Network quality intelligent evaluation method and system, electronic device and storage medium

Abstract

The present application relates to the technical field of communication quality, and especially relates to a network quality intelligent evaluation method and system, an electronic device and a storage medium. The method comprises the following steps: obtaining scene attributes and multi-dimensional index data of a network to be evaluated, wherein the multi-dimensional index data comprises index data of multiple dimensions for representing different aspects of performance of the network to be evaluated; determining evaluation weights corresponding to the index data of each dimension according to the scene attributes of the network to be evaluated; and performing weighted fusion calculation on the index data of each dimension according to the evaluation weights corresponding to the index data of each dimension, so as to obtain a comprehensive quality score of the network to be evaluated. The multi-dimensional index data containing multiple dimensions is adopted, and the comprehensive quality score is obtained through weighted fusion calculation. The previous isolated analysis index is organically fused, the evaluation blind area caused by a single index perspective is avoided, and the accuracy and comprehensiveness of the evaluation result are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of communication quality technology, and in particular to a network quality intelligent assessment method, system, electronic device, and storage medium. Background Technology

[0002] With the large-scale deployment of fifth-generation (5G) mobile communication networks, network quality directly affects user service experience, industry application reliability, and operator network operation and maintenance efficiency. High-quality and efficient network assessment not only supports the accurate allocation of network resources but also provides crucial decision-making support for fault early warning, performance optimization, and service assurance. Currently, 5G network quality assessment mainly relies on the statistics and analysis of key performance indicators.

[0003] Common assessment methods include analyzing signal coverage data from measurement reports, determining cell load levels using physical resource block utilization, or obtaining road coverage quality through drive test data. While these methods can reflect part of the network's operational status from a specific perspective and provide basic data support for daily maintenance, each indicator is usually analyzed independently, lacking effective correlation analysis. This makes it difficult to detect certain complex issues (such as high signal strength but low actual data rate) in a timely manner, resulting in insufficient accuracy and comprehensiveness of the assessment results, and failing to truly reflect the actual service quality of the network. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a network quality intelligent assessment method, system, electronic device and storage medium to solve the above-mentioned technical problem.

[0005] The technical solution of this invention to solve the above-mentioned technical problems is as follows: a network quality intelligent evaluation method, the method comprising: acquiring scene attributes and multi-dimensional indicator data of the network to be evaluated, wherein the multi-dimensional indicator data includes indicator data of multiple dimensions used to characterize the performance of different aspects of the network to be evaluated; determining the evaluation weights corresponding to the indicator data of each dimension according to the scene attributes of the network to be evaluated; and performing weighted fusion calculation on the indicator data of each dimension according to the evaluation weights corresponding to the indicator data of each dimension to obtain a comprehensive quality score of the network to be evaluated.

[0006] The beneficial effects of this invention are as follows: By acquiring the scenario attributes of the network to be evaluated and dynamically determining the evaluation weights of each dimension of the indicator data based on these scenario attributes, the evaluation model can automatically adjust its evaluation focus according to changes in the network environment or business scenario. This overcomes the problem of evaluation results being out of sync with operational needs caused by fixed weights, achieving scenario-adaptive differentiated evaluation. Simultaneously, by using multi-dimensional indicator data containing multiple dimensions and calculating a comprehensive quality score through weighted fusion, the previously isolated indicators are organically integrated, avoiding evaluation blind spots caused by a single indicator perspective and significantly improving the accuracy and comprehensiveness of the evaluation results.

[0007] Based on the above technical solution, the present invention can be further improved as follows.

[0008] Furthermore, the multi-dimensional indicator data includes capacity dimension data, coverage dimension data, structure dimension data, and quality dimension data; wherein, the capacity dimension data describes the network carrying capacity; the coverage dimension data describes the signal reachability and quality; the structure dimension data describes the rationality of the network composition; and the quality dimension data describes the user's actual perception.

[0009] The beneficial effects of adopting the above-mentioned further scheme are: by refining network quality assessment into four orthogonal dimensions of capacity, coverage, structure, and quality, a comprehensive quantification of network carrying capacity, signal quality, rationality of composition, and real user perception is achieved, avoiding the blind spots of single-dimensional assessment and providing a scientific and complete data foundation for subsequent fusion computing.

[0010] Furthermore, determining the evaluation weights corresponding to the indicator data of each dimension based on the scenario attributes of the network to be evaluated includes: matching the weight configuration strategy corresponding to the scenario attributes of the network to be evaluated from a preset weight template library, wherein the preset weight template library contains weight configuration strategies under different scenario combinations; and determining the evaluation weights of each dimension in the matched weight configuration strategy as the evaluation weights corresponding to the indicator data of each dimension.

