| As new latency-sensitive and compute-intensive applications such as autonomous driving,big data analysis and augmented reality rapidly escalate,end devices with limited compute and power capacity cannot meet the computational demands of emerging applications in future wireless systems.Mobile Edge Computing(MEC)can assist resource-and power-constrained end devices with task processing by leveraging the computing power of edge servers near base stations.In addition,with the rapid development of Io T technology,more and more smart devices,such as smartphones and smartwatch,are emerging,which requires us to continuously update and improve the traditional Orthogonal Multiple Access(OMA)to meet the demand for massive parallel connectivity of these devices.NonOrthogonal Multiple Access(NOMA),an enabling technology for 5G,allows multiple devices to multiplex the same time-frequency resources through the power domain.By introducing an additional resource dimension,NOMA has a natural advantage in handling massive access problems.Therefore,this thesis focuses on the task offloading and resource optimization problems in MEC systems in conjunction with NOMA technology.The research content and innovative results of this thesis are as follows:(1)For a single MEC server user task offloading scenario,a delay minimization problem is investigated through the joint optimization of computational resource allocation,transmit power control,sub-channel assignment,and offloading strategy in order to improve the performance of system users in terms of delay.The formulated problem is a Mixed Integer Nonlinear Programming(MINLP)problem,which is difficult to solve directly.To make the problem easy to handle,first,the problem is decoupled into a computational resource allocation subproblem and a joint subchannel allocation and transmit power control subproblem with a fixed offloading decision.Then,a convex optimization technique is used to solve the computational resource allocation subproblem,and an efficient solution is proposed to solve the remaining joint subproblem based on an improved fireworks algorithm.Finally,a low-complexity algorithm is designed to solve the offloading decision problem.Simulation results show that the proposed algorithm has a faster convergence speed and better stability,and the proposed unloading scheme has better delay performance compared with other benchmark schemes in this thesis.(2)The mobile edge computing offloading and resource allocation problem in NOMA heterogeneous networks is studied for a multi-MEC server user task offloading scenario.User offloading benefits are maximized by jointly optimizing computational resource allocation,subchannel allocation,power control,and offloading strategies,which are measured as a weighted sum of the reduction in edge computing latency and energy consumption compared to local computing.For the formulated MINLP problem,this thesis obtains the solution of the original problem by decoupling it into a task offloading subproblem and a resource allocation subproblem,and then,designing an efficient method to iteratively solve the two subproblems.Specifically,the resource allocation problem is further decoupled into the computational resource allocation,joint sub-channel allocation and transmit power control problems with a fixed task offloading decision.First,the computational resource allocation problem is solved based on convex optimization techniques,and then,an efficient scheme is proposed to solve the joint sub-channel allocation and power control problems based on the improved whale optimization algorithm.Finally,a heuristic algorithm is designed to solve the offloading decision problem.The simulation results show that the proposed scheme has obvious advantages in improving the offloading efficiency of the system,and the proposed scheme obtains the best offloading efficiency compared with other benchmark offloading schemes. |
