Distributed wireless network upgrade method, system, device and medium

Abstract

The present disclosure provides a distributed wireless network upgrade method, system, device and medium, which relates to the field of wireless communication technology. An aggregated base station entity sends a model input data request to multiple distributed base station entities; the aggregated base station entity receives the measurement data obtained by multiple distributed base station entities and terminal devices according to the model input data request as the model feature input; the model feature input is input into the AI / ML model for model reasoning, and the adjustment strategy and model performance feedback information of multiple distributed base station entities and terminal devices are output; the performance feedback data after the multiple distributed base station entities and terminal devices execute the adjustment strategy is received and sent to OAM; the AI / ML model is upgraded according to the performance feedback data and the model performance feedback information, and the upgraded AI / ML model is deployed on the aggregated base station entity and multiple distributed base station entities. Model sharing is achieved through the coordination of the distributed base station entities by the aggregated base station entity.

Description

Background Art

[0002] Wireless network intelligence has become a new trend in the development of the Internet industry. The combination of artificial intelligence and 5G technologies has brought tremendous energy to realize efficient, fast and stable interconnection and intercommunication of wireless networks. Operators can rely on their own advantages in scene applications, massive data and computing resources, and use artificial intelligence technology to improve network intelligence and service levels, and expand new scenarios for artificial intelligence applications. Therefore, network intelligence is one of the key research directions for operators in the future.

[0003] The 3rd Generation Partnership Project (3GPP) international standards organization has carried out artificial intelligence (AI) standardization work from different directions. The research topic "Study on enhancement for Data Collection for NR and EN-DC" has been launched, which mainly discusses the relevant principles, functional framework, use cases and solutions of AI-enabled intelligent radio access network (Radio Access Network, RAN) to achieve high-level application cases such as network energy saving, load balancing and mobility management.

[0004] In the related art, for the AI / machine learning (ML)-based network load balancing use case signaling design of non-separated architecture base stations, its model training and model reasoning functions are deployed in the 5G access network NG-RAN entity.

[0005] Currently, various standard organizations have not conducted research on the relevant architecture design or signaling process design for distributed wireless network learning.

[0006] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0007] The present disclosure provides a distributed wireless network upgrade method, system, device and medium, which at least to some extent overcome the problem of the lack of architecture or signaling process of existing distributed wireless network learning.

[0008] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0009] According to one aspect of the present disclosure, a distributed wireless network upgrade method is provided, which is applied to a distributed wireless network system, wherein the distributed wireless network system includes an aggregated base station entity, multiple distributed base station entities, a terminal device, and an operation maintenance management OAM, and the method includes:

[0010] The aggregated base station entity sends a model input data request to the multiple distributed base station entities, where the model input data request is used to perform model reasoning on the AI / ML model of the subscribed use case;

[0011] The aggregated base station entity receives measurement data obtained by the multiple distributed base station entities and terminal devices according to the model input data request as a model feature input quantity;

[0012] The aggregated base station entity inputs the model feature input into the AI / ML model for model inference, and outputs adjustment strategies and model performance feedback information for the multiple distributed base station entities and the terminal devices;

[0013] The aggregated base station entity receives performance feedback data after the multiple distributed base station entities and the terminal equipment execute the adjustment strategy, and sends the performance feedback data and model performance feedback information to the OAM;

[0014] The OAM upgrades the AI / ML model according to the performance feedback data and the model performance feedback information, and deploys the upgraded AI / ML model on the aggregated base station entity and the multiple distributed base station entities.

[0015] In one embodiment of the present disclosure, the AI / ML model is obtained by model training, testing and verification in the aggregated base station entity or the OAM.

[0016] In one embodiment of the present disclosure, before the aggregated base station entity sends a model input data request to the multiple distributed base station entities, the method further includes:

[0017] The aggregation base station entity sends an analysis subscription request message to the OAM to perform model training on the AI / ML model of the subscription use case according to the analysis subscription request message;

[0018] The OAM sends an analysis subscription response message to the aggregate base station entity to indicate the AI / ML model deployment and the model feature input quantity required for model training.