[0011] The beneficial effects of adopting the above-mentioned further solution are: by pre-establishing a weight template library for different scenario combinations and automatically matching the corresponding weights according to the scenario attributes of the network to be evaluated, the evaluation strategy can be quickly and adaptively adjusted, avoiding manual configuration one by one and improving the flexibility and efficiency of the evaluation.

[0012] Furthermore, the step of weightedly fusing and calculating the indicator data of each dimension according to the evaluation weights corresponding to the indicator data of each dimension to obtain the comprehensive quality score of the network to be evaluated includes: for the indicator data of each dimension, normalizing each indicator in the indicator data of the dimension to obtain the normalized value of each indicator; for the indicator data of each dimension, weighted summing of the normalized values ​​of each indicator in the indicator data of the dimension according to the preset indicator weights to obtain the dimension score of the indicator data of the dimension; and weighted summing of the dimension scores of the indicator data of each dimension according to the evaluation weights corresponding to the indicator data of each dimension to obtain the comprehensive quality score of the network to be evaluated.

[0013] The beneficial effects of adopting the above-mentioned further scheme are: by first normalizing each indicator to eliminate the influence of dimensions, and then using a two-layer weighted summation method, the unified fusion calculation of multi-dimensional indicators is realized, so that the comprehensive quality score is both scientific and interpretable, and can truly reflect the overall network performance.

[0014] Furthermore, after obtaining the comprehensive quality score of the network to be evaluated, the method further includes: when the comprehensive quality score is less than a preset score threshold, matching the multidimensional indicator data with each judgment rule in a preset expert rule base; if the matching is successful, generating root causes of problems and network optimization suggestions for the network to be evaluated based on the matched judgment rules; if the matching fails, generating root causes of problems and network optimization suggestions for the network to be evaluated based on the multidimensional indicator data through a pre-trained intelligent reasoning model.

[0015] The beneficial effects of adopting the above-mentioned further solutions are: by setting a preset score threshold to trigger root cause analysis, and by adopting a dual-mode mechanism of rule-priority matching and AI fallback reasoning, it not only ensures the rapid and interpretable diagnosis of common problems, but also effectively deals with complex or new faults, significantly improving the accuracy and coverage of root cause localization.

[0016] Furthermore, the preset expert rule base includes at least one of the following judgment rules: if the physical resource block utilization rate in the multi-dimensional indicator data is greater than a first threshold and the downlink rate in the multi-dimensional indicator data is less than a second threshold, then the root cause of the problem is determined to be a rate drop due to high load, and the network optimization suggestion is to expand carrier capacity or enable load balancing; if the weak coverage grid in the measurement report in the multi-dimensional indicator data is greater than a third threshold and the signal-to-interference-plus-noise ratio in the multi-dimensional indicator data is less than a fourth threshold, then the root cause of the problem is determined to be insufficient edge coverage, and the network optimization suggestion is to adjust the synchronization signal block beam or add micro-stations; if the 5G dwell ratio in the multi-dimensional indicator data is less than a fifth threshold and the proportion of 4G terminals in the multi-dimensional indicator data is greater than a sixth threshold, then the root cause of the problem is determined to be a terminal structure problem, and the network optimization suggestion is to promote terminal upgrades or optimize 4G anchor points.

[0017] The beneficial effects of adopting the above-mentioned further solutions are: by pre-setting judgment rules and corresponding optimization suggestions for typical problems such as high load, weak coverage, and unbalanced terminal structure, it is possible to achieve real-time and accurate identification and handling of common network faults, reduce the cost of manual analysis, and improve the standardization of operation and maintenance response.

[0018] Furthermore, the method also includes: converting network optimization suggestions for the network to be evaluated into structured work order instructions; connecting to the existing network work order system through an application programming interface to automatically create and push optimization tasks according to the structured work order instructions; and tracking the execution status and optimization effect of the optimization tasks.

[0019] The beneficial effects of adopting the above-mentioned further solutions are: by converting optimization suggestions into structured work order instructions and automatically connecting them to the existing network work order system for creation, push and status tracking, a closed loop of the entire chain of evaluation, diagnosis, optimization and verification is opened up, realizing full-process automation from problem discovery to resolution, and greatly improving network operation and maintenance efficiency.

[0020] To address the aforementioned technical problems, the present invention also provides a network quality intelligent assessment system, comprising: The data acquisition module is used to acquire the scene attributes and multi-dimensional indicator data of the network to be evaluated. The multi-dimensional indicator data includes indicator data of multiple dimensions used to characterize the performance of different aspects of the network to be evaluated. The weight adaptive module is used to determine the evaluation weights corresponding to the indicator data of each dimension based on the scene attributes of the network to be evaluated. The quality assessment module is used to perform weighted fusion calculation on the indicator data of each dimension according to the assessment weights corresponding to the indicator data of each dimension, so as to obtain the comprehensive quality score of the network to be evaluated.

[0021] To address the aforementioned technical problems, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the network quality intelligent assessment method as described above.

[0022] To address the aforementioned technical problems, the present invention also provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the network quality intelligent assessment method as described above. Attached Figure Description

[0023] Figure 1 This is a flowchart of the intelligent network quality assessment method of the present invention; Figure 2 This is a flowchart of Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the intelligent network quality assessment system of the present invention; Figure 4 This is a schematic diagram of the electronic device of the present invention. Detailed Implementation

[0024] The principles and features of the present invention are described below. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0025] As mentioned earlier, existing technologies have accumulated some expertise in the field of 5G network quality assessment, but they mainly focus on single-dimensional indicator statistics, static evaluation of specific scenarios, or optimization of underlying routing algorithms. While existing solutions have addressed network coverage or capacity monitoring issues to some extent, significant technical bottlenecks remain in multi-dimensional indicator fusion, scenario-adaptive evaluation, and automated closed-loop optimization.

[0026] Example 1 Based on this, such as Figure 1 As shown in the figure, this embodiment provides a network quality intelligent assessment method, the method including: S101. Obtain the scenario attributes and multi-dimensional indicator data of the network to be evaluated. The multi-dimensional indicator data includes indicator data of multiple dimensions used to characterize the performance of different aspects of the network to be evaluated.

[0027] S102. Based on the scenario attributes of the network to be evaluated, determine the evaluation weights corresponding to the indicator data of each dimension.

[0028] S103. Based on the evaluation weights corresponding to the indicator data of each dimension, perform weighted fusion calculation on the indicator data of each dimension to obtain the comprehensive quality score of the network to be evaluated.

[0029] By acquiring the scenario attributes of the network to be evaluated and dynamically determining the evaluation weights of each dimension of the indicator data based on these attributes, the evaluation model can automatically adjust its evaluation focus according to changes in the network environment or business scenario. This overcomes the problem of evaluation results being out of touch with operational needs caused by fixed weights, achieving scenario-adaptive differentiated evaluation. Simultaneously, by using multi-dimensional indicator data containing multiple dimensions and calculating a comprehensive quality score through weighted fusion, the previously isolated indicators are organically integrated, avoiding evaluation blind spots caused by a single indicator perspective and significantly improving the accuracy and comprehensiveness of the evaluation results.

[0030] Existing assessment methods are mostly limited to a single perspective of coverage or capacity, or only perform static indicator statistics for specific scenarios. For example, some existing technologies only use the Channel Quality Indicator (CQI) to derive the Signal to Interference plus Noise Ratio (SINR) to estimate capacity, or only use drive test data to assess coverage levels, lacking a systematic framework that orthogonally integrates network structure and user experience quality. This leads to the inability to identify complex network problems such as high coverage but low data rate and structural imbalance, creating assessment blind spots.

[0031] Optionally, in the embodiments, the multi-dimensional indicator data includes capacity dimension data, coverage dimension data, structure dimension data, and quality dimension data; wherein, capacity dimension data describes the network carrying capacity; coverage dimension data describes signal reachability and quality; structure dimension data describes the rationality of network composition; and quality dimension data describes the user's actual perception.

[0032] In this embodiment, capacity dimension data includes Physical Resource Block (PRB) utilization, 5G traffic share, number of active users, and proportion of high-load cells. The proportion of high-load cells refers to the percentage of cells whose corresponding PRB utilization exceeds a preset utilization threshold (e.g., 80%).

[0033] Coverage dimension data includes the Reference Signal Receiving Power (RSRP) distribution, SINR distribution, road network coverage, and the proportion of weak coverage grids reported in the Measurement Report (MR). Road network coverage refers to the proportion of sampling points whose corresponding RSRP is greater than or equal to a preset RSRP threshold (e.g., -105dBm) and whose corresponding SINR is greater than or equal to a preset SINR threshold (e.g., -3dB).

[0034] Structural dimension data includes the proportion of 4G / 5G terminals, New Radio (NR) frequency band distribution (e.g., 700MHz / 2.1GHz / 3.5GHz), site spacing rationality, and anchor point configuration consistency. Among these, site spacing rationality refers to the deviation between the actual site spacing and the theoretical site spacing.

[0035] Quality dimension data includes downlink average rate, 5G dwell time ratio, traffic offloading ratio, and VoNR call completion rate. Among them, the traffic offloading ratio refers to the proportion of 5G traffic to the total 4G and 5G traffic.

