Information quality optimization method, device and equipment based on federated learning

Abstract

The present invention provides a method, apparatus, and device for optimizing information quality based on federated learning. The method comprises: obtaining a root cause analysis request sent by a service provider, the root cause analysis request carrying service quality data and service configuration data when the service quality is abnormal; inputting the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; an optimization strategy is generated based on the root cause analysis result; and the service quality and / or network quality are optimized based on the optimization strategy. The solution of the present invention optimizes service quality and / or network quality with high security, high accuracy, high efficiency, and a complete process.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method, apparatus and device for optimizing information quality based on federated learning. Background Art

[0002] In QoE optimization based on RIC (RAN Intelligent Control, wireless intelligent control platform), it is difficult to analyze the actual cause of the jamming based solely on the jamming duration of the QoE prediction results, which is insufficient to generate an effective SLA guarantee strategy and difficult to form a closed-loop SLA guarantee process. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide an information quality optimization method, device and equipment based on federated learning, which obtains a root cause analysis model through vertical federated modeling of federated learning, thereby optimizing service quality and / or network quality with high security, high accuracy, high efficiency and perfect process.

[0004] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0005] A federated learning-based information quality optimization method, applied to a wireless intelligent control platform, includes:

[0006] Obtaining a root cause analysis request sent by a service provider, the root cause analysis request carrying service quality data and service configuration data when the service quality is abnormal;

[0007] Inputting network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning;

[0008] generating an optimization strategy based on the root cause analysis results;

[0009] The service quality and / or network quality are optimized according to the optimization strategy.

[0010] Optionally, the root cause analysis model is obtained through longitudinal federated modeling of federated learning, including:

[0011] When the service provider monitors that the service quality is abnormal, receiving a network quality data collection request sent by the service provider;

[0012] Obtaining sample data of network quality data when service quality is abnormal according to the network quality data collection request;

[0013] Based on expert experience, sample data of the network quality data and sample data of the service provider's service quality data are analyzed to obtain a service quality anomaly label;

[0014] Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

[0015] Optionally, before obtaining the root cause analysis model through longitudinal federated modeling of federated learning, the following steps may be further included:

[0016] Negotiate service level agreement (SLA) guarantee requirements with the service provider;

[0017] Configure the correspondence between services and SLA requirements according to the SLA guarantee requirements;

[0018] The service provider monitors the service quality in real time to see if it is abnormal based on the correspondence between the service and the SLA requirements.

[0019] Optionally, based on expert experience, sample data of the network quality data and sample data of the service provider's service quality data are analyzed to obtain a service quality anomaly label, including:

[0020] Based on expert experience, sample data of the network quality data and sample data of the service quality data of the service provider are analyzed to obtain the cause of the abnormality, and the cause of the abnormality is marked to obtain a service quality abnormality label; the cause of the abnormality includes at least one of the following: abnormal service configuration, poor terminal channel environment, wireless interference, abnormal wireless access network RAN ​​configuration, wireless resource congestion, core network abnormality, and transmission network abnormality.

[0021] Optionally, generating an optimization strategy based on the root cause analysis results includes:

[0022] Generating a first optimization strategy for service quality based on the root cause analysis results of the service quality; and / or

[0023] Based on the root cause analysis results of the network quality, a second optimization strategy for the network quality is generated.

[0024] Optionally, optimizing service quality and / or network quality according to the optimization strategy includes:

[0025] sending the first optimization strategy to a service provider, so that the service provider optimizes the service according to the first optimization strategy; and / or

[0026] The second optimization strategy is sent to the network side, so that the network side optimizes the network according to the second optimization strategy.

[0027] The present invention also provides an information quality optimization method based on federated learning, which is applied to a service provider. The method includes:

[0028] Obtain real-time service quality data sent by the terminal;

[0029] According to the correspondence between the real-time service quality data and the configured services and the service level agreement (SLA) requirements, when a service quality abnormality is detected, a root cause analysis request is sent to the wireless intelligent control platform; the root cause analysis request carries the service quality data and service configuration data at the time of the service quality abnormality;

[0030] An optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request is received, wherein the optimization strategy is that the wireless intelligent control platform inputs the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result, and is generated based on the root cause analysis result; wherein the root cause analysis model is obtained through vertical federated modeling of federated learning.

