| CPC G01S 5/26 (2013.01) | 20 Claims |
|
1. A method for geolocating a mobile computing device within an indoor environment, comprising:
generating a geolocation request audio signal by a speaker of a first mobile device starting at a first point in time, the geolocation request including a fingerprint associated with the first mobile device;
receiving, by a microphone of the first mobile device, a reply audio signal from a second device;
extracting, by one or more processors of the first mobile device, information encoded in the reply audio signal, wherein the information encoded in the reply audio signals comprises a handshake pattern and the fingerprint, and wherein the extracting comprises:
sampling the reply audio signal;
storing the sampled audio signal in a first shift register; and
generating a confidence level that the sampled audio signal in the first shift register contains the handshake pattern, and in response to the confidence level exceeding a predetermined threshold, storing the sampled audio signal in a second shift register, wherein generating the confidence level comprises providing the sampled audio signal in the first shift register to a trained machine learning model executing on the first device;
estimating, by the one or more processors, a receipt time by the first mobile device for the reply audio signal, wherein estimating the receipt time for the reply audio signal comprises matching the handshake pattern to a predetermined data frame; and
calculating a first time of flight (TOF) using the first point in time and the estimated receipt time for a second audio signal.
|