Research On Task Offloading And Data Distribution For Vehicular Networks With Edge Computing

With the popularity of Internet of Things devices and the development of intelligent transportation systems,the challenge of meeting the quality-of-service requirements of vehicular users is becoming increasingly prominent.Vehicular Edge Computing(VEC)is a promising computing paradigm that migrates computing and storage resources to the proximity of vehicular users,thereby reducing service latency and communication overhead.However,due to the geographical dispersion of edge computing resources and the high dynamics of vehicular networks,service provisioning in VEC is a great challenge.In this thesis,the task offloading and data distribution technologies of VEC are deeply studied,aiming to improve the performance of computation-intensive and data-intensive applications.The main contributions of this thesis are as follows:(1)A task offloading strategy based on RSU collaborative edge computing is proposed for the scenario where Road Side Units(RSUs)provide computing services.This thesis designs the network architecture and task offloading process for RSU collaborative edge computing.The effect of execution position and execution order on task completion time and task execution cost is analyzed,and the task offloading utility model is constructed.The optimization model is established with the goal of maximizing the task offloading utility.This thesis proposes a hybrid intelligent optimization algorithm based on partheno-genetic algorithm and heuristic rules to solve the model.The simulation results show that,compared with the existing strategies,the task offloading strategy based on RSU collaborative edge computing improves the offloading utility by about 15%.(2)A task offloading strategy based on vehicle collaborative edge computing is proposed to address the provisioning problem of highly available computing services.This thesis designs the network architecture and task offloading process for vehicle collaborative edge computing.The effect of communication resource allocation and computing resource allocation on task completion time is analyzed,and the task completion time model is constructed.The optimization model is established with the goal of minimizing the average completion time and maximizing the successful completion rate.This thesis proposes a multi-objective swarm intelligent optimization algorithm based on bat algorithm and non-dominated sorting to solve the model.The simulation results show that,compared with the existing strategies,the task offloading strategy based on vehicle collaborative edge computing effectively reduces the task completion time,and also improves the task completion rate.(3)A multicast data distribution mechanism based on D2 D cooperation is proposed to attain high throughput of multimedia data transmission.This thesis integrates multicast technology and relay technology in the edge computing environment,and designs the network architecture and communication process for D2 D cooperative multicast.The relay channel capacity model is constructed by considering factors such as vehicle mobility and channel quality.The optimization model is established with the goal of maximizing the minimum relay channel capacity.This thesis proposes a heuristic algorithm based on clustering to solve the model.The simulation results show that,compared with the existing mechanisms,the multicast data distribution mechanism based on D2 D cooperation effectively improves the throughput of multimedia data transmission.(4)A scalable end-to-end header compression mechanism is proposed to improve the transmission efficiency of multimedia data.This thesis integrates header compression technology and software-defined networking technology in the edge computing environment,and designs the overall architecture of the proposed mechanism based on the idea of separating header compression and packet forwarding.The corresponding compressor architecture and controller management function are designed.The performance of the proposed mechanism in terms of processing capacity,scalability,bandwidth saving,and end-to-end delay are analyzed.The simulation results show that,compared with the existing mechanisms,the scalable end-to-end header compression mechanism effectively improves the transmission efficiency of multimedia data,and reduces the end-to-end delay.