[0036] The above metrics can be obtained directly or through simple statistical calculations by connecting to the existing OSS / BSS platform and accessing multi-source heterogeneous data such as MR (Measurement Report), PM (Performance Management), CM (Configuration Management) and drive test data, after cleaning, alignment and standardization.

[0037] In this embodiment, a four-dimensional orthogonal evaluation framework (3CQ) encompassing capacity, coverage, composition, and quality is constructed. This framework unifies and quantifies network structural and experiential indicators, enabling a comprehensive and accurate portrait of network quality. It breaks away from the traditional KPI-only approach and avoids misjudgments such as high RSRP but low data rate. In particular, the introduction of the structural dimension transcends the traditional classification that focuses solely on signal strength or capacity, achieving a quantitative assessment of potential structural risks in the network.

[0038] Existing technologies generally use fixed weights or static thresholds based on expert experience for evaluation. This approach cannot automatically adjust the evaluation focus according to changes in the network environment or business scenarios, resulting in evaluation results that are out of touch with actual operation and maintenance needs, and making it difficult to meet the requirements of precise policy implementation in different regions.

[0039] Optionally, in an embodiment, determining the evaluation weights corresponding to the indicator data of each dimension based on the scenario attributes of the network to be evaluated includes: matching the weight configuration strategy corresponding to the scenario attributes of the network to be evaluated from a preset weight template library, wherein the preset weight template library contains weight configuration strategies under different scenario combinations; and determining the evaluation weights of each dimension in the matched weight configuration strategy as the evaluation weights corresponding to the indicator data of each dimension.

[0040] Specifically, scenario attributes include region type and dominant service scenario. Region type can be pre-configured in the base station's operating parameter table, such as urban areas, suburbs, rural areas, and high-speed rail lines. Dominant service scenario can be determined based on network slice identifiers or service traffic characteristics, such as enhanced mobile broadband (eMBB), URLLC, and massive machine-type communications (mMTC).

[0041] The weight template library stores the evaluation weights for four dimensions—capacity, coverage, structure, and quality—under different scenario combinations. For example, in an urban eMBB scenario: capacity weight 0.35, coverage weight 0.25, structure weight 0.15, and quality weight 0.25; in a suburban URLLC scenario: coverage weight 0.40, quality weight 0.30, capacity weight 0.20, and structure weight 0.10.

[0042] The matched weights are used as the evaluation weights for each dimension. The weight templates support manual calibration and A / B testing verification to ensure that the evaluation results are highly consistent with actual user complaints and road test data. By establishing a scenario-adaptive dynamic weight mechanism, the optimal evaluation strategy is automatically matched according to the region type and business scenario, achieving accurate evaluation tailored to each location.

[0043] As an alternative, instead of using a pre-defined template matching method, a dynamic game theory algorithm based on real-time service load can be employed. The specific implementation logic is as follows: Real-time collection and monitoring of dynamic load indicators such as instantaneous traffic and the rate of change in active users in the target cell. Based on this real-time data, the weight coefficients for four dimensions—capacity, coverage, structure, and quality—are calculated online using a dynamic game theory algorithm, rather than querying a pre-configured fixed template library.

[0044] For example, when the instantaneous traffic and user count change rate increase significantly, the algorithm automatically increases the weight of the capacity dimension; when the user count change rate is low but the business is sensitive to latency, the algorithm can correspondingly increase the weight of the quality or coverage dimension. The weight coefficients can be updated according to a preset period to meet real-time requirements.

[0045] The advantage of this alternative is that the weights can respond instantly to dynamic changes in network load, offering greater flexibility. However, it should be noted that this approach has relatively high computational complexity, and frequent weight adjustments may lead to decreased stability of the evaluation results. In practical applications, a preset template matching method or a dynamic game algorithm can be chosen based on the specific needs of the operational scenario and the system's computing resources.

[0046] Optionally, in the embodiments, the indicator data of each dimension are weighted and fused according to the evaluation weights corresponding to the indicator data of each dimension to obtain the comprehensive quality score of the network to be evaluated. This includes: for the indicator data of each dimension, normalizing each indicator in the indicator data of the dimension to obtain the normalized value of each indicator; for the indicator data of each dimension, weighted summing of the normalized values ​​of each indicator in the indicator data of the dimension according to the preset indicator weights to obtain the dimension score of the indicator data of the dimension; and weighted summing of the dimension scores of the indicator data of each dimension according to the evaluation weights corresponding to the indicator data of each dimension to obtain the comprehensive quality score of the network to be evaluated.

[0047] First, Min-Max normalization is performed on each specific indicator within each dimension, mapping the original values ​​to the [0,1] interval. For example, for PRB utilization, its historical minimum value is 0% and its maximum value is 100%, so the normalized value of 85% PRB utilization is 0.85.

[0048] Then, for each dimension, the normalized indicator values ​​are weighted and summed according to the preset indicator weights within that dimension (e.g., PRB utilization weight 0.5, 5G traffic share weight 0.3, and high-load cell proportion weight 0.2 within the capacity dimension) to obtain the dimension score.

[0049] Finally, based on the evaluation weights of each dimension determined in the above steps, the dimensional scores of the four dimensions are weighted and summed to obtain the comprehensive quality score (which can be converted to a percentage).

[0050] Existing root cause analysis technologies are mostly limited to diagnosing specific faults and typically stop at generating alarms or outputting fault labels. They lack the ability to transform diagnostic results into specific parameter-level optimization suggestions and execute them automatically. Subsequent processing heavily relies on human experience, resulting in low operational efficiency and failing to achieve a real-time closed loop from problem discovery to resolution.

[0051] Optionally, in this embodiment, after obtaining the comprehensive quality score of the network to be evaluated, the method further includes: when the comprehensive quality score is less than a preset score threshold, matching the multidimensional indicator data with each judgment rule in a preset expert rule base; if the matching is successful, generating root causes of problems and network optimization suggestions for the network to be evaluated based on the matched judgment rules; if the matching fails, generating root causes of problems and network optimization suggestions for the network to be evaluated based on the multidimensional indicator data through a pre-trained intelligent reasoning model.

[0052] like Figure 2 As shown, when the overall quality score is less than a preset score threshold (e.g., 80 points), the root cause analysis process is triggered. First, the multidimensional indicator data is matched with the judgment rules in the preset expert rule base, which contains multiple judgment rules summarized by operation and maintenance experts.

[0053] Optionally, in the embodiments, the preset expert rule base includes at least one of the following judgment rules: if the physical resource block utilization rate in the multi-dimensional indicator data is greater than a first threshold and the downlink rate in the multi-dimensional indicator data is less than a second threshold, then the root cause of the problem is determined to be a rate drop due to high load, and the network optimization suggestion is to expand carrier capacity or enable load balancing; if the weak coverage grid in the measurement report in the multi-dimensional indicator data is greater than a third threshold and the signal-to-interference-plus-noise ratio in the multi-dimensional indicator data is less than a fourth threshold, then the root cause of the problem is determined to be insufficient edge coverage, and the network optimization suggestion is to adjust the synchronization signal block beam or add micro-stations; if the 5G dwell ratio in the multi-dimensional indicator data is less than a fifth threshold and the proportion of 4G terminals in the multi-dimensional indicator data is greater than a sixth threshold, then the root cause of the problem is determined to be a terminal structure problem, and the network optimization suggestion is to promote terminal upgrades or optimize 4G anchor points.

[0054] For example, if the PRB utilization rate is >80% and the downlink speed is <300Mbps, the root cause is determined to be the speed drop due to high load, and the optimization suggestion is to expand the carrier or enable load balancing; if the MR weak coverage grid ratio is >15% and SINR is <-3dB, the root cause is determined to be insufficient edge coverage, and the optimization suggestion is to adjust the Synchronization Signal Block (SSB) beam or add micro base stations; if the 5G dwell ratio is <50% and the 4G terminal ratio is >60%, the root cause is determined to be the terminal structure problem, and the optimization suggestion is to promote terminal upgrades or optimize 4G anchor points.

[0055] If a match is successful, the corresponding root cause and optimization suggestions are directly output. If no rule is matched, the multi-dimensional indicator data is input into the pre-trained intelligent inference model. The pre-trained intelligent inference model can be trained from existing conventional models, such as the lightweight XGBoost classification model, or decision trees, random forests, etc. Conventional model training methods are used to train the model with a large amount of historical fault data, enabling the model to output the root cause label with the highest probability based on the input feature vector (i.e., each indicator value), such as interference-dominated degradation, frequent switching-related disconnections, etc., and provide corresponding optimization suggestions through pre-defined mapping relationships. The final results can be displayed through a web interface.

[0056] This solution balances interpretability and generalization. Prioritizing rules ensures interpretability, making it trustworthy for operations personnel. AI as a fallback provides generalization capabilities, enabling it to handle novel and complex problems. Therefore, its performance significantly outperforms pure rule-based solutions or purely black-box AI solutions.

[0057] As an alternative implementation, a pure expert system or a pure AI model can be used to replace the dual-mode architecture described above. A pure expert system relies solely on an If-Else rule base for matching, while a pure AI model relies solely on a machine learning model (such as a decision tree) for reasoning. However, pure expert systems struggle to cover complex scenarios and are prone to missed judgments. Pure AI models lack interpretability, leading to low trust levels among operations and maintenance personnel. The dual-mode architecture of this method overcomes these shortcomings simultaneously and is a superior implementation.

