System and method for localization and traffic density estimation via segmentation and calibration sampling

摘要

Embodiments are provided for calibration, preprocessing, and segmentation for user localization and network traffic density estimation. The embodiments include sending, from a network component, a request to a plurality of user equipment (UEs) to participate in reporting localization data. Reports for localization data including no-lock reports are received from at least some of the UEs. The no-lock reports indicate indoor UEs among the UEs. The network preprocesses the localization data by eliminating, from the localization data, data that increases the total noise to signal ratio. The localization data is then processed using a model that distinguishes between different buildings. This includes associating, according to a radio map, radio characteristics in the localization data with corresponding bins in a non-uniform grid of coverage. The non-uniform grid is predetermined to maximize uniqueness between the radio characteristics. The indoor UEs are associated with corresponding buildings using the no-lock reports data.

主权项

1. A method for traffic density estimation and user localization, the method comprising:
sending, from a network component, a request to a plurality of user equipment (UEs) to participate in reporting training data, the training data including location and/or fingerprint data; receiving, from at least a portion of the plurality of UEs, a plurality of reports for training data including no-satellite-lock reports and the training data from the at least the portion of the plurality of UEs, wherein the no-satellite-lock reports indicate indoor UEs among the at least the portion of the plurality of UEs; discarding at least some of the training data that includes a noise level higher than a pre-determined threshold; associating, by the network component, the indoor UEs with corresponding buildings in a service area of the network component in accordance with the training data and the no-satellite-lock report to produce building associations; building a radio map in accordance with the training data and the no-satellite-lock report to produce a non-uniform bin grid; and using a soft decision approach for performing traffic density estimation, without performing hard localization, in accordance with the radio map, the non-uniform bin grid, and the building associations.