Social Link Analysis Using Wireless Beaconing and Accelerometer

As smartphones equipped with numerous sensors are ubiquitous and are almost always kept in close proximity of their owners, identifying opportunistic links from the real world and incorporating them in online social networks are now feasible. However, like any other application that involves smartphone sensing, this endeavor also faces the challenge of efficient energy consumption and robustness against data loss due to collisions. In this research work, we try to address these challenges. We have proposed a sensing system that senses human proximity consuming less energy. We have also developed a prediction mechanism that makes our proposed system more robust against sensing data losses. Evaluation results based on both simulation and real world data show that the system demonstrates greater robustness with less power consumption, hence the proposed system is a practical solution that is ready to be deployed.