A regular path recognition method and prediction of user movements in wireless networks

Abstract

Mobile positioning supported control of wireless networks will play an important role in future network optimisation with upcoming location services. We developed a method to predict user movements based on user profiles that have been generated based on measurement reports. The algorithm exploits regular human behaviour in terms of periodic daily activities, such as travelling to work, which results in probability that can be allocated to used paths. The developed scheme can be applied for location and handover support in future wireless networks.