[0058] Optionally, in an embodiment, the method further includes: converting network optimization suggestions for the network to be evaluated into structured work order instructions; connecting to the existing network work order system through an application programming interface to automatically create and push optimization tasks according to the structured work order instructions; and tracking the execution status and optimization effect of the optimization tasks.

[0059] Specifically, the generated network optimization suggestions are converted into structured work order instructions. Then, through an application programming interface (API), it connects to the operator's existing network work order system (such as iOMS) to automatically create and push optimization tasks to the corresponding city-level optimization engineers based on the work order instructions.

[0060] For example, if the system identifies "high load in XX community of Gulou District", it will automatically generate a work order that "suggests expanding capacity by 2 100MHz NR carriers, which is expected to increase capacity by 40%", along with simulation prediction results.

[0061] Furthermore, the execution status of work orders is continuously tracked, and the effect is verified by re-executing steps S101-S103 after optimization. By constructing a root cause localization engine that combines rules with AI dual-mode, the system achieves automatic generation and output of optimization suggestions at the specific parameter level, from problem identification to specific parameter level, thus creating a fully automated closed loop of evaluation, diagnosis, suggestion, and execution.

[0062] This technical solution has been verified on live networks, with 1200 5G macro base station cells selected as pilot sites. It processes an average of 568 million MR sampling data entries daily, covering over 98% of 5G users in the area. Based on the evaluation and diagnostic process of this method, 12 high-load, weak-coverage areas and 8 areas with unbalanced terminal structures were automatically identified. Based on the root cause localization results, optimization suggestions were generated and implemented: two carriers were added to the high-load cells, and the SSB beam direction was adjusted in the weak-coverage areas.

[0063] Compared with before and after optimization, key indicators in the pilot area have improved significantly: network coverage increased from 85% to 90%, an increase of 5 percentage points; the average 5G downlink speed increased from 420Mbps to 470Mbps, an increase of 50Mbps; and the 5G session retention rate increased from 46.26% to 49.43%, an increase of 3.17 percentage points.

[0064] This solution is developed using the common Java / Python languages ​​and can seamlessly integrate with existing OSS / BSS platforms and iOMS ticketing systems without requiring modifications to the underlying hardware. The proposed 3CQ evaluation model and closed-loop optimization process align with the current trend in the communications industry towards intelligent networks, possessing the foundation for conversion into industry or enterprise standards. This method is not only theoretically innovative but also mature in engineering practice. Live network verification has demonstrated the effectiveness of the technical solution, exhibiting extremely high feasibility for implementation and large-scale deployment. It can be directly applied to live network operations and maintenance, significantly improving 5G network quality and operational efficiency.

[0065] Example 2 like Figure 3 As shown, this embodiment provides a network quality intelligent assessment system 200, including: The data acquisition module 201 is used to acquire the scene attributes and multi-dimensional indicator data of the network to be evaluated. The multi-dimensional indicator data includes indicator data of multiple dimensions used to characterize the performance of different aspects of the network to be evaluated. The weight adaptive module 202 is used to determine the evaluation weights corresponding to the indicator data of each dimension based on the scenario attributes of the network to be evaluated. The quality assessment module 203 is used to perform weighted fusion calculation on the indicator data of each dimension according to the assessment weights corresponding to the indicator data of each dimension, so as to obtain the comprehensive quality score of the network to be evaluated.

[0066] Optionally, in the embodiments, the multi-dimensional indicator data includes capacity dimension data, coverage dimension data, structure dimension data, and quality dimension data; wherein, capacity dimension data describes the network carrying capacity; coverage dimension data describes signal reachability and quality; structure dimension data describes the rationality of network composition; and quality dimension data describes the user's actual perception.

[0067] Optionally, in an embodiment, the weight adaptation module 202 includes: The strategy matching unit is used to match the weight configuration strategy corresponding to the scene attributes of the network to be evaluated from the preset weight template library. The preset weight template library contains weight configuration strategies under different scene combinations. The weight determination unit is used to determine the evaluation weights of each dimension in the matched weight configuration strategy as the evaluation weights corresponding to the indicator data of each dimension.

[0068] Optionally, in an embodiment, the quality assessment module 203 includes: The normalization unit is used to normalize each indicator in the indicator data of each dimension, and obtain the normalized value of each indicator. The dimension score calculation unit is used to calculate the weighted sum of the normalized values ​​of each indicator in the dimension's indicator data according to the preset indicator weights, so as to obtain the dimension score of the dimension's indicator data. The quality scoring unit is used to perform a weighted sum of the dimensional scores of the indicator data in each dimension based on the evaluation weights corresponding to the indicator data in each dimension, so as to obtain the comprehensive quality score of the network to be evaluated.

