Research And Application Of Internet Of Vehicles’ Video Transmission Architecture Based On Mobile Edge Computing

With the continuous advancement of sensor technology and wireless communication technology,the Internet of Vehicles technology has achieved rapid development,and has also spawned a series of Internet of Vehicles related applications.Video applications are one of the very popular applications in the Internet of Vehicles.For example,traffic safety-related videos can solve problems such as blind spots for drivers and improve driving safety and road safety;leisure and entertainment videos can increase the entertainment of travel and enhance the users’ travel experience.Video applications often transmit large amounts of data,occupy high bandwidth,and traffic safety-related video applications also have high requirements for latency.However,the traditional video transmission based on the cloud computing framework due to a large amount of traffic back to the core network,resulting in excessively high service delay,and even unable to meet the requirements of high bandwidth and low latency for mass vehicle terminal video applications.In response to the above problems,this thesis mainly did the following research and work:(1)Proposed and implemented the Internet of Vehicles video transmission architecture based on Mobile Edge Computing(MEC).This architecture offloads tasks such as video compression,encoding,transmission,and decompression to edge nodes to provide high-quality video services for video-demanding vehicles.First,in order to accurately count the end-to-end transmission delay,a tool for time synchronization between devices has been designed and implemented,and the synchronization accuracy is actually controlled to the millisecond level.Secondly,in order to facilitate the reception and analysis of the video receiving end,each video frame is deframed at the video sending end to add custom header information,and the video frame receiving buffer is realized at the receiving end.Thirdly,it solves the problem that in the video transmission system under the cloud computing architecture,server resources cannot be recovered when data sticks and abnormally disconnects.Finally,in the indoor environment,through the video transmission comparison test with the cloud computing architecture,the effectiveness and superiority of the architecture proposed in this thesis are verified.(2)Based on the MEC Internet of Vehicles video transmission framework,two real application cases are proposed and their prototype systems are implemented.First,a vehicle to vehicle(V2V)communication-based video sharing(See-Through)application scenario is proposed.The system can expand the driving vision of rear vehicles to assist overtaking and lane change decisions and improve driving safety.Secondly,a cross-section video sharing application scenario based on Infrastructure to Vehicles(I2V)communication is proposed.The system can give drivers a global view of the current intersection to improve the safety and comfort of vehicles passing through the intersection.Through the prototype system test in the real VANET environment,it is verified that the MEC-based video transmission architecture has lower and more stable transmission delay,and is more capable of providing high-bandwidth,low-latency video services.