Optimization Study Of CAN-FD Based Signal Transmission And Network Management

In order to meet the needs of the intelligent and network development of modern cars,the number of various sensors and controllers inside the new generation of high-end models such as Stellar and Aito M5 is increasing,which causes a sudden increase in the amount of data to be transmitted between the controllers in the car,resulting in a higher pressure on the in-vehicle communication network,which in turn leads to data transmission congestion,frequent error frames and easy attacks on the network system of the in-vehicle communication network.If not solved in time,it will affect driving experience and even cause driving accidents in serious cases.Based on this,this paper takes the Controller Area Network with Flexible Data-Rate(CAN-FD)bus as the research object,and analyzes the real-time,accuracy and security problems encountered in the actual application of the bus to establish a more efficient,accurate and secure data transmission mechanism.In the end,it will provide a more stable,practical and reasonable solution to the problem of in-vehicle communication network.The main research contents are:(1)Analyze and summarize the current status of vehicle network research,study the functional characteristics and related technologies of CAN-FD bus used by current vehicle enterprises,and introduce the security mechanism of CAN-FD vehicle bus for the situation that CAN-FD adopts broadcast message transmission mechanism,lacks security protection and is vulnerable to network attacks,and establish the security constraint model according to the vehicle network requirements.(2)Under certain security constraints,a signal packing algorithm based on security considerations is proposed to optimize the signal transmission method as a way to reduce the load rate.The main principle is to address the shortcoming that the classical genetic algorithm is particularly prone to fall into the local optimal solution prematurely,and a hybrid genetic algorithm of the proposed Newton method is introduced to optimize the signal transmission,i.e.,a search technique is added to the genetic algorithm,and through a preset selection mechanism,a selective The Newton-like method is used to accelerate the convergence of the algorithm,so that the algorithm converges to the optimal solution as soon as possible.Subsequently,by comparing the optimization results of each algorithm under the same test conditions,it can be learned that the optimization performance of the hybrid genetic algorithm is better than that of the classical genetic algorithm.(3)In order to reduce the network management load caused by the proliferation of Electronic Control Units(ECUs),AUTOSAR local network management is also introduced to further reduce the load rate under the above established security mechanism.The principle is to dynamically adjust the communication flag bits of the node application based on the real-time bus load rate and the specified security level,and then approve the communication demand of the application based on the flag bits to reduce the load rate caused by the network management while solving the problem of insufficient battery life faced by the current vehicle companies.Based on the above optimization scheme,this paper designs a CANFD based signal transmission and network management optimization scheme and conducts related experiments.The signal transmission optimization experiments show that the use of genetic algorithm can effectively reduce the bandwidth load rate,and as the number of signals increases,both genetic algorithm and hybrid genetic algorithm gradually increase the optimization efficiency compared with the conventional case without using optimization algorithm,reducing the load rate by However,compared with the single genetic algorithm,the hybrid genetic algorithm accelerates the iterative speed while the optimization efficiency is also better than the single genetic algorithm,the maximum difference between the two reduced load rate of 4.1%,the optimization efficiency of 42%.Meanwhile,the experimental results of the network management optimization scheme also show that the optimization efficiency of the local network management optimization scheme reaches 66.39% for the load rate optimization of the classical AUTOSAR network management when the vehicle is in the dormant state,and 20.04% when the bus data is in stable communication.