[0019] In one embodiment of the present disclosure, the analysis subscription request message includes at least one of a base station node ID, a subscription use case ID, an analysis report target, and a parameter type contained in the analysis report, wherein:

[0020] The subscription use case ID includes at least one of a network energy saving use case, a load balancing use case, a mobility management use case, and a slice management use case;

[0021] The analysis report target includes at least one of OAM, a specific aggregated base station entity, a specific group of aggregated base station entities, a specific terminal device, and a group of terminal devices;

[0022] The parameter types contained in the analysis report include at least one of the following:

[0023] Base station temperature, humidity, voltage, and current;

[0024] Base station physical radio resource utilization measurement, cell physical resource block PRB utilization;

[0025] Measurement of the average number of radio resource control (RRC) connections in a cell;

[0026] gNB-DU downlink packet loss rate measurement;

[0027] Downlink Packet Data Convergence Protocol PDCP SDU packet loss rate of gNB-CU-UP;

[0028] Delay measurement;

[0029] Minimization of drive test (MDT) data, wherein the MDT data includes at least one of a reference signal received power (RSRP), a reference signal received quality (RSRQ), a signal to interference plus noise ratio (SINR), and a terminal device location information related to UE measurement;

[0030] New radio NR cell ID;

[0031] Timestamp.

[0032] In one embodiment of the present disclosure, the model feature input includes at least one of the following:

[0033] New radio NR cell ID;

[0034] IP address information and port information used for input data interaction;

[0035] Base station physical radio resource utilization measurement, cell PRB utilization;

[0036] Measurement of the average number of RRC connections in a cell;

[0037] Downlink packet loss rate measurement;

[0038] MDT data;

[0039] Service experience statistics and service usage forecasts for each terminal device.

[0040] In one embodiment of the present disclosure, the aggregated base station entity receives the measurement data obtained by the multiple distributed base station entities and the terminal devices according to the model input data request as the model feature input quantity, including:

[0041] The distributed base station entity configures MDT measurement information for the terminal device according to the model input data request, wherein the measurement information includes at least one of the following: measurement trigger period, measurement area configuration information;

[0042] The terminal device reports the measurement data of the terminal device obtained according to the measurement information to the distributed base station entity.

[0043] In one embodiment of the present disclosure, the method further includes:

[0044] The distributed base station entity sends measurement data of the distributed base station entity and the terminal device to the aggregate base station entity.

[0045] In one embodiment of the present disclosure, the measurement data of the terminal device includes at least one of the following:

[0046] Terminal device ID;

[0047] Serving cell ID;

[0048] Terminal device location information;

[0049] Terminal equipment measures related RSRP, RSRQ, and SINR;

[0050] RSRP and RSRQ of the serving cell;

[0051] RSRP and RSRQ of neighboring cells.

[0052] In one embodiment of the present disclosure, the aggregated base station entity inputs the model feature input into the AI / ML model for model inference, and outputs adjustment strategies and model performance feedback information of the multiple distributed base station entities and the terminal devices, including:

[0053] The aggregation base station entity outputs an analysis report, wherein the analysis report includes at least one of the following:

[0054] Predicted wireless resource information;

[0055] Forecast range;

[0056] List of congested cells within the prediction range;

[0057] A list of target cells carrying handover traffic within the predicted range;

[0058] List of congestion relief cells within the forecast range;

[0059] Prediction confidence;

[0060] List of services;

[0061] Root causes of energy consumption in residential areas;

[0062] Current energy-saving cell status;

[0063] Predicting cell energy efficiency;

[0064] Recommended energy-saving decisions.

[0065] In one embodiment of the present disclosure, the method further includes:

[0066] The multiple distributed base station entities and the terminal devices execute corresponding adjustment strategies according to the analysis report, wherein the adjustment strategies include at least one of UE switching and cell execution of energy-saving strategies.

[0067] In one embodiment of the present disclosure, before the aggregated base station entity receives the performance feedback data after the multiple distributed base station entities and the terminal device execute the adjustment strategy, the method further includes:

[0068] The OAM sends feedback request information to the aggregation base station entity to indicate the required feedback data type;

[0069] The aggregated base station entity sends a feedback request to the multiple distributed base station entities according to the feedback request information;

[0070] The multiple distributed base station entities perform measurements according to the feedback request, obtain performance feedback data after executing the adjustment strategy, and send the performance feedback data to the aggregated base station entity.

[0071] In one embodiment of the present disclosure, the performance feedback data after executing the adjustment strategy includes at least one of the following:

[0072] Cell PRB utilization;

[0073] Measurement of the average number of RRC connections in a cell;

[0074] Cell packet loss rate measurement;

[0075] Average delay;

[0076] MDT data;

[0077] Service experience analysis for each terminal device.

[0078] In one embodiment of the present disclosure, the AI / ML model is obtained by model training, testing and verification in the aggregated base station entity or the OAM, including:

[0079] Acquire historical measurement data of the plurality of distributed base station entities and terminal devices as training samples;

[0080] Input the training samples into the AI / ML model to be trained for model reasoning, and output prediction adjustment strategies and historical model prediction performance parameters of the plurality of distributed base station entities and the terminal devices;

[0081] Receiving historical performance feedback data after the distributed base station entity and the terminal device execute the prediction adjustment strategy;

[0082] Determine whether the preset training stop condition is met;

[0083] When the preset training stop condition is not met, the model parameters of the AI / ML model are adjusted according to the historical model prediction performance parameters and the historical performance feedback data, and the new training samples are input into the adjusted AI / ML model until the preset training stop condition is met;

[0084] When the preset training stop conditions are met, the trained AI / ML model is obtained.

[0085] In one embodiment of the present disclosure, the upgraded AI / ML model is stored in the OAM.

[0086] In one embodiment of the present disclosure, the AI / ML model includes but is not limited to a network energy-saving model, a load balancing model, a mobility management model, and a slice management model.

[0087] According to another aspect of the present disclosure, a distributed wireless network system is also provided, including an aggregated base station entity, a plurality of distributed base station entities communicating with the aggregated base station entity, a terminal device, and an operation maintenance management OAM, wherein:

[0088] The aggregated base station entity is used to send a model input data request to the multiple distributed base station entities, the model input data request is used to perform model reasoning on the AI / ML model of the subscribed use case, and the AI / ML model is obtained by model training, testing and verification at the aggregated base station entity or OAM; receiving measurement data obtained by the multiple distributed base station entities and terminal devices according to the model input data request as model feature input; inputting the model feature input into the AI / ML model for model reasoning, and outputting adjustment strategies and model performance feedback information for the multiple distributed base station entities and the terminal devices; receiving performance feedback data after the multiple distributed base station entities and the terminal devices execute the adjustment strategy, and sending the performance feedback data and model performance feedback information to the OAM;

[0089] The multiple distributed base station entities and terminal devices are used to receive the model input data request sent by the aggregated base station entity, perform measurements according to the model input data request, obtain measurement data and send it to the aggregated base station entity; and execute the adjustment strategy obtained by the aggregated base station entity through model reasoning, and send performance feedback data after executing the adjustment strategy to the aggregated base station entity;

[0090] The OAM is used to upgrade the AI / ML model according to the performance feedback data and the model performance feedback information, and deploy the upgraded AI / ML model on the aggregated base station entity and the multiple distributed base station entities.

[0091] According to another aspect of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the above-mentioned distributed wireless network upgrade method by executing the executable instructions.

[0092] According to another aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the distributed wireless network upgrade method described above is implemented.

[0093] According to another method of the present disclosure, a computer program product is provided, including executable instructions, which are stored in a computer-readable storage medium. A processor of an electronic device reads the executable instructions from the computer-readable storage medium, and the processor executes the executable instructions, so that the electronic device executes the above-mentioned distributed wireless network upgrade method.

[0094] The embodiments of the present disclosure provide a distributed wireless network upgrade method, system, device and medium, which realize data sharing and model sharing among multiple network entities by coordinating distributed base station entities through aggregated base station entities, thereby improving the accuracy, robustness and confidence of the model and enhancing the performance of the intelligent network.

