Abstract: In mobile edge computing, if mobile users choose the strategy of unloading all the applications to the Edge server, it will not only increase the delay, energy consumption, but also cause the server to accumulate too much information, which will lead to the leakage of Mobile users' privacy. To solve this problem, considering the user's privacy tolerance differences, an edge computing unloading strategy is proposed on the basis of considering user differences. According to the delay and energy, this paper formulated a pricing model and a utility function combined the privacy model and different privacy breach tolerance. Based on game theory, this system will choose the need unloading mobile users. This simulation results show that this method can reduce the local delay and energy consumption, and protect the privacy of mobile users to a certain extent.