An Energy-Efficient Elderly Tracking Algorithm

Location and tracking of healthy or infirm elders is a potentially useful application as society aging is accelerating worldwide and as technologies such as cellular positioning and Assisted Global Positioning System (AGPS) are maturing. For a wearable device to be used in such an application, the battery operating hour is often a key consideration. This paper describes an energy-efficient cellular and AGPS tracking algorithm based on the concept of Personal Common Location (PCL) that we define as the locations at which an elder spends most of his/her time, as recognized from historic cell ID record. During the tracking process, AGPS fixes are avoided if the elderly carrier is determined to be located within a PCL, if the destination PCL can be predicted, or if AGPS signals are likely to be unavailable. This paper also describes our approach for off-line PCL recognition, the necessity of cell ID clustering, as well as the results from our tracking experiments.