Research on Computing Offloading and Resource Joint Optimization Under Delay Constraints

In mobile edge computing, the traditional multiple access and computation offload methods based on average performance constraints can no longer meet the demand of delay sensitive services of mobile user devices. Based on Lyapunov optimization theory, a delay constrained dynamic computing offloading algorithm is proposed for the scenario of edge network with single access point and single mobile edge host. This algorithm minimizes the long-term average power consumption problem of user devices by jointly optimizing the allocation of computing and communication resources under average delay and out-of-service probability constraints. Through formula derivation, the original optimization problem is defined as a convex optimization problem within each time slot, and the solution of the problem is obtained by solving the convex optimization problem in each time slot. The simulation experiment results show that compared with the joint resource allocation scheme based on the constraint of stable average rate of task buffer, the proposed scheme has better potential gains and improve system performance.