Energy Efficient Sensor Management Strategies in Mobile Sensing

Mobile device based human-centric sensing and participatory sensing provide a vast context about the user state information. To capture and recognize any user state, then to classify it requires operating all existing sensors in a mobile device continuously. Nonetheless, constantly running sensors drain the mobile device’s battery rapidly. Therefore, it is imperative to construct a framework to utilize sensors efficiently and try to recognize user states very accurately by consuming less power. In this paper, some strategies are proposed to correct the mentioned deficiency. It is aimed to detect and recognize user state transitions by first decreasing computational complexity during data processing stages, prompting time-varying efficient duty cycle schedules and setting adaptive sampling time periods to satisfy trade-off between accuracy vs. energy consumption.