Refining and selecting timing data to increase accuracy of timing-based location for user equipment

摘要

Machine learning on user equipment candidate locations provides more accurate location estimation in timing-based location estimation. For instance, timing signals from a user equipment are collected, and adjusted using calibration data, previously known for pairs of fixed-located cells. The adjusted timing data are processed into a location candidate dataset. The location candidate data set is iteratively processed using machine learning technology to eliminate candidate locations until an estimated location is determinable. The location estimation via the described technology is significantly more accurate than other timing-based methods.