Trajectory Predictions in Mobile Networks

This paper proposes a mobility model and a neural location predictor (NLP) to predict the location of a mobile host (MH). The seemingly random movement is actually a logical function of the user’s position, speed, acceleration and direction. (x,y) of a coordinate should be treated independently such that NLP can be designed with fewer mathematic operations and achieves high computing efficiency. Accurate and speedy location prediction can also minimize signaling traffic required by QoS maintenance during seamless handoff.