[0031] Optionally, the root cause analysis model is obtained through longitudinal federated modeling of federated learning, including:

[0032] Obtaining sample data of service quality data when service quality is abnormal and sample data of network quality data on the terminal side;

[0033] Based on expert experience, sample data of the service quality data and sample data of the network quality data on the terminal side are analyzed to obtain a service quality anomaly label;

[0034] Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

[0035] Optionally, the SLA requirement includes at least one of the following:

[0036] Network SLA requirements;

[0037] Business SLA requirements: Different business types have different SLA requirements.

[0038] The present invention also provides a quality optimization device based on federated learning, which is applied to a wireless intelligent control platform. The device includes:

[0039] An acquisition module is used to acquire a root cause analysis request sent by a service provider, wherein the root cause analysis request carries service quality data and service configuration data when the service quality is abnormal;

[0040] A processing module is used to input the network quality data and network configuration data on the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; an optimization strategy is generated based on the root cause analysis result; and the service quality and / or network quality is optimized according to the optimization strategy.

[0041] The present invention also provides an information quality optimization device based on federated learning, which is applied to a service provider, and the device includes:

[0042] An acquisition module is used to acquire real-time service quality data sent by the terminal;

[0043] a transceiver module configured to send a root cause analysis request to the wireless intelligent control platform when abnormal service quality is detected based on the correspondence between the real-time service quality data and the configured service and service level agreement (SLA) requirements; the root cause analysis request carries the service quality data and service configuration data at the time of abnormal service quality; and

[0044] An optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request is received, wherein the optimization strategy is that the wireless intelligent control platform inputs the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result, and is generated based on the root cause analysis result; wherein the root cause analysis model is obtained through vertical federated modeling of federated learning.

[0045] The present invention further provides a communication device, comprising: a processor and a memory storing a computer program, wherein the computer program executes the method described above when executed by the processor.

[0046] The present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above.

[0047] The above solution of the present invention includes at least the following beneficial effects:

[0048] The above-mentioned solution of the present invention obtains a root cause analysis request sent by a service provider, and the root cause analysis request carries service quality data and service configuration data when the service quality is abnormal; the network quality data and network configuration data on the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, are input into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; an optimization strategy is generated based on the root cause analysis result; and the service quality and / or network quality is optimized based on the optimization strategy. The root cause analysis model is used to process the network quality data and network configuration data on the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, to obtain a root cause analysis result. A corresponding optimization strategy is further generated based on the root cause analysis result, and the service quality and / or network quality are optimized according to the optimization strategy. This realizes federated reasoning through the terminal service experience quality, wireless network quality data and configuration data when the service quality fails to meet the SLA guarantee requirements (or when a service failure occurs), analyzes the root cause, and generates a wireless network optimization strategy in real time, with high security, high accuracy, high efficiency and a complete process. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a flowchart of a federated learning-based information quality optimization method for a wireless intelligent control platform provided by an embodiment of the present invention;

[0050] Figure 2 This is a sample collection flow chart provided by an embodiment of the present invention when a service provider monitors abnormal service quality;

[0051] Figure 3 is a flow chart of SLA configuration according to an embodiment of the present invention;

[0052] Figure 4 is a flowchart of root cause analysis federated modeling according to an embodiment of the present invention;

[0053] Figure 5 is a schematic diagram of federated modeling of a service provider and a wireless intelligent control platform according to an embodiment of the present invention;

[0054] Figure 6 This is a schematic diagram of the main functional modules of the service provider, wireless intelligent control platform and terminal according to an embodiment of the present invention;

[0055] Figure 7 is a flowchart of a method for optimizing information quality based on federated learning applied to a service provider according to an embodiment of the present invention;

[0056] Figure 8This is a specific flow chart of a method for optimizing information quality based on federated learning according to an embodiment of the present invention;

[0057] Figure 9 It is a module diagram of a quality optimization device based on federated learning according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The following describes exemplary embodiments of the present disclosure in more detail with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0059] like Figure 1 As shown, an embodiment of the present invention provides an information quality optimization method based on federated learning, which is applied to a wireless intelligent control platform. The method includes:

[0060] Step 11: Obtain a root cause analysis request sent by the service provider, wherein the root cause analysis request carries service quality data and service configuration data when the service quality is abnormal;

[0061] Step 12: Inputting the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning;

[0062] Step 13: generating an optimization strategy based on the root cause analysis results;

[0063] Step 14: Optimize service quality and / or network quality according to the optimization strategy.

