Research On Distributed Tasks Offloading And Communication Resource Allocation Via Vehicular Fog Computing

With the continuous development of Internet of Vehicle(Io V)and wireless communication technology,a series of computational intensive and delay-sensitive applications such as ultra-high-definition video,autonomous driving,assisted driving and real-time traffic management have emerged.These applications require network support and have strict requirements on low latency and high bandwidth to ensure the quality of service(Qo S)of the system and improve the quality of experience(Qo E)of users.Due to the limitations of resources such as battery capacity,computing and storage,mobile terminals cannot well support these emerging applications.The traditional centralized cloud computing architecture cannot meet the low latency,high bandwidth,and local processing required by massive mobile terminals and emerging applications due to excessive backhaul link load and long service response delay.Some studies have proposed the combination of Internet of Vehicles technology and Fog Computing(FC)to form Vehicular Fog Computing(VFC),which offloads computing,storage,and communication functions from the cloud data center to the edge of the wireless network.It can provide adjacent real-time computing and local processing services for mobile terminals,thereby reducing network burden,reducing response delay,and improving user experience.As a new application paradigm,VFC is promising in supporting task offloading and real-time service applications in Io V.However,since resources such as computing and storage are sinking to the edge of the wireless network,and the vehicles in the architecture are both computing service requestors and service parties,the original task offloading and wireless communication resource allocation schemes do not applicable.Accordingly,this article will conduct research on VFC related technologies,analyze the advantages of VFC and supportable application services,and further study how to design distributed task offloading and communication resource allocation algorithms in heterogeneous Io V environments.The main research contents and innovations of this thesis include:(1)Distributed task offloading based on dynamic programming via VFC architectureIn view of the problems such as limited storage and computing resources of mobile terminals and Road Side Units(RSU)in the Io V,the mobility of vehicles,the difference in the resources of VFC servers,and the low latency requirements of the tasks,we study the distributed task offloading of multi-task and multi-server.First,a two-layer distributed network architecture based on VFC is built.Both mobile vehicles and roadside infrastructures are equipped with computing,storage,and communication resources to support both local computing and task offloading.Second,we formulate a distributed task offloading problem,which the vehicle network dynamics,low-latency requirements of tasks,and resource heterogeneity of fog nodes into consideration,and aims at maximizing the overall resource utilization of system.Then,the problem is linearized into 0-1 integer linear programming(ILP)by transformation,and Lingo software used to obtain the optimal solution to the problem with high computational cost.Finally,based on sub-problem decomposition and dynamic programming algorithms,an efficient and feasible distributed task offloading algorithm is designed,which optimizes both task offloading decision and workload balancing.Experimental results show that the distributed task offloading algorithm proposed in this thesis can quickly obtain an approximately optimal offloading strategy with low computational cost,and demonstrate the effectiveness and scalability of the proposed algorithm.(2)Distributed communication resource allocation based on potential game model via vehicle fog computing architectureIn a V2V/V2I communication based VFC scenario,a distributed communication resource allocation algorithm based on potential game model is proposed to solve the problem of co-channel interference(CCI)between V2 V users and V2 I users.First,we formulate a problem to consider channel scheduling and power allocation jointly,and establish a game model that maximizes user’s payoff.It aims to minimize the total CCI among users while meeting the transmission rate requirements of each user during task offload.Second,we show that the formulated game can be mathematically described by a potential game and proved the existence of Nash equilibrium.Based on the potential game model,a distributed probability-based best response algorithm is designed,where the limited payoff and local information are used to update the user’s strategy.Finally,simulation results show that the algorithm can meet the transmission rate requirements of each user,maximize user benefits and minimize co-channel interference.