the behavioural characteristics and preferred strategies of each vehicle node in the Telematics network and to make conflicting policy choices,this paper,based on the theory of non-cooperative game theory,treats different vehicle nodes in the Telematics network as different game participants and sets the parameters such as the number of transmission slots per vehicle in the V2 I network and the transmission probability per vehicle in the V2 V network as the strategy of the non-cooperative game.The number of transmission time slots per vehicle and the probability of transmission per vehicle of V2 V network are set as the strategy of the non-cooperative game,so that the competitive conflict behaviour between different vehicles can be effectively portrayed and each vehicle can make the best response strategy according to the complete or incomplete network information it has and its own demand.For V2 I communication networks,this paper adopts a polling communication method to coordinate the data transmission order between each vehicle node and the base station,while allowing each node to send data to the base station multiple times in succession to effectively increase the chance of successful data transmission.Based on the Ao I and energy efficiency of each vehicle node,this paper designs a full information dynamic game model,rigorously proves the existence of sub-game Nash equilibrium in the model,and proposes two algorithms to solve the sub-game Nash equilibrium,such as forward game iteration and reverse game pruning.The simulation analysis of the algorithms based on MATLAB platform shows that the results obtained by both forward and inverse game algorithms are consistent with the simulation results,and the performance of the forward game algorithm achieves better overall network utility than the inverse game algorithm with higher computational complexity.To address the possible conflicts under V2 V communication networks,this paper designs a pure strategy-based non-complete information static game model based on random competitive access and rigorously proves that only weakly dominant Nash equilibria exist in this model.By transforming this game model into a Bayesian game model based on mixed strategies through a Hessiani transformation,we further prove the existence of a Bayesian Nash equilibrium for each node based on a uniformly distributed selection of sending probabilities,and derive a closed-form solution to the Nash equilibrium for each node.Simulations based on the MATLAB platform verify that the utility function of this Bayesian equilibrium is consistent with the simulation results and can effectively approximate the perfect information equilibrium result in the ideal state.