[0064] In this embodiment, the wireless intelligent control platform obtains a root cause analysis request from a service provider, carrying service quality data when service quality is abnormal. The wireless intelligent control platform inputs the service quality data and network quality data when service quality is abnormal into a pre-established root cause analysis model obtained through vertical federated modeling using federated learning, performs root cause processing, obtains a root cause analysis result, generates an optimization strategy based on the root cause analysis result, and optimizes service quality and / or network quality according to the optimization strategy. In this way, a root cause analysis model is obtained through vertical federated modeling using federated learning. The root cause analysis model processes the network quality data and network configuration data on the base station side when service quality is abnormal, as well as the service quality data and service configuration data of the service provider, to obtain a root cause analysis result. Furthermore, a corresponding optimization strategy is generated based on the root cause analysis result, and service quality and / or network quality is optimized according to the optimization strategy. This achieves federated reasoning based on terminal service experience quality and wireless network quality data when service quality fails to meet SLA requirements (or when a service failure occurs), analyzes the root cause, and generates a wireless network optimization strategy in real time, with high security, high accuracy, high efficiency, and a complete process.

[0065] Here, the wireless intelligent control platform may include multiple main functional modules. Specifically, the wireless intelligent control platform may include at least one of the following modules: a service level agreement (SLA) management module, a data collection module, a federated learning module, a root cause analysis module, and an assurance implementation module.

[0066] Among them, the service SLA management module exists in the form of xAPP (software) and is used to maintain the correspondence between service quality data and SLA requirements after the operator and the service provider negotiate the SLA guarantee agreement.

[0067] The data acquisition module is used to collect and maintain network quality data and network configuration data reported by the network side. The network quality data includes but is not limited to: base station side UE (User Equipment, terminal) channel information, protocol stack status, cell status, etc.; the network configuration data includes but is not limited to: QoS (Quality of Service) priority, base station pre-scheduling, frame structure, etc.

[0068] Federated learning module: Exists in the form of xAPP (software). The federated learning module of the wireless intelligent control platform is used to jointly train the root cause analysis model with the federated learning module of the service provider.

[0069] Root cause analysis module: exists in the form of xAPP (software). When a network anomaly occurs, it analyzes the real-time service quality data and network quality data to determine the root cause of the network anomaly.

[0070] Assurance implementation module: Used to determine the root cause analysis results. If it is a business-side problem, the business side will be notified for processing. If it is a network-side problem, a network optimization strategy will be generated and configured.

[0071] In an optional embodiment of the present invention, the root cause analysis model in step 12 is obtained through vertical federated modeling of federated learning, specifically including:

[0072] Step 12-1: When the service provider monitors that the service quality is abnormal, receiving a network quality data collection request sent by the service provider;

[0073] Step 12-2, obtaining sample data of network quality data when service quality is abnormal according to the network quality data collection request;

[0074] Step 12-3: Analyze the sample data of the network quality data and the sample data of the service quality data of the service provider based on expert experience to obtain a service quality abnormality label;

[0075] Step 12-4: performing vertical federated modeling of federated learning based on the abnormal service quality label to obtain the root cause analysis model.

[0076] In this embodiment, the service provider continuously monitors the terminal service quality through service quality data collected from the terminal. When the service provider monitors a service quality anomaly, where the service quality anomaly refers to service quality that is worse than the SLA requirement, the wireless intelligent control platform receives a network quality data collection request sent by the service provider. Based on the network quality data collection request, the wireless intelligent control platform obtains sample data of the network quality data when the service quality anomaly occurs, where the sample data includes network quality data and network configuration data on the network side at the time the anomaly occurs. Based on expert experience, the wireless intelligent control platform analyzes the sample data of the network quality data and the sample data of the service provider's service quality data to obtain a service quality anomaly label. The wireless intelligent control platform performs vertical federated modeling of federated learning based on the service quality anomaly label to obtain the root cause analysis model.

[0077] It should be noted that after the service provider detects abnormal service quality, it may also:

[0078] The service provider records at least one of the real-time service quality data, the terminal network configuration data, the terminal service configuration data, and the exception number.

[0079] like Figure 2As shown, in a specific embodiment of the present invention, a service provider continuously monitors the terminal service quality by collecting service quality data from the terminal. When the service provider detects that the service quality is worse than the SLA requirement, the service provider records the real-time service quality data, the terminal network configuration data, the terminal service configuration data, and the exception number. In addition, the wireless intelligent control platform receives a network data collection request initiated by the service provider. Based on the network quality data collection request, the wireless intelligent control platform obtains the network quality data and network configuration data on the network side at the time of the exception. Based on expert experience, the wireless intelligent control platform analyzes sample data of the network quality data and sample data of the service provider's service quality data to obtain a service quality anomaly label. The wireless intelligent control platform performs vertical federated modeling of federated learning based on the service quality anomaly label to obtain the root cause analysis model. By jointly establishing the root cause analysis model with service quality data and network quality data, the root cause analysis model is more accurate. The root cause analysis model is obtained through vertical federated modeling of federated learning, without data security risks, and has high security.

