An adaptive location management strategy using forwarding pointers

We present an adaptive location management strategy using forwarding pointers to locate mobile users in Personal Communications Systems (PCS). The proposed strategy aims at reducing the overall location cost by adapting itself according to the user mobility and calling patterns. The work is motivated by the observation that the set of callers initiating most of the calls for a mobile host is actually small and relatively stationary. Thus, a saving in the lookup delay can be realized by informing this set of callers at each move. This saving is significant for high Call to Mobility Ratios (CMR) but imposes excessive update messages for low CMR. A reduction in the number of updates can be made possible by using forwarding pointers and only informing the set of frequent callers whenever the pointers chain length reaches its maximum. This modification reduces the cost of updates for low CMR but at the expense of a slight increase in the lookup cost. The algorithm determines the frequent caller set dynamically using a simple on-line algorithm. Moreover, to tradeoff between lookup and update cost, it dynamically opts for the use of forwarding pointers according to the rate of mobility. We quantify the costs and benefits of the proposed strategy for a range of CMR values. It has been shown that, under certain assumptions, for low CMR, our strategy can result in 30% to 60% savings in the total cost, and for high CMR, the improvement ranges from 20% to 50%.