Crowd Estimation Based on RSSI Measurements Using kNN Classification

In this paper, crowd estimation based on the received signal strength indicator (RSSI) is presented. Crowd is estimated for a given measured RSSI based on the trained kNN classifier. For training, a measurement setup consisting of one transmitting and two receiving wireless nodes is proposed. Moreover, RF data is collected to form a Fingerprint database consisting of features. For accurate feature extraction and optimization of the time required for training Bootstrap resampling technique is used. The presented kNN classifier has shown 94% accuracy using n-fold cross validation for up to 14 people crowd density.