[0080] In another optional embodiment of the present invention, before obtaining the root cause analysis model through vertical federated modeling of federated learning, an SLA configuration process may also be included:

[0081] like Figure 2 As shown, the process includes:

[0082] Step 12-01, negotiate service level agreement (SLA) guarantee requirements with the service provider;

[0083] Step 12-02, configuring the correspondence between the service and the SLA requirement according to the SLA guarantee requirement;

[0084] In step 12-03, the service provider monitors the service quality in real time to see if it is abnormal based on the correspondence between the service and the SLA requirements.

[0085] In this embodiment, the wireless intelligent control platform negotiates with the service provider about the service level agreement (SLA) guarantee requirements; the wireless intelligent control platform and the service provider configure a correspondence between the service and the SLA requirements based on the SLA guarantee requirements; and the service provider monitors the service quality in real time for abnormalities based on the correspondence between the service and the SLA requirements.

[0086] In another optional embodiment of the present invention, step 12-3 includes:

[0087] Step 12-31: Analyze the sample data of the network quality data and the sample data of the service quality data of the service provider based on expert experience to obtain the abnormal cause, mark the abnormal cause, and obtain a service quality abnormality label;

[0088] The abnormal reason includes at least one of the following: abnormal service configuration, poor terminal channel environment, wireless interference, abnormal radio access network RAN ​​configuration, wireless resource congestion, core network abnormality, and transmission network abnormality.

[0089] Here, as Figure 3 As shown, the process of collecting sample data of network quality data and sample data of service quality data by the service provider includes:

[0090] After the service is launched, the service provider continuously monitors the terminal service quality through terminal collection. When a quality abnormality occurs (the service quality is worse than the SLA requirement), the service provider records the real-time service quality data, terminal network configuration data, terminal service configuration data and abnormality number. At the same time, the wireless intelligent control platform interface is called to record the network quality data and network configuration data on the network side at the time of the abnormality.

[0091] Through expert experience analysis, the causes of the anomalies are determined, including abnormal service configuration, poor UE channel environment, wireless interference, abnormal RAN configuration (QoS, base station pre-scheduling, frame structure, SR cycle, etc.), wireless resource congestion, and other anomalies (core network, transmission network problems). The anomaly causes are used as labels.

[0092] like Figure 4 and Figure 5 As shown in Figure 2, the root cause analysis federated modeling process includes: when there is sufficient sample data, the operator and service provider load the sample data into their respective federated learning modules;

[0093] The operator initiates vertical federated modeling for root cause analysis from the federated learning module of the wireless intelligent control platform. The operator serves as the modeler and the service provider serves as the data provider. They can select an algorithm suitable for multi-classification for model training. This example uses the random forest algorithm for training.

[0094] First, both parties read the sample data and use anomaly numbers to align the samples; N sub-sample sets are randomly selected from the aligned samples, and the gini algorithm is used as the feature selection algorithm. Decision trees are created using the N sub-sample sets to complete the root cause analysis multi-classification model training.

[0095] After model training is completed, the service provider and operator deploy sub-models respectively, and the operator publishes the model to generate online inference services for service providers to call.

[0096] In another optional embodiment of the present invention, step 13 may include:

[0097] Step 131: generating a first optimization strategy for service quality based on the root cause analysis result of service quality; and / or

[0098] Step 132: Generate a second optimization strategy for network quality based on the root cause analysis result of the network quality.

[0099] In this embodiment, the first optimization strategy here is an optimization strategy for service quality; the second optimization strategy here is an optimization strategy for network quality. The first optimization strategy can be used to optimize service quality, and the second optimization strategy can be used to optimize network quality.

[0100] In another optional embodiment of the present invention, step 14 may include:

[0101] Step 141: Send the first optimization strategy to the service provider, so that the service provider optimizes the service according to the first optimization strategy; and / or

[0102] Step 142: Send the second optimization strategy to the network side, so that the network side optimizes the network according to the second optimization strategy.

[0103] Specifically, such as Figure 6 As shown in the figure, the SLA guarantee and exception analysis process includes:

[0104] The terminal collects service quality data;

[0105] The service provider monitors service quality and initiates root cause analysis when an anomaly occurs (service quality does not meet SLA requirements). The request should include real-time service quality data, terminal network configuration data, and terminal service configuration data.

