On Participant Selection for Minimum Cost Participatory Urban Sensing with Guaranteed Quality of Information

Exploring vehicles to conduct participatory urban sensing has become an economic and efficient sensing paradigm to pursue the smart city vision. Intuitively, having more vehicles participate in one sensing task, higher quality-of-information (QoI) can be achieved. However, more participation also implies a higher sensing cost, which include the cost pay to participated vehicles and 3G traffic cost. This paper introduces an interesting problem on how to select an appropriate set of vehicles to minimize the sensing cost while guaranteeing the required QoI. In this paper, we define a new QoI metric called coverage ratio satisfaction (CRS) with the consideration of coverage from both temporary and spatial aspects. Based on the CRS definition, we formulate the minimum cost CRS guaranteeing problem as an integer linear problem and propose a participant selection strategy called Vehicles Participant Selection (VPS). The high efficiency of VPS is extensively validated by real trace based experiments.