Adaptive Location Prediction Strategies Based on a Hierarchical Network Model in a Cellular Mobile Environment

Open Access
Publisher Website
Google Scholar
Abstract
We present four efficient heuristics (one basic scheme and its three variants) to predict the location of a mobile user in a cellular wireless network. These location management schemes assume a hierarchy of location areas, which changes dynamically with traffic patterns, and estimate the location probabilities of each user. Depending on the movement profiles for the last τ time units, the most probable location area (MPLA) and the future probable location area (FPLA) are computed for the user. The basic scheme and its first variant attempt to find the user in the MPLA first and then in the FPLAs before searching the whole system area. The second variant attempts to predict the location probabilities of the user in the future cells, which are in turn predicted from the user-profile while forming the FPLA. It then forms an MPLA consisting of cells traversed in the last τ time units as well as the future cells. The third variant is a hybrid of the first and second variants. A probabilistic analysis of total paging cost is also presented for the basic scheme and its first variant. Finally, the proposed heuristics are validated and compared through extensive simulation studies under various traffic patterns.

This publication has 0 references indexed in Scilit: