CN-121985362-A - Mobile communication network quality monitoring system

CN121985362ACN 121985362 ACN121985362 ACN 121985362ACN-121985362-A

Abstract

The invention relates to a mobile communication network quality monitoring system. The system mainly comprises a multi-source scene binding type acquisition module, a data storage and processing module, a linkage type service management and control module and a double-linkage type situation visualization module, wherein a multi-index collaborative network quality amplitude estimation model is built by acquiring network quality data of multi-source scene sampling, an estimated value of the network quality amplitude is acquired, the linkage type service management and control module acquires monitoring information of the network quality through a dynamic mapping mechanism, predicts abnormal duration and associated abnormal indexes, carries out classified information pushing through a hierarchical authority display mechanism, the double-linkage type situation visualization module acquires network quality monitoring data corresponding to a test place through front-end interface scene layout design of a management system, and visual presentation and structural tracing of the data are obtained by utilizing real-time situation monitoring and multi-dimensional data interaction. Quality monitoring of the mobile communication network is achieved.

Inventors

WANG QIANQIAN
LIANG CHAOHUI
XU YI
YU ZHIHAO
XU DIYUAN
CHEN HONGYI

Assignees

上海泰峰检测认证有限公司

Dates

Publication Date: 20260505
Application Date: 20251225

Claims (12)

