The invention discloses an edge cooperation cache optimization method based on user mobility prediction, and the method comprises the following steps: 1, building a Markov prediction model according to the movement track of a user, and obtaining the probability that the user moves to each micro-cell base station; 2, grouping all the micro-cellular base stations in the network by using a hierarchical clustering algorithm, wherein the micro-cellular base stations in each group are connected to the same service gateway; And step 3, considering the file popularity distribution, and adopting a distributed cooperative caching method in all the micro-cellular base station groups based on the movement probability of the user, and enabling the micro-cellular base stations in the same micro-cellularbase station group to jointly execute a caching decision and cooperate to provide service for the user. The method has the advantages that the movement probability and the file popularity of the userare considered, the method that cooperative caching is adopted in each group of the base station is provided, the utilization rate of a caching unit is improved, and the time delay of user access content is reduced.