Research On Task Offloading And Multi-Dimensional Resource Management In Mobile Edge Computing

With the rapid developments of mobile communication technologies and internet services,new intelligent applications such as virtual reality,augmented reality,autonomous driving,and telemedicine are rapidly popularized.These applications are generally computation-intensive and delay-sensitive,i.e.,they require powerful computing capacities and extremely low delay requirements.However,terminal devices as the major application supporting platform,have limited battery capacities and computing abilities,thus cannot meet the requirements of these applications.By deploying high performance servers at the edge of wireless networks(e.g.,base stations(BSs),access points(APs),etc.)to provide computing services for the users,mobile edge computing(MEC)has been considered as a promising technology to tackle the above challenges.With MEC,users can offload their computing tasks to edge servers for execution,thereby achieving low-latency services and saving energy consumption.However,as the number of various terminal devices increases dramatically,the existing spectrum resources are insufficient to support large-scale user access.Non-orthogonal multiple access(NOMA)has been regarded as a key technology to improve spectrum efficiency and accommodate massive connectivity in future wireless networks.In order to improve the access capability of MEC networks and guarantee the service demands of massive users,therefore,this paper first considers a NOMA-based heterogeneous MEC network.Aiming at the problems of the severe interference caused by the multiple offloading users and the limited computation resources of the edge server,a joint task offloading and resource allocation approach is proposed to minimize the system energy consumption while guaranteeing the quality of service of all users.Offloading decision,power control,local CPU frequency,computation resource and subchannel resource allocation are jointly considered in the proposed approach.The simulation results prove that the proposed approach can effectively reduce system energy consumption and meet the task delay requirement.In addition,to serve the users in remote areas or cell edges,this paper further studies the deployment of edge servers on unmanned aerial vehicles(UAVs)to achieve more flexible access.Considering that executing users' tasks on edge servers requires the corresponding computing programs,this paper studies the service caching-based task offloading and resource management in the multi-UAV assisted MEC network.In order to minimize task completion delay and meet the energy consumption requirements of the users and the UAVs,a joint service caching,task offloading,computing resource,spectrum resource,and UAV location scheduling optimization problem is formulated,and an iterative-based distributed algorithm is then proposed to solve the problem.The simulation results show that the proposed algorithm can lower system latency while guaranteeing the quality of service of all users.