[0095] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0097] Figure 1 A distributed wireless network system architecture diagram is shown in an embodiment of the present disclosure;

[0098] Figure 2 A flow chart of a distributed wireless network upgrade method according to an embodiment of the present disclosure is shown;

[0099] Figure 3 A flow chart of another distributed wireless network upgrade method according to an embodiment of the present disclosure is shown;

[0100] Figure 4 A flowchart of AI / ML model training in an embodiment of the present disclosure is shown;

[0101] Figure 5 A signaling diagram showing a distributed wireless network upgrade method in an embodiment of the present disclosure;

[0102] Figure 6 A signaling diagram showing a distributed wireless network load balancing use case in an embodiment of the present disclosure is shown;

[0103] Figure 7 A schematic diagram of a distributed wireless network system disclosed in the present invention is shown;

[0104] Figure 8 A structural block diagram of an electronic device in an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0105] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the disclosure will be more comprehensive and complete and to fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0106] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0107] Figure 1 A schematic diagram showing an exemplary system architecture that can be applied to a distributed wireless network upgrade method or a distributed wireless network system according to an embodiment of the present disclosure is shown.

[0108] like Figure 1 As shown, the system architecture may include a terminal device 103, a radio access network (RAN) and a core network 701, wherein the radio access network RAN ​​includes an aggregate base station entity 101 and multiple distributed base station entities 102 communicating with the aggregate base station entity 101, and the terminal device 103 communicates with the distributed base station entity 102.

[0109] The wireless access network is used to provide a medium for a communication link between the terminal device 103 and the core network 701 .

[0110] Optionally, the wireless access network described above uses standard communication technologies and / or protocols. The wireless access network is typically the Internet, but may also be any network, including but not limited to a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a dedicated network or any combination of a virtual private network. In some embodiments, technologies and / or formats including Hyper Text Mark-up Language (HTML), Extensible Markup Language (XML), etc. are used to represent data exchanged over the network. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. may be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies may also be used to replace or supplement the above data communication technologies.

[0111] It can be understood that for a wireless communication system, it is a network that provides wireless communication functions. The communication system can adopt different communication technologies, such as Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency-division Multiple Access (OFDMA), Single Carrier FDMA (SC-FDMA), Carrier Sense Multiple Access with Collision Avoidance. According to the capacity, rate, delay and other factors of different networks, the network can be divided into 2G (generation) network, 3G network, 4G network or future evolution network, such as 5G network, 5G network is referred to as network or system. In the present disclosure, the wireless access network includes network equipment, which can be, for example, a wireless access network entity; the core network 701 includes core network equipment. Among them, the wireless access network entity can also be called a base station, such as Figure 1 The aggregated base station entity 101 and the multiple distributed base station entities 102. The wireless access network can provide network services for the terminal device 103 through network equipment. Different operators can provide different network services for the terminal device 103. It can also be understood that different operators correspond to different operator networks.

[0112] Furthermore, the wireless access network can communicate with the core network through the bearer network, which includes: Synchronous Digital Hierarchy (SDH), Packet Transport Network (PTN), and Optical Transport Network (OTN), etc.; among them, 2G services are mainly carried on SDH, and 3G services are mainly carried on PTN.

[0113] The network structures of SDH, PTN, and OTN all include three parts: the core layer, the aggregation ring, and the access ring. Specifically, the core layer of SDH is mainly carried by bare optical fiber, the aggregation ring of SDH is generally carried by optical cable, the long-distance section is carried by OTN, and the access ring of SDH is carried by optical cable; the core layer of PTN is carried by OTN or bare optical fiber, the aggregation ring of PTN is generally carried by OTN or optical cable, and the access ring of PTN is carried by optical cable; therefore, SDH, PTN, and OTN are relatively independent.

[0114] The terminal device 103 may be various electronic devices, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, wearable devices, augmented reality devices, virtual reality devices, etc.

[0115] Optionally, the client of the application installed in different terminal devices 103 is the same, or the client of the same type of application based on different operating systems. Based on the different terminal platforms, the specific form of the client of the application can also be different, for example, the application client can be a mobile client, a PC client, etc.

[0116] Those skilled in the art will know that Figure 1 The number of terminal devices 103, base station entities of the wireless access network, and core network devices is only illustrative, and any number of terminal devices, wireless access network base station entities, and core network devices may be provided according to actual needs. This disclosure embodiment does not limit this.

[0117] The solution provided in the embodiment of the present application involves technologies such as wireless communication, big data processing of artificial intelligence, and machine learning. It is a software program applied to a computer, which uses measurement data and performance data sent by distributed base station entities to build models and predict adjustment strategies for subscription use cases, and then compares the performance feedback data after the adjustment strategy with the model performance feedback information, upgrades the AI / ML model, and deploys the upgraded AI / ML model to the aggregated base station entity and multiple distributed base station entities, thereby realizing data sharing and model sharing among multiple network entities, improving the accuracy, robustness and confidence of the model, and improving the performance of the intelligent network, which is specifically described by the following embodiments:

[0118] First, an embodiment of the present disclosure provides a distributed wireless network upgrade method, which can be executed by any system with computing and processing capabilities.

[0119] Figure 2 A flow chart of a distributed wireless network upgrade method according to an embodiment of the present disclosure is shown as follows: Figure 2As shown, the distributed wireless network upgrade method provided in the embodiment of the present disclosure is applied to a distributed wireless network system, which includes an aggregated base station entity, multiple distributed base station entities communicating with the aggregated base station entity, a terminal device, and an operation maintenance management OAM, including the following steps:

[0120] S202. The aggregate base station entity sends a model input data request to multiple distributed base station entities, where the model input data request is used to perform model reasoning on the AI / ML model of the subscribed use case.

[0121] The distributed wireless network system in this embodiment includes an aggregated base station entity and multiple distributed base station entities. Model reasoning is performed on the aggregated base station entity side. Before receiving test data sent by multiple distributed base station entities for model reasoning, the pre-trained, verified and tested AI / ML model is deployed on the aggregated base station entity and multiple distributed base station entities, and a model input data request required for model reasoning is obtained. The aggregated base station entity sends a request carrying the model input data to the multiple distributed base station entities to instruct the multiple distributed base station entities and terminal devices to collect relevant test data for model reasoning analysis.

[0122] Each subscription use case corresponds to a pre-trained, verified, and tested AI / ML model. Usually, each subscription use case corresponds to a subscription use case ID, that is, the subscription use case ID corresponds to the AI / ML model one-to-one.

[0123] It should be noted that AI / ML models include but are not limited to network energy-saving models, load balancing models, mobility management models, slice management models, etc., and this application does not make specific limitations.

[0124] Among them, the AI / ML model is obtained by model training, testing and verification in the aggregated base station entity or OAM. It should be noted that the execution subject of the AI / ML model training can be determined according to the actual situation. In this disclosure, the execution subject of the AI / ML model training is set in the aggregated base station entity or OAM to effectively improve the system efficiency.

[0125] S204. The aggregated base station entity receives measurement data obtained by multiple distributed base station entities and terminal devices according to model input data requests as model feature input quantities.

[0126] In this embodiment, the measurement data sent by the distributed base station entity to the aggregated base station entity includes the measurement data obtained by the distributed base station entity according to the model input data request and the measurement data obtained by the terminal device according to the MDT measurement information configured by the distributed base station entity. The measurement data is uniformly sent by the distributed base station entity to the aggregated base station entity, and the measurement data sent by each distributed base station entity is a set of data in the model feature input quantity.

[0127] The model feature input includes at least one of the following:

[0128] New radio NR cell ID;

[0129] IP address information and port information used for input data interaction;

[0130] Base station physical radio resource utilization measurement, cell PRB utilization;

[0131] Measurement of the average number of RRC connections in a cell;

[0132] Downlink packet loss rate measurement;

[0133] MDT data, where the MDT data includes UE location information (longitude information, latitude information, altitude information), RSRP and RSRQ of the serving cell and neighboring cells, and RSRP, RSRQ and SINR tested by the UE;

[0134] Service experience statistics and service usage forecasts for each terminal device.

[0135] Before the terminal device measures the data, the distributed base station entity performs Minimization Drive Test (MDT) measurement configuration on the terminal device in the local cell. The MDT measurement configuration information includes but is not limited to the measurement trigger period, measurement area configuration information, etc. The terminal device uses the Radio Resource Management (RRM) mechanism to report the measurement data to the distributed base station entity.

[0136] It should be noted that the measurement data of the terminal device includes but is not limited to the terminal device ID, service cell ID, UE location information (longitude information, latitude information, altitude information), UE measurement-related reference signal receiving power (RSRP), LTE reference signal receiving quality (RSRQ) and signal to interference plus noise ratio (SINR), RSRP and RSRQ of the service cell, and RSRP and RSRQ of the neighboring cell, etc.

[0137] The measurement data of the distributed base station entity includes but is not limited to the New Radio (NR) cell ID, IP address information and port information used for input data interaction, base station physical radio resource utilization measurement, cell physical resource block (PRB) utilization, cell average radio resource control (RRC) connection number measurement, cell downlink packet loss rate measurement, MDT data, each UE (source / target cell) service experience statistics and service usage prediction, etc.

[0138] The terminal device sends the measurement data to the distributed base station entity. The distributed base station entity sends the measurement data of the terminal device and the measurement data of the distributed base station entity to the aggregate base station entity. The aggregate base station entity performs data preprocessing on the received measurement data. The data preprocessing methods include filling missing values, data type conversion, formatting, normalization and equalization, etc., to construct model feature input.

[0139] The aggregated base station entity is responsible for data collection, model training, reasoning and intelligent prediction, and outputs analysis results for network, network management and application execution policy decisions; each distributed base station node can perform joint modeling and iterative optimization while maintaining independence. Efficient modeling between distributed base station nodes avoids problems such as unstable model performance, waste of network resources and poor user experience caused by the scarcity of training samples and insufficient computing power of individual network infrastructure, which is conducive to promoting the realization of AI endogenous to the network.

[0140] S206. The aggregated base station entity inputs the model feature input into the AI / ML model for model reasoning, and outputs adjustment strategies and model performance feedback information for multiple distributed base station entities and terminal devices.

[0141] In this embodiment, the model feature input is input into a pre-trained, verified and tested AI / ML model, which performs model inference and outputs an analysis report, wherein the analysis report includes but is not limited to predicted wireless resource information, prediction range (prediction of a time window in advance), a list of congested cells within the prediction range, a list of target cells carrying switching traffic within the prediction range, a list of cells with congested environments within the prediction range, prediction confidence, a service list (a list of supported service classes or corresponding slice IDs in a list of possible target cells), root causes of cell energy consumption, current energy-saving cell status, predicted cell energy efficiency, recommended energy-saving decisions, etc.

[0142] It should be noted that multiple distributed base station entities and terminal devices perform model reasoning and output analysis reports based on the AI / ML model, so that the distributed base station entities and terminal devices execute corresponding adjustment strategies to realize the subscription use case.

[0143] The adjustment strategies executed by the distributed base station entities and terminal devices include but are not limited to UE switching and cell energy-saving decisions, such as carrier shutdown or deep sleep.

[0144] S208. The aggregated base station entity receives performance feedback data after the multiple distributed base station entities and terminal devices execute the adjustment strategy, and sends the performance feedback data and model performance feedback information to OAM.

[0145] The performance feedback data after the multiple distributed base station entities in this embodiment execute the adjustment strategy include but are not limited to the cell PRB utilization, the cell average RRC connection number measurement, the cell packet loss rate measurement, the average delay, the MDT data, the service experience analysis of each UE (source / target cell), etc., wherein the MDT data includes the UE location information (longitude information, latitude information, altitude information), the RSRP and RSPQ of the serving cell and the neighboring cell, and the RSRP, RSRQ and SINR measured by the UE.

[0146] After the aggregated base station entity receives the performance feedback data sent by the distributed base station entity, the aggregated base station entity sends the performance feedback data of the distributed base station entity and the model performance feedback information obtained by model reasoning to OAM, so that OAM can judge the performance indicators such as model accuracy, robustness, and confidence, and confirm whether the AI / ML model needs to be upgraded.

[0147] S210. OAM upgrades the AI / ML model according to the performance feedback data and the model performance feedback information, and deploys the upgraded AI / ML model in the aggregated base station entity and multiple distributed base station entities.

[0148] In this embodiment, if the performance feedback data is inferior to the model performance feedback information, OAM optimizes the model parameters of the AI / ML model according to the performance feedback data and upgrades the AI / ML model; if the performance feedback data is better than the model performance feedback information, there is no need to upgrade the AI / ML model.

[0149] OAM optimizes and upgrades the model based on model performance feedback information and performance feedback data. OAM deploys the optimized and upgraded AI / ML model to aggregated base station entities and distributed base station entities, thereby achieving collaboration between base stations for model training and model reasoning.

[0150] The distributed wireless network upgrade method provided by the embodiment of the present disclosure performs comprehensive data collection, model training, model reasoning, and model upgrade processes on distributed base station entities in different areas by aggregating base station entities, thereby realizing common modeling among distributed base stations and continuous iterative optimization of algorithms, realizing data sharing and model sharing among multiple network entities, improving the accuracy, robustness, and confidence of the model, and improving the performance of intelligent networks. Model sharing is performed for distributed base station entities with the participation of multiple parties, realizing lightweight model training and model reasoning for distributed base station entities, and enhancing the performance of the entire network by improving the intelligent performance of the local entity network.

[0151] Figure 3 FIG. 2 shows another flow chart of a distributed wireless network upgrade method according to an embodiment of the present disclosure. Figure 3 As shown, in this embodiment, before the aggregate base station entity sends a model input data request to the multiple distributed base station entities, the method further includes:

[0152] S302, the aggregation base station entity sends an analysis subscription request message to the OAM to perform model training on the AI / ML model of the subscription use case according to the analysis subscription request message;

[0153] S304. OAM sends an analysis subscription response message to the aggregated base station entity to indicate the AI / ML model deployment and the model feature input quantity required for model training.

[0154] In this embodiment, the AI / ML model is stored in OAM, and the aggregated base station entity sends an analysis subscription request message to OAM. The analysis subscription request message carries a subscription use case type that represents the subscription use case. The subscription use case type is used to distinguish different subscription use cases. Each subscription use case type corresponds to an AI / ML model. OAM searches for the AI / ML model corresponding to the subscription use case type. OAM deploys the pre-trained AI / ML model to the aggregated base station entity and multiple distributed base station entities, thereby preparing for model reasoning and model training.

[0155] It should be noted that the analysis subscription request message includes but is not limited to the distributed base station entity ID, subscription use case type, analysis report target, and parameter types contained in the analysis report. Specifically, AI / ML models include network energy saving model, load balancing model, mobility management model, slice management model, etc.; analysis report targets include OAM, specific NG-RANnode, a specific group of NG-RAN nodes, specific UE, a group of UEs, etc.; the parameter types contained in the analysis report vary according to different AI / ML models, including base station temperature, humidity, voltage, current, base station physical resource utilization measurement, cell PRB utilization, cell average RRC connection number measurement, gN-DU downlink packet loss rate measurement, gNB-CU-UP downlink PDCP SDU packet loss rate, delay measurement, MDT data, NR cell ID, and timestamp (timestamp for each measurement report), etc. Among them, MDT data includes RSRP, RSRQ, SINR, and UE location information related to UE measurement.

[0156] OAM deploys the AI / ML model corresponding to the subscription use case type in the analysis subscription request message to the aggregated base station entity and multiple distributed base station entities. At the same time, OAM sends an analysis subscription response message to the aggregated base station entity. The analysis subscription response message is used for model deployment and relevant instructions on feature input required for model training.

