Blind-area Elimination in Video Surveillance Systems by WiFi Sensing with Minimum QoS Loss

Video surveillance systems have demonstrated their great importance in security protection these years. However, due to limited budget, installing cameras in every place of surveillance region is not practical and then blind areas become inevitable. As a result, eliminating blind areas with the lowest cost has become a tough challenge. To the best of our knowledge, this is the first work that fixes the blind spots of video surveillance systems based on WiFi sensing technique. By taking existing WiFi infrastructure as the sensing device, this paper attempts to eliminate blind spots with tiny hardware cost. Moreover, in order to completely fix blind area with minimum loss of video communication QoS, the WiFi sensing device’s location boundary that satisfies the above objective is modeled and estimated. To that end, a visitor-disturbed channel model is first derived for precisely describing the inherent relation between the appearance of visitor and the change of wireless channel. Then a location boundary model satisfying both blind-area elimination and QoS maximization is further derived. Based on the derived model, a practical system is designed to estimate the real location boundary. The simulation and experiment results have not only verified the correctness of our derived location boundary model, but also showed its good performance on both blind-area elimination and communication QoS optimization.