[0106] The wireless intelligent control platform uses a federated learning module in conjunction with the service provider's federated learning module to use a root cause analysis model for online reasoning to determine the root cause of anomalies. The data uses the service provider's real-time service quality data, terminal network configuration data, terminal service configuration data, and the wireless intelligent control platform's network quality and configuration data.

[0107] The wireless intelligent control platform uses the network optimization module to derive service quality optimization strategies based on the root causes of abnormalities;

[0108] Execute the optimization strategy to the service provider or base station respectively according to the optimization strategy.

[0109] In this embodiment, since the service quality is the service quality of the service provider, the first optimization strategy is sent to the service provider, and the service provider optimizes the service quality according to the first optimization strategy; since the network quality is the network quality on the network side, the second optimization strategy is sent to the network side, and the network side optimizes the network quality according to the second optimization strategy.

[0110] The above-mentioned embodiments of the present invention use federated learning for model training, eliminating data security risks, making them easier to implement for industry customers and providing high security. By jointly modeling network-side and business-side data, the model accuracy is even higher.

[0111] Compared with the QoE prediction solution, the SLA guarantee process is more complete. After QoE determines that a service is abnormal, it is difficult to perform root cause analysis due to the lack of business-side data.

[0112] This solution is based on a wireless intelligent control platform, and its millisecond-level data quality can provide faster business assurance. This solution securely aggregates the data value of business and network parties through federated learning, replacing the original solutions that rely solely on business data to predict network capabilities or solely on network data to predict business quality, greatly improving SLA assurance efficiency.

[0113] like Figure 7 As shown, in a specific embodiment of the present invention, a system for optimizing service and / or network quality based on federated learning may include: a service provider, a wireless intelligent control platform, a terminal, a base station, and a core network and its main functional modules;

[0114] The main functional modules of the service provider include:

[0115] Service SLA management module: used for maintaining SLA guarantee requirement information in the service SLA management module after the service provider and the operator sign the service SLA agreement;

[0116] The SLA guarantee requirement information includes but is not limited to: business SLA requirements, indicators in the business SLA requirements include but are not limited to: network latency, uplink and downlink rates, packet loss rate, latency jitter, etc., as well as business-class first frame display duration, frame loss rate, average freeze duration, MOS (Mean Opinion Score), etc., where business-class indicators vary according to business type.

[0117] Service quality monitoring module: used to monitor the service quality data collected from the terminal; and compare it with the SLA requirements corresponding to the service quality data. If the service quality data does not meet the SLA requirements, it will be used as sample data in the training phase to initiate service anomaly root cause analysis in the assurance phase.

[0118] Federated Learning Module: Serves as a terminal-side federated learning engine, supporting decentralized vertical federated modeling and online reasoning. It is used to jointly train root cause analysis models with the federated learning module of the wireless intelligent control platform.

[0119] The main functional modules of the wireless intelligent control platform (RIC) include:

[0120] Service SLA management module: exists in the form of xAPP and is used to maintain the correspondence between service quality data and SLA requirements after the operator and the service provider negotiate the SLA guarantee agreement.

[0121] Data acquisition module: used to collect and maintain network quality data and network configuration data reported by the network side. The network quality data includes but is not limited to: base station side UE channel information, protocol stack status, cell status, etc.; the network configuration data includes but is not limited to: QoS priority, base station pre-scheduling, frame structure, etc.

[0122] Federated Learning Module: This module exists in the form of an xAPP and serves as a network-side federated learning engine. It supports decentralized vertical federated modeling and online reasoning, and is used to train root cause analysis models together with the service provider's federated learning module.

[0123] Root Cause Analysis Module: This module exists in the form of an xAPP. When a network anomaly occurs, it performs real-time reasoning based on real-time service quality data and network quality data to analyze the root cause of the network anomaly.

[0124] Assurance implementation module: Used to determine the root cause analysis results. If it is a business-side problem, the business side will be notified for processing. If it is a network-side problem, a network optimization strategy will be generated and configured.

[0125] The main functional modules of the terminal include:

[0126] Service quality collection module: used to collect real-time service quality data, terminal-side network quality data, terminal network configuration information, and service configuration information determined by service type, and upload the collected data to the service provider;

[0127] The terminal-side network quality data includes but is not limited to: RSRP (Reference Signal Received Power), RSRQ (Reference Signal Received Quality), SINR (signal-to-noise and interference ratio), MCS (Modulation and coding scheme), RANK, etc.;

[0128] The terminal network configuration information includes but is not limited to: terminal uplink and downlink speed limits;

[0129] The service configuration information includes but is not limited to: video bit rate, FPS (Frames Per Second), control instruction packet sending frequency, data packet collection frequency, etc.

