Optimizing Position Updates for kNN - Is It Worth It?

Keeping a mobile device always up-do-date about the k nearest neighbor (kNN) points of interest (PoIs) is a common challenge of location based services. The challenge lies in the fact that the set of PoIs available at the device needs to always remain correct and fresh while the mobile user may be constantly moving about. To this end, in a very simple approach, the position of the device could be continuously updated with the server- side, whenever it changes. Since this simple approach creates a potential communication overhead, several optimizations have been proposed in literature, all of which significantly reduce the amount of messages sent over the air. However, location based service (LBS) developers have been left clueless until today about whether deploying those complex strategies into their systems actually improves battery life sig- nificantly - which is arguably the most important aspect for the end user. As a way to remedy, this paper critically examines optimized kNN tracking, based on real-life data sets and in the light of actual battery drainage.