A Hybrid Strategy of Location Database Clustering and Dynamic Hierarchical Caching in PCS Networks

Abstract

The location database (LDB) organization has much impact on the performance of Personal Communication Services (PCS) network. In order to optimize the LDB organization, users' moving and calling patterns should be considered together. In this paper, we propose a hybrid strategy of location database clustering and dynamic hierarchical caching for a hierarchical database structure in PCS network. The main goal of our strategy is to reduce the cost of location management and network signaling traffic of data access and communication. The strategy consists of two steps: (1) clustering location databases based on users' moving patterns in the system; (2) arranging the clusters and then caching the users' location information of the adjacent cluster at each cluster's representative node based on users' calling patterns. A threshold required for beneficial caching is used to determine whether to cache. Through experimental simulation, the comparison results demonstrate that our strategy can result in total cost reduction.