[0130] An embodiment of the present invention obtains a root cause analysis request sent by a service provider, the root cause analysis request carrying service quality data when the service quality is abnormal; the service quality data and network quality data when the service quality is abnormal are input into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; an optimization strategy is generated based on the root cause analysis result; and the service quality and / or network quality are optimized based on the optimization strategy. This can achieve optimization of service quality and / or network quality with high security, high accuracy, high efficiency and a complete process.

[0131] like Figure 8 As shown, an embodiment of the present invention further provides an information quality optimization method based on federated learning, which is applied to a service provider, and the method includes:

[0132] Step 81, obtaining real-time service quality data sent by the terminal;

[0133] Step 82: When a service quality anomaly is detected based on the correspondence between the real-time service quality data and the configured services and the service level agreement (SLA) requirements, a root cause analysis request is sent to the wireless intelligent control platform; the root cause analysis request carries the service quality data and service configuration data at the time of the service quality anomaly.

[0134] Step 83: Receive an optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request. The optimization strategy is that the wireless intelligent control platform inputs the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result, and generates the root cause analysis result according to the root cause analysis result; wherein, the root cause analysis model is obtained through vertical federated modeling of federated learning.

[0135] In this embodiment, a service provider obtains real-time service quality data sent by a terminal. When the service provider detects a service quality anomaly based on the service quality data and the corresponding relationship between the configured service and the service level agreement (SLA) requirements, where the service quality anomaly refers to service quality that is worse than the SLA requirements, the service provider sends a root cause analysis request containing the service quality data at the time of the anomaly to the wireless intelligent control platform. The service provider then receives an optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request. The optimization strategy involves the wireless intelligent control platform inputting the service quality data and network quality data at the time of the anomaly into a pre-established root cause analysis model for root cause processing, thereby obtaining a root cause analysis result, and generating a root cause analysis result based on the root cause analysis result. The root cause analysis model is obtained through vertical federated modeling using federated learning. This can optimize service quality and / or network quality with high security, accuracy, efficiency, and a streamlined process.

[0136] Optionally, the root cause analysis model is obtained through longitudinal federated modeling of federated learning, including:

[0137] Obtaining sample data of service quality data when service quality is abnormal and sample data of network quality data on the terminal side;

[0138] Based on expert experience, sample data of the service quality data and sample data of the network quality data on the terminal side are analyzed to obtain a service quality anomaly label;

[0139] Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

[0140] Optionally, the SLA requirement includes at least one of the following:

[0141] Network SLA requirements;

[0142] Business SLA requirements: Different business types have different SLA requirements.

[0143] It should be noted that the method on the service provider side is a method corresponding to the above-mentioned method applied to the wireless intelligent control platform side. All implementation methods in the above-mentioned method embodiments applied to the wireless intelligent control platform side are applicable to the embodiments of the method on the service provider side and can achieve the same technical effect.

[0144] like Figure 9 As shown, the present invention also provides a quality optimization device 90 based on federated learning, which is applied to a wireless intelligent control platform. The device includes:

[0145] An acquisition module 91 is configured to acquire a root cause analysis request sent by a service provider, wherein the root cause analysis request carries service quality data and service configuration data when the service quality is abnormal;

[0146] Processing module 92 is used to input the network quality data and network configuration data on the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; based on the root cause analysis result, an optimization strategy is generated; and the service quality and / or network quality is optimized according to the optimization strategy.

[0147] Optionally, the root cause analysis model is obtained through longitudinal federated modeling of federated learning, including:

[0148] When the service provider monitors that the service quality is abnormal, receiving a network quality data collection request sent by the service provider;

[0149] Obtaining sample data of network quality data when service quality is abnormal according to the network quality data collection request;

[0150] Based on expert experience, sample data of the network quality data and sample data of the service provider's service quality data are analyzed to obtain a service quality anomaly label;

[0151] Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

[0152] Optionally, before obtaining the root cause analysis model through longitudinal federated modeling of federated learning, the following steps may be further included:

[0153] Negotiate service level agreement (SLA) guarantee requirements with the service provider;

[0154] Configure the correspondence between services and SLA requirements according to the SLA guarantee requirements;

[0155] The service provider monitors the service quality in real time to see if it is abnormal based on the correspondence between the service and the SLA requirements.

