Abstract：Mobile edge computing （MEC） can effectively serve task computing by sinking the computing center to the edge of the network. However， the computing resources of MEC are not unlimited， which brings many problems. This paper analyses a joint optimization problem between unloading and resource allocation to maximize the total revenue of MEC for computational resources finite MEC systems. Firstly， the interaction between MEC and users is described through the Stackelberg model， the differential pricing strategy is used to increase the constraints on unloading， and then the unloading problem is transformed into a binary backpack problem. Finally， the computing resources are allocated by improving the simulated annealing algorithm， and the optimal scheme is obtained by iteration. Simulation results show that the proposed scheme can maximize the number of user unloading and effectively improve the benefits of MEC systems.