1. The mobile communication network quality monitoring system is characterized by comprising a multi-source scene binding type acquisition module, a data storage and processing module, a linkage type service management and control module and a double linkage situation visualization module; the multi-source scene binding type acquisition module constructs a three-dimensional data acquisition system through a preset data test point to acquire network quality data of multi-source scene sampling; the data storage and processing module performs field level mapping and alignment on the associated database based on the network quality data, and performs normalized storage on the multidimensional field according to the space-time label to construct a network quality prediction associated data set; performing multi-order normalization on the data based on the associated data set to generate a normalized steady-state data stream, and extracting and extrapolating time sequence trend by utilizing sliding window energy to obtain a network quality characteristic vector; The linkage type service management and control module acquires monitoring information of network quality through a dynamic mapping mechanism based on the estimated value of the network quality amplitude, manages and binds the monitoring information of the network quality with multidimensional data, predicts abnormal duration time and associated abnormal indexes, and acquires an early warning information packet; The dual-linkage situation visualization module acquires network quality monitoring data of a corresponding test place through the scene layout design of a front-end interface of the management system, and utilizes real-time situation monitoring and multi-dimensional data interaction to construct a dual-view linkage network quality visualization system so as to acquire visual presentation and structural traceability of the data.
2. The system of claim 1, wherein the constructing a stereoscopic data collection system through the preset data test points includes collecting multiple indexes of multiple preset data test points, associating administrative areas with scene types, and requiring to collect field environment data in a special test scene, and binding index data, physical scenes with depths of terminal devices, so as to construct a stereoscopic data collection system.
3. The system of claim 1, wherein the specific process of field-level mapping and alignment of the associated database comprises converting network quality data fields to associated database target fields, through a pre-trained network data semantic encoder, into high-dimensional semantic vectors, and obtaining automatic mapping across naming systems based on vector cosine similarity; The field level alignment adopts a self-adaptive time granularity alignment algorithm, dynamically adjusts a time window according to data density, corrects an asynchronous time stamp through linear interpolation, associates multi-source field data of the same time node, binds the multi-source field data to a digital twin body of a physical network node based on a network topology digital twin model, and obtains space position alignment of cross-equipment through space coordinate coding.
4. The system of claim 1, wherein the process of constructing the network quality prediction related dataset comprises assigning time granularity labels and space topology labels to the multidimensional fields and anchoring the fields and network physical scenes by a space-time label dynamic binding mechanism, and constructing a double-layer index by the space-time labels and field characteristics by a related dataset self-adaptive storage architecture to obtain the related dataset of network quality monitoring of quick traceability and multidimensional related query.
5. The system of claim 1, wherein the multi-order normalization method is characterized in that the association data set monitored by network quality is used as initial input data, de-abnormal data obtained after processing by a time-space domain outlier detection model is used as a first-order normalization value, a multi-scale dynamic smoothing algorithm is adopted based on the first-order normalization value, a smoothing window size is automatically adjusted according to data fluctuation frequency to balance data smoothness and feature retention, a smoothed intermediate data stream obtained after processing is used as a second-order normalization value, a time stamp calibration algorithm is utilized by a time stamp calibration engine to unify multi-device asynchronous time stamps to a standard time axis based on the second-order normalization value, and multi-dimensional field missing data is complemented by means of a attention mechanism time sequence interpolation model, so that a time-continuous and feature-complete normalized steady-state data stream is obtained.
6. The system of claim 1, wherein the process of obtaining the network quality feature vector comprises using a fluctuation response sliding window mechanism, taking a sudden change threshold of the index in the steady-state data stream as a trigger condition, dynamically adjusting a window, extracting steady energy features through a sliding average energy value and an energy attenuation rate, and obtaining a binary energy feature set of a sudden energy parameter and a steady energy baseline; constructing a multisource anchoring trend deduction model, selecting three types of reference data from the associated data set based on the space-time labels of the steady-state data flow, and obtaining a trend feature set of trend directions and fluctuation amplitudes through linear fitting of three-way anchor points; And sorting the screened features according to a two-dimensional matrix of energy intensity and trend stability, and giving feature weights to generate network quality feature vectors with the two-state energy parameters, the three-dimensional anchoring trend parameters and the space-time label features.
7. The system of claim 1, wherein the process of obtaining the estimated value of the network quality amplitude comprises taking as input a bi-state energy parameter, a tri-anchoring trend parameter and a spatio-temporal tag feature in the network quality feature vector, searching a contemporaneous scene sample in a historical database by using a spatio-temporal tag, calculating a contribution ratio of the two types of parameters in an event, proportionally distributing dynamic weights, and obtaining the calibrated estimated value of the network quality amplitude by weighting summation and residual correction.
8. The system of claim 1, wherein the process of obtaining the anomaly information of the network quality is to establish a two-dimensional dynamic mapping system of an anomaly amplitude quantization value and a scene adaptation threshold, wherein the method comprises the steps of utilizing a space-time tag characteristic to call historical contemporaneous scene data, and generating a dynamic threshold of a corresponding scene through a quantile dynamic calibration algorithm; And constructing a three-dimensional association rule of abnormal amplitude, influence range and index cooperativity, and obtaining mapping scores through weighting matrix operation based on the abnormal amplitude quantized values by combining the number of affected network nodes, coverage scale, bi-state energy parameters and multi-index cooperativity abnormal conditions reflected by the ternary anchoring trend parameters associated with the space-time labels, so as to obtain abnormal information of network quality.
9. The system of claim 1, wherein the acquisition process of the early warning information package comprises the steps of retrieving historical contemporaneous scene anomaly data from a network quality prediction association data set based on space-time tag characteristics of anomaly early warning information, extracting corresponding relations of historical anomaly, amplitude change curves and duration, associating binary energy and ternary anchoring trend parameters of the steady-state data stream with anomaly index and occurrence frequency statistics data, calculating similarity of the current anomaly characteristics and the historical anomaly feature, combining a real-time trend slope change rate, obtaining a minute anomaly duration by utilizing a scene adaptation piecewise linear interpolation algorithm, and acquiring the structured early warning information package by utilizing the early warning information and management-bound multidimensional data according to historical anomaly index probability and current parameter cooperative fluctuation conditions.
10. The system of claim 1, wherein the classification information pushing process is to dynamically generate an associated multidimensional identity map for each user based on the coordinated type service management and control and network quality area data bound with the early warning information, and the network quality monitoring system is to perform atomic level disassembly on the early warning content according to the bound network quality area information, screen out information modules matched with different area authorities, perform role level encapsulation according to the user multidimensional identity map, and push the network quality early warning information.
11. The system of claim 1, wherein the process of obtaining network quality monitoring data corresponding to the test site includes the front-end interface of the management system preconfiguring the test site information through a site management function, presenting a speed measurement interface in a scene layout design, displaying positioning information of the current test site on the interface, providing a site selection drop-down option to associate the test site preconfigured by the management system, supporting uploading of functional supplementary site information on a location photo, and obtaining binding of the test site and the network quality monitoring data.
12. The system of claim 1, wherein the constructing the network quality visualization system comprises performing region positioning by using a large-scale real-time situation monitor, selecting corresponding administrative regions, displaying network quality data in a target region after screening, and synchronously updating data statistics information into the target region data, wherein data interaction supports preset screening conditions, and constructing the network quality visualization system in a multi-dimensional manner.

