A QoS adaptive mobility prediction scheme for wireless networks

Abstract

Mobility prediction based on an individual's movement history has been reported as an effective means to decrease the call-dropping probability and to shorten handover latency. Applying various basic prediction schemes to a realistic office environment, it is shown that mobility prediction using an individual's movement history has limitations, and the statistical randomness of the user motion can prevent accurate prediction performance. We propose a QoS adaptive mobility prediction scheme to resolve these problems. A stricter QoS compliance can be achieved through a pro-active probability update mechanism, a supplementary correlative movement history and the concept of prediction confidence ratio.