[0069] Optionally, in this embodiment, after obtaining the comprehensive quality score of the network to be evaluated, the method further includes: when the comprehensive quality score is less than a preset score threshold, matching the multidimensional indicator data with each judgment rule in a preset expert rule base; if the matching is successful, generating root causes of problems and network optimization suggestions for the network to be evaluated based on the matched judgment rules; if the matching fails, generating root causes of problems and network optimization suggestions for the network to be evaluated based on the multidimensional indicator data through a pre-trained intelligent reasoning model.

[0070] Optionally, in the embodiments, the preset expert rule base includes at least one of the following judgment rules: if the physical resource block utilization rate in the multi-dimensional indicator data is greater than a first threshold and the downlink rate in the multi-dimensional indicator data is less than a second threshold, then the root cause of the problem is determined to be a rate drop due to high load, and the network optimization suggestion is to expand carrier capacity or enable load balancing; if the weak coverage grid in the measurement report in the multi-dimensional indicator data is greater than a third threshold and the signal-to-interference-plus-noise ratio in the multi-dimensional indicator data is less than a fourth threshold, then the root cause of the problem is determined to be insufficient edge coverage, and the network optimization suggestion is to adjust the synchronization signal block beam or add micro-stations; if the 5G dwell ratio in the multi-dimensional indicator data is less than a fifth threshold and the proportion of 4G terminals in the multi-dimensional indicator data is greater than a sixth threshold, then the root cause of the problem is determined to be a terminal structure problem, and the network optimization suggestion is to promote terminal upgrades or optimize 4G anchor points.

[0071] Optionally, in an embodiment, the system further includes: The instruction conversion module is used to convert network optimization suggestions for the network to be evaluated into structured chemical instructions. The task creation module is used to connect to the existing work order system through an application programming interface to automatically create and push optimized tasks based on structured work order instructions; The tracking module is used to track the execution status and optimization effects of optimization tasks.

[0072] In some embodiments, the network quality intelligent assessment system 200 of the present invention can be implemented in a combination of hardware and software. As an example, the network quality intelligent assessment system 200 of the present invention can be a processor in the form of a hardware decoding processor, which is programmed to execute the network quality intelligent assessment method of the present invention. For example, the processor in the form of a hardware decoding processor can be one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0073] The modules described in the embodiments of this invention can be implemented in software or hardware. The names of the modules are not, in some cases, limiting the scope of the module itself.

[0074] Example 3 like Figure 4 As shown, this embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the network quality intelligent assessment method as described in Embodiment 1.

[0075] In other words, an electronic device according to an embodiment of the present invention may include, but is not limited to, a processor and a memory; the memory is used to store computer programs; the processor is used to execute the network quality intelligent assessment method shown in any embodiment of the present invention by calling the computer programs.

[0076] In one alternative embodiment, an electronic device is provided. Figure 4 The illustrated electronic device 300 includes a processor 301 and a memory 303. The processor 301 and the memory 303 are connected, for example, via a bus 302. Optionally, the electronic device 300 may further include a transceiver 304, which can be used for data interaction between the electronic device and other electronic devices, such as sending and / or receiving data. It should be noted that in practical applications, the transceiver 304 is not limited to one type, and the structure of the electronic device 300 does not constitute a limitation on the embodiments of the present invention.

[0077] Processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. Processor 301 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0078] Bus 302 may include a path for transmitting information between the aforementioned components. Bus 302 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 302 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The bus 302 is represented by only one thick line, but this does not mean that there is only one bus or one type of bus.

[0079] The memory 303 may be a ROM (Read Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0080] The memory 303 is used to store application code (computer program) for executing the present invention, and its execution is controlled by the processor 301. The processor 301 is used to execute the application code stored in the memory 303 to implement the content shown in the foregoing method embodiments.

[0081] Among them, electronic devices can also be terminal devices, which can be any device that can install applications, including at least one of smartphones, tablets, laptops, desktop computers, smart speakers, smartwatches, smart TVs, and smart in-vehicle devices.

[0082] It should be noted that, Figure 4 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0083] Example 4 This embodiment provides a non-transitory computer-readable storage medium that stores computer instructions for causing a computer to execute the network quality intelligent assessment method as described in Embodiment 1.

[0084] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0085] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the aforementioned intelligent network quality assessment method.

[0086] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer. In cases involving a remote computer, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0087] It should be understood that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0088] The computer-readable storage medium provided in this invention can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EEPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0089] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the method shown in the above embodiments.

