Abstract: Mobile edge computing (MEC) provides users with low latency and low energy consumption services by unloading computing tasks to MEC servers, overcoming the shortcomings of traditional cloud computing. In mobile edge computing, how to make unloading decision is one of the key technologies to provide low latency and low energy consumption services. In addition, due to the limited bandwidth resources of wireless channel, unreasonable bandwidth allocation will increase the energy consumption and delay of user devices. Therefore, how to reasonably allocate resource is also the key to the implementation of edge computing. To jointly optimize delay, energy consumption, and computational resources, this paper proposes a multi-channel search algorithm based on Monte Carlo tree search (MCS-MCTS). Firstly, with the cost of delay and energy consumption as the optimization objective, the decision on computing resource allocation and transmission power modeling is modeled as a convex optimization problem. The optimal transmission power allocation problem is solved by using gradient descent method, and the optimal computational resource allocation problem is solved by means of Lagrange multiplier method and Karush-Kuhn-Tucker (KKT) condition. Subsequently, the MCS-MCTS algorithm is used to deal with the binary unloading decision problem. Besides, the simulated annealing algorithm is introduced to avoid the search results falling into local optimum. The numerical results show that MCS-MCTS algorithm can obtain near optimal unloading decision and resource allocation decision in linear coherent time. Compared with existing heuristic search algorithms, this algorithm can achieve near optimal performance while reducing time complexity and improving system energy efficiency.