[0156] Optionally, based on expert experience, sample data of the network quality data and sample data of the service provider's service quality data are analyzed to obtain a service quality anomaly label, including:

[0157] Based on expert experience, sample data of the network quality data and sample data of the service quality data of the service provider are analyzed to obtain the cause of the abnormality, and the cause of the abnormality is marked to obtain a service quality abnormality label; the cause of the abnormality includes at least one of the following: abnormal service configuration, poor terminal channel environment, wireless interference, abnormal wireless access network RAN ​​configuration, wireless resource congestion, core network abnormality, and transmission network abnormality.

[0158] Optionally, generating an optimization strategy based on the root cause analysis results includes:

[0159] Generating a first optimization strategy for service quality based on the root cause analysis results of the service quality; and / or

[0160] Based on the root cause analysis results of the network quality, a second optimization strategy for the network quality is generated.

[0161] Optionally, optimizing service quality and / or network quality according to the optimization strategy includes:

[0162] sending the first optimization strategy to a service provider, so that the service provider optimizes the service according to the first optimization strategy; and / or

[0163] The second optimization strategy is sent to the network side, so that the network side optimizes the network according to the second optimization strategy.

[0164] It should be noted that the device on the wireless intelligent control platform side is a device corresponding to the method on the above-mentioned wireless intelligent control platform side. All implementation methods in the above-mentioned method embodiment on the wireless intelligent control platform side are applicable to the embodiments of the device on the wireless intelligent control platform side and can achieve the same technical effect.

[0165] The present invention also provides a quality optimization device based on federated learning, which is applied to a service provider, and the device includes:

[0166] An acquisition module is used to acquire real-time service quality data sent by the terminal;

[0167] a transceiver module configured to send a root cause analysis request to the wireless intelligent control platform when abnormal service quality is detected based on the correspondence between the real-time service quality data and the configured services and the service level agreement (SLA) requirements; the root cause analysis request carries the service quality data and service configuration data at the time of the abnormal service quality;

[0168] An optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request is received, wherein the optimization strategy is that the wireless intelligent control platform inputs the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result, and is generated based on the root cause analysis result; wherein the root cause analysis model is obtained through vertical federated modeling of federated learning.

[0169] Optionally, the root cause analysis model is obtained through longitudinal federated modeling of federated learning, including:

[0170] Obtaining sample data of service quality data when service quality is abnormal and sample data of network quality data on the terminal side;

[0171] Based on expert experience, sample data of the service quality data and sample data of the network quality data on the terminal side are analyzed to obtain a service quality anomaly label;

[0172] Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

[0173] Optionally, the SLA requirement includes at least one of the following:

[0174] Network SLA requirements;

[0175] Business SLA requirements: Different business types have different SLA requirements.

[0176] It should be noted that the device on the service provider side is a device corresponding to the method on the service provider side mentioned above. All implementation methods in the method embodiment on the service provider side mentioned above are applicable to the embodiment of the device on the service provider side and can achieve the same technical effect.

[0177] An embodiment of the present invention further provides a communication device comprising: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0178] An embodiment of the present invention further provides a computer-readable storage medium, including stored instructions, which, when executed on a computer, cause the computer to execute the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0179] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0180] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0181] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0182] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0183] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0184] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, ROM, RAM, a magnetic disk, or an optical disk.

[0185] In addition, it should be noted that, in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it will be understood that all or any steps or components of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in hardware, firmware, software or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0186] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general-purpose device. Therefore, the purpose of the present invention can also be achieved simply by providing a program product containing program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but do not necessarily need to be performed in chronological order. Certain steps can be performed in parallel or independently of each other.

[0187] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for optimizing information quality based on federated learning, characterized in that: Applied to a wireless intelligent control platform, the method includes: Obtaining a root cause analysis request sent by a service provider, the root cause analysis request carrying service quality data and service configuration data when the service quality is abnormal; Inputting network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; generating an optimization strategy based on the root cause analysis results; The service quality and / or network quality are optimized according to the optimization strategy.

2. The information quality optimization method based on federated learning according to claim 1, characterized in that: The root cause analysis model is obtained through vertical federated modeling of federated learning, including: When the service provider monitors that the service quality is abnormal, receiving a network quality data collection request sent by the service provider; Obtaining sample data of network quality data when service quality is abnormal according to the network quality data collection request; Based on expert experience, sample data of the network quality data and sample data of the service provider's service quality data are analyzed to obtain a service quality anomaly label; Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

3. The information quality optimization method based on federated learning according to claim 1 or 2, characterized in that: Before obtaining the root cause analysis model through longitudinal federated modeling of federated learning, the following steps are also included: Negotiate service level agreement (SLA) guarantee requirements with the service provider; Configure the correspondence between services and SLA requirements according to the SLA guarantee requirements; The service provider monitors the service quality in real time to see if it is abnormal based on the correspondence between the service and the SLA requirements.

