Adaptive Location Management in Mobile Environments

Abstract

Location management in mobile environments consists of two major operations: location update and paging. The more up-to-date the location information, the less paging becomes necessary and vice versa. The conventional approach is the location area based approach (LA-based approach), where a location area (LA) consists of multiple cells. When the mobile station (MS) enters a new location area, the MS immediately updates its location information at the new location’s visitor location register (VLR) and this update is propagated to the MS’s home location register (HLR). The major drawback of the LA-based approach is that it does not consider any mobility patterns, or call arrival patterns. Moreover, MS updates frequently when it roams only within the boundary cells of different location areas, resulting in unnecessary location updates. So, there is a need for an efficient algorithm which can eliminate the drawbacks of the LA-based approach. To this end, an adaptive location management algorithm is described in this paper: an MS dynamically determines whether or not to update when it moves to a new LA, so that each location update becomes a necessary location update, i.e., in each updated location area at least one call is made. We have conducted experiments to capture the effect of mobility and call arrival patterns on the new location update strategy. We have also tested our algorithm with SUMATRA (Stanford University Mobile Activity TRAces), which has been validated against real data on call and mobility traces. Experimental results show that our adaptive location management algorithm considerably reduces the location management cost, by avoiding unnecessary location updates.