Chapter 11: Location Prediction Algorithms for Mobile Wireless Systems

Christine Cheng, Ravi Jain, and Eric van den Berg

Abstract

Predicting the location of a mobile wireless user is an inherently interesting and challenging problem. Location prediction has received increasing interest over the past decade, driven by applications in location management, call admission control, smooth handoffs, and resource reservation for improved quality of service. It is likely that location prediction will receive even more interest in the future, especially given the increased availability and importance of location estimation hardware and applications.

In this chapter, we present an overview of location prediction in mobile wireless systems. We do not attempt to provide a comprehensive survey of all techniques and applications, but offer instead a description of several types of algorithms used for location prediction. We classify them broadly into two types of approaches: (1) domain-independent algorithms that take results from Markov analysis or text compression algorithms and apply them to prediction, and (2) domain-specific algorithms that consider the geometry of user motion as well as the semantics of the symbols in the user's movement history. We briefly mention other algorithms using Bayesian or neural network approaches, and end with some concluding remarks.