| With the improvement of mobile devices(MD)performance,the rapid development of various computation-intensive applications has been promoted,such as 3D online games,face recognition,augmented reality/virtual reality,and video transcoding.When executing these applications,in order to ensure strict latency requirements,sufficient computing resources are usually required,which consume a lot of energy.However,MD often has limited computing/storage resources and energy,and it is difficult to complete these computationintensive applications independently.Through traditional cloud computing,computing tasks can be offloaded to the cloud center for execution,but as MD is far from the cloud center,it will inevitably bring a large delay and increase energy consumption,which will affect the user's computing quality experience.At the same time,when a large number of computing tasks are offloaded to the cloud center at the same time,network congestion will also be caused.So in this context,mobile edge computing(MEC)was proposed.By installing an MEC server with sufficient computing and storage resources on base stations(BSs),wireless access points(APs),and road side units(RUSs)near mobile users,computing tasks can be directly offloaded to the MEC server for execution,thereby solving the problems of long delay,high energy consumption,and network congestion caused by cloud computing.Although MEC can effectively reduce latency,decrease energy consumption,and improve the user's computing experience,there may be interruptions when MDs with limited battery capacity offloading tasks.In order to solve this problem,wireless power transfer(WPT)is introduced in MEC systems.However,most of the existing works only consider the performance of the system from the perspective of the user and ignore the benefits of the operator.On the other hand,the security and privacy issues brought by the migration of tasks and services in the MEC server are a big challenge.The decentralized advantages of blockchain can solve these problems.However,existing research schemes that support blockchain-enabled MEC systems only optimize the performance of MEC systems and blockchain systems separately,resulting in suboptimal results.Based on the above reasons,this thesis is devoted to the research of efficient resource management technology in mobile edge computing and conducts research from two aspects: MEC system based on wireless power transmission and MEC system supporting blockchain.In the MEC system based on wireless power transfer,this thesis considers the performance of the system from the perspective of the operator,the purpose is to maximize the data utility of the operator and minimize its energy consumption at the same time.In the blockchain-enabled MEC systems,under the scenario that the MEC system and the blockchain system are performing tasks at the same time,the optimal compromise between the performance of the MEC system and the blockchain system is achieved.The main contents of this thesis include:(1)A resource allocation framework for a MEC system based on wireless power transmission is proposed,which solves the problems of maximizing data utility and minimizing the energy consumption of operators.Specifically,in order to prevent the mobile device from unloading interruption due to insufficient energy when computing tasks are unloaded,this article introduces WPT technology to supplement the energy for mobile devices.Carriers perform mobile device offload data as their revenue.In order to increase revenue,operators actively encourage mobile devices to offload more data by supplying energy to mobile devices.This article jointly optimizes the wireless power allocation at the base station side,the size of the offload data and power allocation at the user end,while ensuring the constraints of the offload delay while maximizing the operator's data utility and minimizing its energy consumption for performing the offloaded task.In order to solve this problem efficiently,we first convert the computational offload delay constraint into an offloaded data rate constraint.Then,by using the Lagrangian dual method,we propose an efficient iterative algorithm to derive the offload data size and power scheme at the user end.Finally,the results obtained are used to obtain wireless power allocation at the base station.(2)A framework for computing offload and resource allocation in a MEC system that supports the blockchain is proposed,which solves the delay/time to finality(DTF)minimization in the blockchain system and the energy in the MEC system.Minimize the problem and achieve the best compromise between the performance of the two subsystems.The blockchain is deployed on the MEC server,so the two subsystems share computing and storage resources,so there is a competition for resources between the two subsystems.Specifically,in the MEC system,the mobile device offloads the computing task to the MEC server,that is,the MEC system performs the computing offloading task,and the blockchain system needs to generate blocks and agree on the generated blocks.After reaching an agreement,add the blocks to in the blockchain,the blockchain performs block generation and consensus tasks.In chapter 3,by jointly optimizing user connection,data rate allocation,block producer scheduling,and computing resource allocation,the research problem is modeled as an optimization problem to achieve the optimal trade-off between the energy consumption of the MEC system and the DTF of the blockchain system.Since the proposed problem is a mixed-integer non-linear programming problem,in order to reduce the complexity of directly solving this problem,we designed an efficient algorithm by decoupling the optimization variables.The simulation results show the convergence of the proposed algorithm,and the proposed scheme can well achieve the compromise between the performance of the MEC system and the blockchain system.(3)A collaborative unloading and resource allocation framework that supports the blockchain MEC system is proposed,which solves the problems of maximizing the throughput in the blockchain system and maximizing the calculation rate in the MEC system.In this framework,in order to ensure the data security and privacy of nodes in the MEC system and the blockchain system,we have introduced a trust model.At the same time,the framework solves some existing challenges in existing MEC systems that support the blockchain,such as 1)joint optimization of the MEC system and the blockchain system,2)collaborative computing offloading,and 3)dynamic optimization.In this framework,by jointly optimizing the offloading decision,power allocation,block size and block interval,we propose a multi-objective optimization problem to simultaneously maximize the computation rate of the MEC system and the transaction throughput of the blockchain system.Due to the dynamic characteristics of wireless fading channels and processing queues,the proposed joint optimization problem is modeled as a Markov Decision Process(MDP).Aiming at the dynamics and complexity of the MEC system supporting the blockchain,an A3C-based collaborative computing offload and resource allocation algorithm was proposed to solve this MDP problem.Finally,the convergence of the proposed algorithm is demonstrated through simulation,and the proposed algorithm is compared with the existing schemes. |
PDF Full Text Request