Abstract: In this paper, a joint optimization algorithm of base station selection, computation offloading and resource allocation in multi-cell cellular networks is proposed in the multi-layer computation offloading scenario composed of the multi-base station and remote cloud. The algorithm considers the base station selection of multi-base station overlapping coverage users and constructs the minimization problem of the weighted sum of energy consumption and delay under the constraint of edge server computing resources, which is NP-hard. We start from the single user multi-cell computation offloading problem solved by the Lagrange multiplier method. Secondly, for the multi-user multi-base station scenario, considering the user's selection of base station and the competition of computing resources of edge server, the access base station is selected based on the defined selection function, and the suboptimal iterative heuristic algorithm is adopted to make the dynamic correction for the offloading decision in the single-user scenario, so as to obtain the offloading decision and resource allocation of the edge server. Simulation results show that the proposed algorithm can effectively reduce the task completion delay and terminal energy consumption.