Research On Computing Offloading And Resource Allocation Algorithms In Mobile Edge Computing

With the development of the Internet of Things(Io T)and the continuous advancement of science and technology,a large number of new applications have appeared in people's lives,such as unmanned driving,smart homes,and telemedicine.The traditional cloud computing architecture can not meet the low latency,high bandwidth,and low power consumption requirements of new applications.Mobile edge computing(MEC)has become an important technical means to solve this problem.The computing offloading technology uses computing servers placed at the edge of the MEC network to help users calculate nearby,saves user task computing time,and enables the network to support application services with high latency requirement and has become an important research direction of the MEC system.Based on this,how to reasonably allocate tasks to different computing servers has become a key issue that needs to be solved in mobile edge computing.This thesis mainly studies the computing offloading of user tasks and resource allocation algorithms in MEC networks.First,for single-user MEC networks,the delay minimization and cost(total user delay and total system energy consumption)minimization algorithms when tasks have priority are studied respectively;then,the single-user circumstance is extended to a multi-user tone,and a task-priority considered cost(user delay cost and processor resource cost)minimization algorithm is proposed next.The specific research contents are as follows:1)For computing offloading in single-user MEC networks,two resource allocation algorithms are designed:a)First,the delay minimization algorithm is studied.Considering the priority relationship between user terminal tasks,the system's computing resources and communication resources are used as constraints to minimize the total user delay Model for purpose.Based on this,a hybrid genetic algorithm is proposed to solve the problem.The algorithm first obtains the cumulative probability of the selected processor according to the computing power of the processor,and then calculates the initial allocation set,finally obtains the solution of the optimization problem through the iterative selection-cross-mutation cycle.In the iterative process,the roulette decision algorithm is used to select the best allocation plan in the current set,then the crossover operation based on the tabu search algorithm is applied to improve the search ability of the algorithm,and next,the mutation operation based on the simulated annealing algorithm is used so that the overall distribution plan set has a shorter delay,and finally a task distribution plan with the smallest delay is obtained.The simulation results show that the proposed algorithm has better convergence rate and lower latency.b)On the basis of a),the cost minimization algorithm of single-user MEC network is further studied.The total user delay and total system energy consumption(user energy consumption +MEC server energy consumption + MCC server energy consumption)is taking as the optimization goal to establish an optimization model,and a hybrid particle swarm algorithm is proposed to solve the model.Different from the traditional particle swarm algorithm,the algorithm uses nonlinear weights to update the particle velocity,and uses the simulated annealing algorithm when iteratively updating the particle position and velocity to improve the search ability of the algorithm.The simulation results show that in the single-user MEC network scenario,the algorithm can obtain a task allocation scheme with lower total user delay and lower total system energy consumption.2)For computing offloading in multi-user MEC networks,considering multiple users with multiple sub-tasks,and the sub-tasks with mutual priorities,with the constraints that server computing resources and communication resources between MEC servers and mobile cloud server,an optimization model is established with the goal of minimizing the user's total delay and the total processor resource cost.An improved gravity search algorithm which is based on the idea of the law of universal gravitation is proposed to solve the optimization problem.In the algorithm,the crossover operation of genetic algorithm is used to obtain a more optimal set of feasible solutions,and the convergence factor is used when calculating the resultant force,which improves the algorithm's performance.The search capability ultimately minimizes the users' total delay and the total processor resource cost.The simulation results show that the proposed algorithm is significantly better than the existing algorithms in terms of the final optimization results.