Mobile Crowd Sensing

People-centric sensing with smart phones can be used for large scale sensing of the physical world at low cost by leveraging the available sensors on the phones. Despite its benefits, mobile people-centric sensing has two main issues: (i) incentivizing the participants, and (ii) reliability of the sensed data. Unfortunately, the existing solutions to solve these issue either requires infrastructure support or adds significant overhead on user phones. We believe that mobile crowd sensing will become a widespread method for collecting sensing data from the physical world once the data reliability issues are properly addressed. We present the concept of mobile crowd sensing and its applications to everyday life. We describe the design and implementation of McSense, our mobile crowd sensing platform, which was used to run a user study at the university campus for a period of two months. We also discuss the data reliability issues in mobile crowd sensing by presenting several scenarios involving malicious behavior. We present a protocol for location reliability as a step toward achieving data reliability in sensed data, namely, ILR (Improving Location Reliability). ILR also detects false location claims associated with the sensed data. Based on our security analysis and simulation results, we argue that ILR works well at various node densities. The analysis of the sensed data collected from the users in our field study demonstrate that ILR can efficiently achieve location data reliability and detect a significant percentage of false location claims.