Research On User-edge Collaborative Resource Allocation And Offloading Strategy

With the advent of the era of big data and the development of Internet of Things technology,the demand for data and services from users has exploded.The user's mobile device cannot obtain continuous power supply,and limited battery capacity prevents many computationally intensive smart applications from being executed locally.Cloud computing technology can provide centralized services for users.However,because the data and service centers are far away from users,the high delay caused by long-distance communication reduces the user's experience quality.Edge computing provides a new solution.A large number of servers are deployed at the edge of the network,that is,near user equipment,to provide users with lowlatency computing and services and reduce core network pressure.In this paper we mainly studied the computational offloading decision and computing resource allocation of edge computing.Computing offloading refers to the process by which user equipment transfers computing tasks to edge servers.How to perform offloading is called the offloading decision problem.A single edge server has limited resources.How to utilize and allocate computing resources is a resource allocation problem.In this paper we studied edge computing architecture and proposed a user-edge collaborative computing offloading process.In this process,the user equipment assists the edge server to collect global information,and the edge server makes accurate offload decisions.We studied the edge computing scenarios of single-edge server multi-user devices,and performed mathematical modeling and problem formulation.For the formulated mixed-integer non-linear programming problem,we divided it into two sub-problems and used the function monotonicity and genetic algorithm to solve them.We described the genetic algorithm process and formulated three core operators: selection,intersection,and mutation.For the underutilization of computing resources caused by the interaction between variables,we proposed an iterative method based on genetic algorithms to solve it.For the method proposed in this paper,we designed simulation experiments and compared a variety of computing offloading strategies and resource allocation methods.Experimental results show that our method can achieve lower task execution costs and show better performance.