Client location tracking with k-step prediction and cache policy in ubiquitous information service network

The paper presents a feasible and efficient methodology that improves client mobility management under an ubiquitous information service network. The authors base the system on the open integrated service network architecture and propose a hybrid infrastructure with facilities of both a centralized network management center (NMC) and a distributed home base node (HBN). They use LSDMk for choosing a HBN with minimum cost by computing the expected registration cost and location tracking cost according to users´ future k-step movement and called frequencies. LSDM is a special case of LSDMk when k is equal to 0. In addition, they use a cache strategy to save location tracking cost. Running simulations, results show that the methodology is able to save up to 86% and 54% of the cost comparing to that of IS-41 in best and average cases, respectively, when the value of call-mobility ratio (CMR) is in the range of (1, 100), which means either clients are called with high frequencies or are moving with low mobilities. As the value of CMR goes down to the range of (0.01, 1), which means either clients are called with low frequencies or are moving with high mobilities, one may save up to 83%, 60%, and 10% costs in best, average, and worst cases, respectively