An Efficient Mobility Management by Prediction of User Mobility in Mobile Computing Environments

摘要

Mobility prediction is an important issue for mobility management in mobile computing systems. Because of the ntrinsic characteristics of mobile computing systems associated with user mobility, it is required that mobility management be more intelligent nad be more effective. In this paper, we propose an intelligent mobility prediction scheme that can predic future locations from the movement patterns of a mobile user for past several days or several months, i.e., the location history of a mobile user. To model the movement of a mobile user, we use a neuro-fuzzy inference system. Since the neuro-fuzzy inference system predicts a future location where a mobile user will move, our scheme can be used for supporting the pre-allocation of communication channels and the pre-assignment of deficient resources, etc.