Resource Allocation And Optimization In Mobile Edge Computing

As mobile terminals are gaining enormous popularity,more applications have demands for low delay and energy consumption.In order to satisfy demand,recent years have seen a paradigm shift in mobile computing,from the centralized mobile cloud computing toward mobile edge computing(MEC).MEC provides an IT service environment and cloud computing capabilities at the edge of mobile network in close proximity to mobile terminals.However,MEC needs to spend much cost on deploying local MEC servicers.To reduce the cost,general mobile edge computing is proposed.In the general mobile edge computing architecture,edge service providers are extended,combined with mobile terminals which has idle computing resources(i.e.,edge service provider),which make general mobile edge computing become an active area of research.Due to mobility and complex network topology of mobile terminals,it is a significant challenge to schedule resource allocation of mobile terminals.In order to overcome the challenge,in our paper,we focus on the static and dynamic resource allocation of mobile terminals(i.e.,cloud service providers).The main work of this paper is listed as follows.1)We consider the scenario where parked vehicles are considered as mobile edge nodes.A parked vehicular edge computing system is proposed.The MEC server recruits the parked vehicles(PVs)located in a parking lot to help offload tasks.As the PVs' private information(e.g.,parking time)is not visible to the MEC server,an adverse selection problem arises with asymmetric information between the MEC server and the PVs.In order to encourage the PVs to participate in the system and improve their energy efficiency,a contract-based incentive mechanism is proposed to overcome the asymmetric information and address the adverse selection problem.The necessary and sufficient conditions for the feasibility of the contracts are to the proved.Numerical results show that the proposed contract-based incentive mechanism is feasible and effective.2)We consider the scenario where Public buses are considered as mobile edge nodes.We leverage the characteristics of buses and consider a scalable mobile edge computing architecture in bus networks.The on-board bus servers are motivated to accomplish the offloading computation tasks from MEC servers.By this way,the computing capability of the MEC server significantly extends.We consider an incentive strategy using genetic algorithm(GA).With the strategy,the MEC server spends the minimal cost to offload its redundant computation tasks.Meanwhile,the user experience of mobile terminals is guaranteed.The simulations validate the effectiveness of our incentive strategy.3)We consider the scenario where private vehicles and mobile edge server corporately are considered as mobile edge nodes.We design an optimized workload allocation strategy through a sequential Stackelberg game.With the sequential Stackelberg game,a MEC server,resource demander terminals,and resource provider terminals achieve an efficient coordination of the workload allocation.The sequential Stackelberg game is proven to reach two sequential Nash Equilibriums.The simulation results validate the efficiency of the optimized workload allocation strategy.