4. The information quality optimization method based on federated learning according to claim 2, characterized in that: Based on expert experience, sample data of the network quality data and sample data of the service provider's service quality data are analyzed to obtain service quality anomaly labels, including: Based on expert experience, sample data of the network quality data and sample data of the service quality data of the service provider are analyzed to obtain the cause of the abnormality, and the cause of the abnormality is marked to obtain a service quality abnormality label; the cause of the abnormality includes at least one of the following: abnormal service configuration, poor terminal channel environment, wireless interference, abnormal wireless access network RAN ​​configuration, wireless resource congestion, core network abnormality, and transmission network abnormality.

5. The information quality optimization method based on federated learning according to claim 1, characterized in that: Based on the root cause analysis results, an optimization strategy is generated, including: Generating a first optimization strategy for service quality based on the root cause analysis results of the service quality; and / or Based on the root cause analysis results of the network quality, a second optimization strategy for the network quality is generated.

6. The information quality optimization method based on federated learning according to claim 5, characterized in that: Optimizing service quality and / or network quality according to the optimization strategy includes: sending the first optimization strategy to a service provider, so that the service provider optimizes the service according to the first optimization strategy; and / or The second optimization strategy is sent to the network side, so that the network side optimizes the network according to the second optimization strategy.

7. A method for optimizing information quality based on federated learning, characterized in that: Applied to a service provider, the method includes: Obtain real-time service quality data sent by the terminal; According to the correspondence between the real-time service quality data and the configured services and the service level agreement (SLA) requirements, when a service quality abnormality is detected, a root cause analysis request is sent to the wireless intelligent control platform; the root cause analysis request carries the service quality data and service configuration data at the time of the service quality abnormality; An optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request is received, wherein the optimization strategy is that the wireless intelligent control platform inputs the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result, and is generated based on the root cause analysis result; wherein the root cause analysis model is obtained through vertical federated modeling of federated learning.

8. The information quality optimization method based on federated learning according to claim 7, characterized in that: The root cause analysis model is obtained through vertical federated modeling of federated learning, including: Obtaining sample data of service quality data when service quality is abnormal and sample data of network quality data on the terminal side; Based on expert experience, sample data of the service quality data and sample data of the network quality data on the terminal side are analyzed to obtain a service quality anomaly label; Based on the service quality anomaly label, vertical federated modeling of federated learning is performed to obtain the root cause analysis model.

9. The method for optimizing information quality based on federated learning according to claim 7, characterized in that: SLA requirements include at least one of the following: Network SLA requirements; Business SLA requirements: Different business types have different SLA requirements.

10. An information quality optimization device based on federated learning, characterized in that: Applied to a wireless intelligent control platform, the device includes: An acquisition module is used to acquire a root cause analysis request sent by a service provider, wherein the root cause analysis request carries service quality data and service configuration data when the service quality is abnormal; A processing module is used to input the network quality data and network configuration data on the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result; the root cause analysis model is obtained through vertical federated modeling of federated learning; an optimization strategy is generated based on the root cause analysis result; and the service quality and / or network quality is optimized according to the optimization strategy.

11. An information quality optimization device based on federated learning, characterized in that: Applied to a service provider, the device includes: An acquisition module is used to acquire real-time service quality data sent by the terminal; a transceiver module configured to send a root cause analysis request to the wireless intelligent control platform when abnormal service quality is detected based on the correspondence between the real-time service quality data and the configured service and service level agreement (SLA) requirements; the root cause analysis request carries the service quality data and service configuration data at the time of abnormal service quality; and An optimization strategy fed back by the wireless intelligent control platform based on the root cause analysis request is received, wherein the optimization strategy is that the wireless intelligent control platform inputs the network quality data and network configuration data of the base station side when the service quality is abnormal, as well as the service quality data and service configuration data of the service provider, into a pre-established root cause analysis model for root cause processing to obtain a root cause analysis result, and is generated based on the root cause analysis result; wherein the root cause analysis model is obtained through vertical federated modeling of federated learning.

12. A communication device, characterized in that: include: A processor and a memory storing a computer program, wherein when the computer program is executed by the processor, the method according to any one of claims 1 to 6 or the method according to any one of claims 7 to 9 is performed.

13. A computer-readable storage medium, characterized in that The device stores instructions, which, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 6 or the method according to any one of claims 7 to 9.

Images