Description

Mobile communication network quality monitoring system Technical Field The invention belongs to the technical field of communication, and particularly relates to a mobile communication network quality monitoring system. Background At the moment of digital transformation acceleration, a mobile network has become a core infrastructure of social operation, economic development and civil service, and the quality of the mobile network directly influences user experience, enterprise operation efficiency and industry digital progress. However, the current mobile communication network quality monitoring field faces multiple bottlenecks, which severely restrict network optimization efficiency and user experience improvement, and are specifically embodied in the following aspects: Network test data are scattered in multiple channels such as special test, public test, OTT platform feedback, user complaint report and the like, multisource data are scattered and stored, a unified association mechanism is lacked, field naming systems of different data sources are large in difference, time stamps are asynchronous, spatial position association is fuzzy, data values are difficult to stack, and full-link quality assessment is lacked complete data support. The scene binding accuracy is insufficient, namely the binding of the data acquisition and the physical scene and the terminal equipment is lack of deep association, administrative region division and scene type classification are not clear enough, and the field environment data acquisition is missing in special test, so that the regional quality bottleneck is difficult to accurately position, the fault investigation has a blind area, and the problem solving period is long. The reliability of data processing is poor, namely, the problems of abnormal value, missing value, asynchronous time stamp and the like exist in the original data, the traditional data processing method is difficult to consider both the data smoothness and the feature retention, the accuracy of the subsequent analysis result is affected, and a stable and reliable data base cannot be provided for network quality evaluation. The method has the advantages of low efficiency of abnormality early warning and pushing, lack of abnormality judgment standards dynamically adapting to different scenes, low prediction precision of abnormal duration time period, incomplete identification of associated abnormality indexes, no combination of user permission and management requirements in early warning information pushing, information overload or insufficient pertinence, and difficulty in acquiring key treatment information by first-line maintenance personnel. The screening dimension is single, the screening function of the traditional monitoring tool is limited to a single dimension or a few core indexes, and the requirements of accurate data retrieval of multiple scenes and multiple conditions are difficult to meet. For example, the inability to combine multi-dimensional combinatorial screening of areas, operators, network types, test time periods, rate intervals, signal strengths, etc. simultaneously results in difficulty in locating network problems in a particular scenario quickly. The visualization and tracing experience is poor, the visualization of the monitoring result shows the lack of hierarchical and multidimensional interactive design, deep drill-down from the overall situation to the local test point cannot be realized, the data tracing flow is complex, the multi-system switching operation is complex, and the full-link data tracing and problem positioning are difficult to complete rapidly. Disclosure of Invention In order to solve the above problems in the prior art, the present invention provides a mobile communication network quality monitoring system. The invention can realize the purposes by the following technical proposal that a multisource scene binding type acquisition module, a data storage and processing module, a linkage type business management and control module and a double linkage situation visualization module; the multi-source scene binding type acquisition module constructs a three-dimensional data acquisition system through a preset data test point to acquire network quality data of multi-source scene sampling; the data storage and processing module performs field level mapping and alignment on the associated database based on the network quality data, and performs normalized storage on the multidimensional field according to the space-time label to construct a network quality prediction associated data set; performing multi-order normalization on the data based on the associated data set to generate a normalized steady-state data stream, and extracting and extrapolating time sequence trend by utilizing sliding window energy to obtain a network quality characteristic vector; The linkage type service management and control module acquires monitoring information of network quality through a