Opportunistic crowd sensing in WiFi-enabled indoor areas

Crowd sensing in indoor areas is becoming more and more fundamental for flow management, security and surveillance, or building usage statistics. This paper deals with a simple crowd sensing approach, which opportunistically exploits the already deployed WiFi networks, thus avoiding dedicated wiring and installations. The proposed algorithm is based on a two-step procedure that first applies a Wavelet decomposition of the signal strength data and then exploits the obtained coefficients to learn the unknown relation between crowd presence and signal changes. To this end, a customized learning-by-example (LBE) algorithm is trained for successive real-time crowd detection. The results of the experimental validation are presented to assess system potentialities and current limitations.