Failure-aware mobile crowd sensing: A social relationship-based transfer approach

As an appealing sensing paradigm, Mobile Crowd Sensing (MCS) which provides a cost-efficient solution for large-scale urban sensing tasks has gained significant attention in recent years. However, in practice, many MCS applications usually suffer from the failure of sensing task execution, ranging from the randomness and autonomous in participant users' behavior, to lacking of prior experience and monetary reward, etc. To mitigate the impact of these failures, in this paper, we propose and study a novel problem, namely failure-aware mobile crowd sensing. To solve our problem, we devise a two-stages framework, including offline task allocation and online task transfer. Towards enhancing task completion ratio, we propose an indeterminate fitness proportionate based task allocation approach FPSAll, and an utility evaluation-based task transfer approach FTASKTraf, respectively. Through extensive experiments, we demonstrate the efficiency and effectiveness of our proposed approaches on real-world data set.