Research On Optimization Of Spectrum Allocation Strategy In Cognitive Vehicular Network

The Vehicular ad hoc network(VANET)is an important part of the intelligent transportation system.The increasing demand for various applications of the vehicular network leads to the spectrum resources in the vehicular network become more scarce.Introducing cognitive radio technology into the VANET can effectively alleviate the problem of insufficient spectrum resources in the vehicular network.Rational and efficient spectrum allocation strategy in Cognitive Radio for Vehicular Ad hoc Network(CR-VANET)has great application prospects for improving road safety,user travel experience and traffic efficiency.However,it also faces with challenges,such as fast moving speed of vehicles,short duration of communication links,and different QoS requirements of diversified vehicular services.This study takes CR-VANET as the research background,aiming at the problem of spectrum resource scarcity and the traditional cognitive wireless network spectrum allocations are difficult to fully adapt to the VANET environment,on the basis of considering vehicle mobility,the CR-VANET spectrum allocation strategies are studied,the main research contents are as follows:(1)The model construction of CR-VANET and the application of classical algorithms to solve the spectrum allocation problem in CR-VANET.The influence of vehicle mobility on the performance of VANET is analyzed,then the correlation between vehicle mobility and spectrum allocation is defined and the vehicular network is modeled from two aspects: application scenario and vehicle movement model.The classical color sensitivity graph coloring(CSGC)algorithm and the game theory algorithm are respectively analyzed theoretically,and the two algorithms are used to simulate the spectrum allocation of VANET.The CSGC algorithm focuses on maximizing the benefit of the final objective function,while the game theory algorithm focuses on the fairness of vehicular users.Based on the analysis of these two classical algorithms,this paper studies the spectrum allocation algorithms for the VANET for the expressway scenario.(2)Research on spectrum allocation strategy optimization in CR-VANET based on an improved cuckoo search algorithm.Considering the available time of the Primary Background User(PBU)and the required duration of the Vehicle User Request(VUR)in the heterogeneous environment,an Improved Adaptive Binary Cuckoo Search(IABCS)algorithm is proposed by combing the graph theory model and an improved cuckoo search algorithm,which optimizes the throughput of VANET.The characteristics of 4G spectrum and TV white space(TVWS)are analyzed,and then a heterogeneous cognitive vehicle network system combining roadside control unit,4G cellular base station and TV base station is proposed.Based on this,the PBU channel usage model is constructed to optimize the target spectrum allocation model.Considering that the feasible distribution solution that ultimately satisfies the constraint is not unique,it will lead to the transformation of the target problem into a NP problem.So a fast convergence IABCS optimization algorithm is proposed.The IABCS algorithm uses the adaptive step size control factor to enhance the global search ability of the algorithm in the Levy flight process.The improvement rate of the solution is introduced as a feedback factor in the random preference walk process.The discovery probability is adjusted dynamically and the variable neighborhood search method is used to strengthen the local search ability of the algorithm,in addition,the algorithm is combined with the simplex method to accelerate the convergence speed.The simulations of the IABCS algorithm with the standard CS algorithm,ASCSA algorithm and the SACS algorithm prove that the IABCS algorithm can improve the VANET throughput effectively and has better convergence performance.(3)Considering the different QoS requirements of diverse vehicular services,a Vehicular Service Differentiation Spectrum Allocation(VSDSA)algorithm is proposed.The priority classification system is established for different types of vehicular services,and the demand differences of QoS attributes such as throughput,packet loss rate and delay of different vehicular services are analyzed.The preference weight corresponding to QoS attributes of the three types of vehicular services is calculated with the analytic hierarchy process,and the service model of vehicle nodes is constructed.Quantitatively analyze the satisfaction degree of the residual usable time of the channel to the actual transmission time of the vehicle users in the channel,establish the benefit matrix of the expected throughput and packet loss rate,and the fuzzy comprehensive judgment is used to conduct normalization processing,so as to obtain a comprehensive target benefit.The simulation results of the VSDSA algorithm with the GAS algorithm show that the VSDSA algorithm can provide better services for vehicular users with different priorities according to different spectrum resource conditions while improving the efficiency of spectrum allocation.The packet loss rate requirements of vehicular services for emergency security and real-time traffic and the demand for throughput of vehicular entertainment services both got a high degree of satisfaction under the condition that the spectrum resources are relatively sufficient.