Research On Task Offloading Strategies Orient To Internet Of Vehicle MEC

The emergence of mobile edge computing(MEC)solves the problems of ultra-long transmission latency and network congestion of traditional cloud computing for applications in Internet of Vehicles(Io V),laying a good foundation for Io V to achieve smarter applications.However,computation task offloading in Io V suffers from limited resources of heterogeneous edge servers and mismatch between tasks and resources.Meanwhile,the highly dynamic Io V is vulnerable to malicious attacks causing information leakage during task offloading.To address the above problems,this paper firstly proposes a digital twin and blockchain based Io V architecture to ensure efficient and secure task offloading process.Secondly,in the case that tasks can be partitioned,a task partial offloading strategy with edge cooperation is proposed to reduce the network cost while making full use of edge resources.Then,in the case that the task is not divisible,a task offloading strategy with vehicle collaboration is proposed to solve the problem of mismatch between task and vehicle resources.The specific research content and innovation points of this paper are as follows:(1)To address the problem that the existing Io V architecture ignores the dynamics and security of the network environment,this paper constructs a four-layer architecture of Io V based on digital twin and blockchain,including physical layer,blockchain layer,digital twin layer and application layer.Firstly,the state information of the physical layer in Io V is collected to build a digital twin layer that is highly similar to the physical layer,simplifying the deployment of Io V.Secondly,physical entities such as vehicles are simulated and predicted in the digital twin layer with the help of intelligent algorithms to provide intelligent decisions for task offloading in the application layer.Finally,the blockchain layer protects the data privacy of the remaining three layers by means of smart contracts and provides security for performing efficient task offloading.(2)To address the problem that the existing Io V ignores the limited resources of heterogeneous edge servers,this paper proposes a task partial offloading strategy with edge cooperation.Firstly,the limited resources of edge servers in Io V are fully utilized through edge cooperation,and a task offloading algorithm based on improved cuckoo algorithm is proposed to reasonably divide and offload subtasks to reduce task offloading cost.Secondly,a blockchain-based smart contract is used to protect the transaction information to achieve safe and reliable task offloading,and a greedy strategy-based blockchain resource allocation scheme is used to select appropriate edge server verifiers to reduce block consensus cost.The experimental results show that the strategy proposed in this paper can minimize the network cost in the whole task offloading process on the basis of reducing the offloading cost and consensus cost.(3)To address the problem that the existing Io V ignores the mismatch between tasks and vehicular resources due to multi-vehicle and multi-task,this paper proposes a task offloading strategy for vehicle collaboration.Firstly,the network capacity and privacy protection of Io V are enhanced using the space-air-ground integrated network with blockchain,which is conducive to the flexible scheduling of resources.Secondly,in order to balance the matching relationship between tasks and vehicular resources,the effect of the urgency of the tasks,the busyness of the vehicles and DT deviation on task offloading are considered comprehensively.A selection algorithm based on Circle chaotic mapping is used to select vehicles suitable for task execution,and a task allocation algorithm based on improved genetic algorithm is proposed to obtain the best task offloading decision.Experimental results show that the strategy proposed in this paper can maximize the system utility on the basis of reducing the total task execution time.