Research On Key Technologies Of Distributed MAC In V2X Communications

The Ultra-Reliable Low-Latency Communication(URLLC)technology of the Vehicle to Everything(V2X)network is significant to security applications in intelligent transportation systems.The 3rd Generation Partnership Project(3GPP)specifies Quality of Service(Qo S)requirements with higher priority for delay-sensitive communication scenarios.Moreover,the open network architecture of 5G/6G systems defines new air interface,namely,direct communication to support new delay-sensitive services.Therefore,it is significant to explore new distributed Media Access Control(MAC)technology for promoting the efficient radio resources uses and diverse wireless accesses in the future V2X network.However,variable factors such as dynamic network topology,vehicle mobility and diverse services cause challenges to V2X communications.Firstly,the frequently changing network topology would intensifie the instability of the communication link between vehicles,causing serious radio resource contention,packet collisions and access performance deterioration.Secondly,in the complex V2X environment,the unpredictable driving behaviors of vehicle nodes may cause problems such as link differences,unequal communication opportunities and unfair use of network resources.Thirdly,due to the lack of the precise link information of each vehicle node in the high dynamic V2X network,it would be difficult to coordinate the contention access between the vehicles nodes with the inefficient radio resource allocation.Moreover,the strict Qo S requirements on V2X security services may put more attention on URLLC qualities,but the radio resource optimization theory is classically used to optimize throughput.In addition,in the open distributed network environment,the eavesdropping vehicles may capture the message transmissions of the legitimate vehicles and interfere the normal communication,e.g.in the security message transmissions,causing wireless link security problems.In short,affected by the above factors,it is still urgent to overcome the dynamic limitations of the MAC strategies when solving the resource allocation problems in the V2X networks in order to enable the dynamic efficiency of the radio resource access.Facing with the challenges on distributed dynamic networking and URLLC quality of V2X networks,this thesis mainly explores the distributed MAC in order to solve the key scientific issues on the dynamic radio resource allocation and optimization methods with perfoemance improvements on scalability,fairness and link security in the complex and dynamic network enviroment.The main contributions and innovation achievements of this thesis include:1)In view of the dynamic scalability issue of V2X communication quality with limited link quality information,an efficient Bidirectional Backoff-State Transition(BBST)model is firstly proposed.Next,this thesis establishes the BBST two-dimensional Markov chain extension model based on Bianchi model with the theoretical analysis,derives the accurate transmit probability and the theoretical normalized average throughput expression of the BBST expansion model,and gives the closed analytic formula of transmit probability's relationship with link supervision factor and matching avoidance factor.Furthermore,a novel adaptive Collision Supervision and Avoidance algorithm(CSA)is proposed based on the theoretical analysis.Both Link Supervision(LS)and Matching Backoff(MB)mechanisms enable nodes to dynamically perceive changes of link quality within a certain historical transmission period,and enable the Matching Backoff Stage(MBS)to be dynamically consistent with changes of link quality,realizing the adaptive transit probability.The simulation results indicate that the dynamic performance of CSA is surprior to that of adaptive ECA.Even in the 100km/h and 100veh/km high-mobility and high-density environment,CSA can achieve the 8%throughput improvement,Packet Delivery Ratio(PDR)>0.86 and access delay<6.5ms.CSA improves the scalability of dynamic V2X network and more effectively adapts to the dynamic application scenarios of V2X network.2)In view of the radio resource access fairness issue of V2X communication with no link quality information,an intelligent BBST vehicle agent model is firstly proposed according to Markov Decision Process(MDP)model,and a Q-learning based Collision Avoidance(QCA)algorithm is designed by using the Reinforcement Learning(RL)theory.QCA introduces BBST bidirectional backoff reward model into Q function Bellman equation,motivates vehicle agents with different backoff states to optimize their access priorities,and improves the adjustment efficiency of transmit probability.Through the optimal strategy obtained by solving the problem of the maximum value Q_max,the vehicle agent can take corresponding action to transit from current backoff state to best backoff state BSI_opt.Finally,the?-greedy algorithm is used to solve the maximum value Q_max and the convergence of the algorithm is proved.Through self-optimization and intelligent update of vehicle agents,QCA improves the Jain Fairness Index(JFI)of V2X communications.The simulation results indicate that QCA can obviously improve JFI fairness,PDR and delay in the dynamic V2X environment such as vehicle density,vehicle speed,packet payload and date rate.Under the impact of packet payload,the saturated JFI of QCA reaches 0.7,which is75%higher than QLMAC.Therefore,QCA can improve fairness within URLLC and adapt to diverse service application requirements of the V2X network.3)In view of the link security problem that normal messages may be eavesdropped by malicious vehicle nodes in the open wireless V2X environment,this thesis proposes a V2X low-complexity link-security multi-antenna transmission scheme AN-STSK based on secrecy capacity theory,and adopts low-complexity Maximum Likelihood(ML)receiver.Through using Artificial Noise(AN)in the transmit signal to impact the malicious channel mutual information,the Signal-to-Noise Ratio(SNR)of receiving signal of the malicious node decreases,therefore suppressing the receiving performance of the malicious node and improving the link security of V2X networks.Moreover,the secrecy capacity of the AN-STSK system is derived theoretically,obtaining the influence of the power proportion of AN on both Bit Error Ratio(BER)and secrecy capacity.Further,the MAC wireless link security control frame is designed to adjust the power proportion of AN,efficiently suppressing the block ratio of malicious communications.The simulation results indicate that MAC can make a good balance between BER and the secrecy capacity by optimizing the power proportion of AN,providing a basis for the trade-off design between V2X communication reliability and link security.In summary,in view of dynamic complex V2X environment factors such as high vehicle speed,high vehicle density,high service payload and variable data rate,this thesis efficiently solves the problems of dynamic distributed radio resource allocation,and proposes the scalable,fair,security and intelligent distributed MAC key technology in V2X networks.The obtained related theoretical innovation results and protocol algorithms can lay a scientific theoretical foundation for the development of the high-quality V2X communication technology,has good reference value of the application design,and has good application prospects for a variety of V2X wireless communication fields.