A wireless communication network feature portrait method, device and computer program product

Abstract

The present invention provides a wireless communication network feature portrait method, device and computer program product, comprising: S1, acquiring high-dimensional features of the wireless communication network and performing vectorization to obtain high-dimensional feature vectors of the wireless communication network; S2, based on training The deep self-encoding model of the wireless communication network performs feature extraction and feature dimensionality reduction; S3. Based on the clustering algorithm, similar wireless network features are aggregated, and the aggregated wireless communication network features are drawn. S4. Detect abnormal features of the wireless communication network based on the reconstruction error of the deep self-encoding model. The label description is carried out through the feature description link, and the image and multi-layer mapping coding of complex network features are realized. At the same time, the reconstruction error can be calculated through the trained self-encoding model, and the special scene characteristics of the wireless network can be realized through the error evaluation mechanism. Or the detection and identification of abnormal cell characteristics.

Description

technical field [0001] The present invention relates to the technical field of analysis and optimization of a wireless communication network, and more specifically, to a method, device and computer program product for characterizing a wireless communication network. Background technique [0002] The mobile communication network is a dynamic and complex wireless network. Due to factors such as uneven distribution of traffic density, increasingly tight frequency resources, complex and diverse network configurations, and long-term changes, the service quality of the existing network cannot achieve the best, and the huge investment has not obtained the highest return. All of these Operators are required to collect and analyze data on the mobile communication network to find out the reasons that affect network quality in order to make various adjustments, optimize resource allocation, and set network parameters reasonably to make the network reach the best operating state and obt...

AI Technical Summary

Problems solved by technology

[0006] The present invention provides a wireless communication network feature profiling method, device, and computer program product that overcome the above-mentioned problems or at least partially solve the above-mentioned problems, and solve the problem of extensive network feature analysis in the prior art with more and more network dimensions , too simple to achieve refined and scenario-based network analysis and optimization

Images