Fingerprint in the Air: Using the RSS Data for Uniqueness Identification

Indoor localization, device identification, and wireless attendance security systems are widely used in recent years. There is always a premise that each person can carry only one wireless device by himself, which is no longer valid nowadays. To detect the uniqueness identification problem, the bio-assisted methods such as fingerprint, face or gait recognition systems are deployed near the entrance which are difficult to use. This paper studies such problem using RF RSS fingerprints which can be collected and modeled as time series. Then we can calculate the similarity of the time series to judge the uniqueness identification problem. Firstly, a naive algorithm using dynamic time warping is presented to simply compute the similarity the asynchronous time series. Then an improved algorithm is proposed to reduce the computational complexity while keeping the robustness. Simulation and experiments results show that our algorithms can perfectly detect the uniqueness identification problem with a reasonable cost.