In this paper, we propose a methodology for estimating the crowd speed usingWiFi devices, and without relying on people to carry any device (passively).Our approach not only enables speed estimation in the region where WiFi linksare, but also in the adjacent possibly WiFi-free regions where there may be noWiFi signal available. More specifically, we use a pair of WiFi links in oneregion, whose RSSI measurements are then used to estimate the crowd speed, notonly in this region, but also in adjacent WiFi-free regions. We first prove howthe cross-correlation and the probability of crossing of the two linksimplicitly carry key information about the pedestrian speeds and develop amathematical model to relate them to pedestrian speeds. We then validate ourapproach with 108 experiments, in both indoor and outdoor, where up to 10people walk in two adjacent areas, with a variety of speeds per region, showingthat our framework can accurately estimate these speeds with only a pair ofWiFi links in one region. For instance, the NMSE over all experiments is 0.18.Furthermore, the overall classification accuracy, when crowd speed iscategorized as slow, normal, and fast, is 85%. We also evaluate our frameworkin a museum-type setting, where two exhibitions showcase two different types ofdisplays. We show how our methodology can estimate the visitor speeds in bothexhibits, deducing which exhibit is more popular. We finally run experiments inan aisle in Costco, estimating key attributes of buyers' behaviors.
展开▼
