Accurate user location identification method based on neural network

Abstract

The invention discloses an accurate user location identification method based on a neural network. The method comprises the following steps: establishing a position fingerprint database by using collected locations of users and network environments of the users, and then providing accurate locations of the users through the network environments of users having no location. By adoption of the accurate user location identification method, different buildings and different heights in the buildings can be accurately located, the real perception of the users can be accurately reflected, so that we can know the current situation of the network and accurately locate the problem of the present network, an optimization scheme is efficiently output based on matching multiple data sources, and the network problems are discovered prior to customers, thereby reducing the complaints. A lot of manpower and material resources are reduced, meanwhile, a strong support is provided for achieving overall, comprehensive and scientific planning of indoor coverage, the influence of various subjective factors in artificial planning is reduced, and precise planning and optimization of the network can be realized.

Description

technical field [0001] The present invention relates to the field of data information technology, in particular to a neural network-based accurate user location identification method, which is a method for calculating the user's precise location by using the user's mobile network environment. Background technique [0002] The current telecommunications industry can only obtain the location of user businesses through user app uploads, base station location discovery, and POI information of various map service providers. Each of these methods has advantages and disadvantages. Uploading through the user app is an active way for the user, and the user can upload it or not. Moreover, this method generally uses GPS positioning, and the outdoor accuracy is generally about 10 meters. Of course, there are also positioning errors of hundreds of meters, but it is basically unavailable indoors; the base station location discovery method is actively discovered by the operator, which can...

AI Technical Summary

Problems solved by technology

[0023] Pruning: The complete decision tree may describe the characteristics of training samples "too precise" (affected by noise data), lack of general representativeness and cannot be used to make classification predictions for new data, and "overfitting" occurs

Images