Significant Location Identification Based on User Behavior Modeling

Abstract

We describe a cost-effective method which enables location-aware applications in PLMN environments. The proposed approach is assuming no terminal support and it is based on signaling traffic analysis in respect to selected invariant patterns of human behavior, IHAP, and network topology. The experiments carried out in real-life network environment are revealing robust performance, in terms of accuracy and quality of experience (QoE), for certain classes of context aware applications.