Joint Optimization Of Communication And Computation Resources In Wirelless Network

With the rapid development of wireless communication technology,mobile internet has already come into our life silently,we begin to enjoy the convenience,timeliness and accuracy that the mobile internet brings.We can access to the internet by variety of technologies to enjoy the services provided anytime and anywhere.Moreover,we can find what we want easily based on the big data technology.However,mobile network is limited by the processing ability and battery capacity of mobile equipment,while which can be solved by mobile cloud computing.By this new technology,the computation tasks of mobile equipment can be offloaded to the cloud to save energy and enhance performance.Furthermore,with the evolution of vehicle network,the vehicles can communicate with each other by the advanced technologies and share their multiple kinds of resources.In order to utilize the resources fully,e.g.,communication,computation and storage resources,mobile cloud computing technology is applied in the vehicle network.The vehicle user can be regarded as two kinds of character.One of them is computation resource provider,the mobile terminals can offload their computation tasks to the vehicle to save the energy and enhance the processing capability.The other is computation resource requester,the vehicle can offload its computation task to the other vehicles or other cloud computing providers.In this thesis,we provide several resource allocation schemes under the two specific scenes above which are aimed at different services,the proposed allocation schemes can maximize the corresponding objective function effectively.The thesis firstly focuses on the computation resource allocation problem in vehicle cloud computing system.The vehicle cloud computing system is composed with a vehicle cloud and a remote cloud.The mobile terminals can offload their computation tasks to each one of them to save energy and improve the processing capability.In addition,when the computation task is allocated to the remote cloud,in order to save the energy of mobile terminals,it will be sent to the vehicle cloud firstly and then be transmitted to the remote cloud by the vehicle cloud.In this thesis,we propose a computation resource allocation scheme to maximize the long-term expected discounted reward.The allocation problem is modeled by semi-Markov dynamic decision process(SMDP).The numerical results show that the performance of the proposed scheme is better than the Greedy algorithm scheme and the Simulated Annealing scheme.Moreover,in this,a cloud-assisted network architecture is proposed,under which the system can schedule all the resources in each resource pool.Therefore,the communication and computation co-optimization problem under this architecture is what we focuses on.This architecture includes a mobile cloud which is composed of mobile vehicles or terminals,a local cloud which is deployed next to access points,and a traditional remote cloud.The vehicle offloading request can be accessed to any layer of the architecture based on the wireless channel state and the idle resources in each cloud.Moreover,the system should allocate the corresponding amount of resources to the request.The resource allocation problem is modeled by semi-Markov dynamic process and then the SMDP problem is uniformly transformed to normal MDP problem.The numerical results prove that the proposed scheme is better than the Greedy one and the other heuristic ones.