Wireless device detection systems and methods incorporating streaming survival modeling for discrete rotating identifier data

Abstract

A wireless device detection system can include processors, memory devices, and sensors to receive temporary identifiers transmitted between a base station and a mobile wireless device. The memory devices include instructions that cause the processors to build a cumulative distribution function for survival probability based on delta times between multiple access events for each of multiple temporary identifiers received from the sensors. In response to a new access event, the processors add a corresponding temporary identifier to a streaming list and assign a survival probability value, based on the cumulative distribution function, to a latest access event for each temporary identifier contained in the streaming list. The processors remove temporary identifiers from the streaming list that have a survival probability value less than a threshold value, compare the number of temporary identifiers contained in the streaming list to a number of devices known to be present, and calculate the probability that a device corresponding to each of the temporary identifiers contained in the streaming list is present.