[0090] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0091] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0092] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this invention can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0093] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A network quality intelligent assessment method, characterized in that, The method includes: Obtain the scenario attributes and multidimensional indicator data of the network to be evaluated. The multidimensional indicator data includes indicator data of multiple dimensions used to characterize the performance of different aspects of the network to be evaluated. Based on the scenario attributes of the network to be evaluated, determine the evaluation weights corresponding to the indicator data of each dimension; Based on the evaluation weights corresponding to the indicator data of each dimension, the indicator data of each dimension are weighted and fused to obtain the comprehensive quality score of the network to be evaluated.

2. The intelligent network quality assessment method according to claim 1, characterized in that, The multi-dimensional indicator data includes capacity dimension data, coverage dimension data, structure dimension data, and quality dimension data; wherein, the capacity dimension data describes the network carrying capacity; the coverage dimension data describes the signal reachability and quality; the structure dimension data describes the rationality of the network composition; and the quality dimension data describes the actual user perception.

3. The intelligent network quality assessment method according to claim 1, characterized in that, The step of determining the evaluation weights corresponding to the indicator data of each dimension based on the scenario attributes of the network to be evaluated includes: The weight configuration strategy corresponding to the scene attributes of the network to be evaluated is matched from a preset weight template library, which contains weight configuration strategies for different scene combinations. The evaluation weights of each dimension in the matched weight configuration strategy are determined as the evaluation weights corresponding to the indicator data of each dimension.

4. The intelligent network quality assessment method according to claim 1, characterized in that, The step involves weighting and fusing the indicator data of each dimension according to the evaluation weights corresponding to the indicator data of each dimension to obtain the comprehensive quality score of the network to be evaluated, including: For the indicator data of each dimension, each indicator in the indicator data of that dimension is normalized to obtain the normalized value of each indicator. For each dimension of indicator data, the normalized values ​​of each indicator in the indicator data of that dimension are weighted and summed according to the preset indicator weights to obtain the dimension score of the indicator data of that dimension. Based on the evaluation weights corresponding to the indicator data of each dimension, the dimensional scores of the indicator data of each dimension are weighted and summed to obtain the comprehensive quality score of the network to be evaluated.

5. The intelligent network quality assessment method according to claim 1, characterized in that, After obtaining the overall quality score of the network to be evaluated, the following steps are also included: When the overall quality score is less than a preset score threshold, the multidimensional index data is matched with each judgment rule in the preset expert rule base; If a match is successful, then based on the matched judgment rules, root causes of problems and network optimization suggestions for the network to be evaluated are generated. If the matching fails, the pre-trained intelligent reasoning model will be used to make inferences and judgments based on the multi-dimensional index data to generate root causes of problems and network optimization suggestions for the network to be evaluated.

6. The intelligent network quality assessment method according to claim 5, characterized in that, The preset expert rule base includes at least one of the following decision rules: If the physical resource block utilization rate in the multidimensional indicator data is greater than the first threshold and the downlink rate in the multidimensional indicator data is less than the second threshold, then the root cause of the problem is determined to be the rate drop caused by high load, and the network optimization suggestion is to expand the carrier or enable load balancing. If the weak coverage grid in the measurement report of the multidimensional index data is greater than the third threshold and the signal-to-interference-plus-noise ratio in the multidimensional index data is less than the fourth threshold, then the root cause of the problem is determined to be insufficient edge coverage, and the network optimization suggestion is to adjust the synchronization signal block beam or add a micro station. If the 5G retention rate in the multidimensional indicator data is less than the fifth threshold and the proportion of 4G terminals in the multidimensional indicator data is greater than the sixth threshold, then the root cause of the problem is determined to be a terminal structure problem, and the network optimization suggestion is to promote terminal upgrades or optimize 4G anchor points.

7. The intelligent network quality assessment method according to claim 5, characterized in that, Also includes: The network optimization suggestions for the network to be evaluated are converted into structural chemical single instructions; By connecting to the existing work order system through the application programming interface, optimized tasks can be automatically created and pushed according to the structured work order instructions; Track the execution status and optimization effect of the optimization task.

8. A network quality intelligent assessment system, characterized in that, include: The data acquisition module is used to acquire the scene attributes and multi-dimensional indicator data of the network to be evaluated. The multi-dimensional indicator data includes indicator data of multiple dimensions used to characterize the performance of different aspects of the network to be evaluated. The weight adaptive module is used to determine the evaluation weights corresponding to the indicator data of each dimension based on the scene attributes of the network to be evaluated. The quality assessment module is used to perform weighted fusion calculation on the indicator data of each dimension according to the assessment weights corresponding to the indicator data of each dimension, so as to obtain the comprehensive quality score of the network to be evaluated.

9. An electronic device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the network quality intelligent assessment method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to execute the network quality intelligent assessment method according to any one of claims 1 to 7.

Images