Research On Multi-user Computation Offloading Methods Based On Edge Computing

With the development of smart phones,more and more new applications such as face recognition have emerged.Such mobile applications are often resource-hungry,require a lot of computing resources,and will generate higher energy consumption.However,due to the physical size limitations of mobile devices,their computing resources and battery life are limited.The contradictory relationship between resource-hungry applications and resource-constrained mobile devices poses a big challenge to the development of mobile platforms.Therefore,people have proposed the concept of edge cloud computing.It provides cloud computing in the vicinity of mobile users to enhance the computing capacity of mobile devices,and it saves processing delays and energy consumption of the mobile devices.However,there are few studies on computation offloading from the perspective of multi-user multi-radio,and the profits of users and operators.Multi-ratio offloading can greatly improve the transmission efficiency of users.Besides,studying the multi-user and multi-operator profits problem can effectively improve their profits.This paper studies the multi-user computation offloading problems from above two perspectives.(1)A multi-user computation offloading method for resource optimization is studied.And it is considered under a multi-user multi-radio scenario,with the goal of maximizing the number of beneficial users by optimizing user decisions and transmission resources.First,we model it as a non-convex mixed integer nonlinear programming problem.Second,it is transformed into an equivalent bilinear programming problem.Then,the McCormick envelope method is used to relax it into a mixed integer linear programming problem and the optimal solutions can then be obtained.Finally,we use the branch and bound algorithm to solve it.(2)A multi-user computing offloading method for economic profits is studied.And it is considered under a multi-user multi-operator scenario,with the goal of minimizing operators loss and users computing cost,while optimizing computing prices and user decisions.First,we consider the profits of operators and users and model them as a bi-level optimization problem.The upper-level operator profits problem is a continuous linear programming problem,and the lower-level user decisions problem is an integer linear programming problem.Secondly,the underlying integer problem is transformed into an equivalentcontinuous linear programming problem.Then,the KKT condition is used to replace the lower level problem,and the bi-level optimization problem is thus transformed into a single-level optimization problem.Finally,we solve the problem by using the spatial branch and bound algorithm,which can obtain the ? optimal solution.(3)In this paper,experimental verification and analysis are carried out.The experimental results corresponding to the resource optimization model show that the proposed multi-radio communication model can greatly increase the number of beneficial users and alleviate the pressure of insufficient bandwidth resources.At the same time,the experimental results show that it greatly reduces the computation cost(the weighted sum of delays and energy)of overall users.The branch and bound algorithm is also verified.The results show that it can obtain the optimal solution and is stable compared with the heuristic algorithm such as diving.Besides,the experimental results corresponding to the economic benefits model show that the proposed bi-level optimization model can help users make reasonable decisions and greatly reduce the overall execution cost(the weighted sum of energy and prices).At the same time,it can help operators to make reasonable pricing and minimize their losses.