A Method to Evaluate the Energy-Efficiency of Wide-Area Location Determination Techniques Used by Smartphones

Location-based services (LBS) are one of the most useful applications for smart phone users, but there is a significant energy cost in acquiring the user location. Continuous location sampling using GPS, Wi-Fi positioning system (WPS), and Global System for Mobile Communications Positioning System (GSMPS) typically deplete the battery within 12, 46, and 63 hours respectively, compared to 284 hours when GPS and Wi-Fi location positioning are turned-off. We present the design, implementation, and evaluation of a novel method to evaluate the energy-efficiency of GPS, WPS, and GSMPS location sensing technologies used by smart phones based upon a user-centred metric, battery depletion time. This metric depends upon three main factors: location sensor usage, user mobility context determination and location accuracy. Using solely the embedded smart phone accelerometer our pattern recognition model can within 2 seconds identify the user mobility state which in-turn manages the activation and deactivation of the location determination. Our results show that our hybrid energy efficient location sensing architecture can achieve energy-savings of up to 57% in typical circumstances.