[0157] As an optional embodiment of the present disclosure, the aggregated base station entity periodically sends an analysis subscription request message to the OAM. That is, when the periodic time is up, the aggregated base station entity sends an analysis subscription request message to the OAM, such as performing model reasoning every 1 hour, 4 hours, 12 hours, etc. The period is determined according to actual needs and is not specifically limited in this application.

[0158] In addition, the following method can also be used to determine the timing of the aggregated base station entity sending an analysis subscription request message to OAM: the aggregated base station entity monitors the performance feedback data sent by multiple distributed base station entities in real time, and when the performance feedback data is less than a preset performance threshold, sends an analysis subscription request message to OAM. It should be noted that the preset performance threshold can be determined according to actual needs, and this application does not make a specific limitation. Generally, the higher the preset performance threshold, the more frequent the model update.

[0159] In one embodiment of the present disclosure, step S204, the aggregated base station entity receives measurement data obtained by multiple distributed base station entities and terminal devices according to the model input data request measurement as the model feature input quantity, including:

[0160] The distributed base station entity configures MDT measurement information for the terminal device according to the model input data request, wherein the measurement information includes at least one of the following: measurement trigger period, measurement area configuration information;

[0161] The terminal device reports the measurement data of the terminal device obtained according to the measurement information to the distributed base station entity.

[0162] Specifically, the aggregated base station entity sends model input data requests to the multiple distributed base station entities respectively to instruct the multiple distributed base station entities and terminal devices to perform measurements to obtain measurement data for model reasoning.

[0163] It should be noted that the measurement data includes test data of multiple distributed base station entities forcibly tested according to model input data and measurement data of terminal equipment, which has been pointed out in the above embodiments and will not be repeated here.

[0164] In this embodiment, the method further includes:

[0165] The distributed base station entity sends measurement data of the distributed base station entity and the terminal device to the aggregated base station entity.

[0166] In one embodiment, before the aggregate base station entity receives the performance feedback data after the multiple distributed base station entities and the terminal devices execute the adjustment strategy, the method further includes:

[0167] The OAM sends a feedback request message to the aggregation base station entity to indicate the type of feedback data required;

[0168] The aggregation base station entity sends a feedback request to multiple distributed base station entities according to the feedback request information;

[0169] The multiple distributed base station entities perform measurements according to the feedback request, obtain performance feedback data after executing the adjustment strategy, and send the performance feedback data to the aggregated base station entity.

[0170] It should be noted that after multiple distributed base station entities and terminal devices execute the adjustment strategy, the performance of the aggregated base station entity, distributed base station entities and service cells tends to be stable. For example, after multiple distributed base station entities execute the adjustment strategy for a preset period of time, the OAM sends a feedback request message to the aggregated base station entity. The aggregated base station entity sends a feedback request to the multiple distributed base station entities based on the feedback request information to instruct the multiple distributed base station entities and terminal devices to measure the performance feedback data after executing the adjustment strategy. It will not be repeated here.

[0171] In this embodiment, the AI / ML model is obtained by model training, verification and testing in the aggregated base station entity or OAM.

[0172] When the AI / ML model is called for the first time, offline model training is required. When it is called again, online model training can be performed to determine the applicable model for the subscription use case or to optimize and upgrade the relevant model. It should be noted that offline model training uses historical data as the training data set. The historical data includes test data of one or more distributed base station entities, measurement data reported by the terminal device to the corresponding distributed base station entity, adjustment strategy of the AI / ML model after model inference, and performance feedback data after the distributed base station entity executes the adjustment strategy.

[0173] Figure 4 A flowchart of AI / ML model training in an embodiment of the present disclosure is shown.

[0174] In this embodiment, if Figure 4 As shown, the AI / ML model in step S202 is obtained by model training, testing and verification in the aggregated base station entity or OAM, including:

[0175] S402, obtaining historical measurement data of multiple distributed base station entities and terminal devices as training samples;

[0176] S404: Input the training samples into the AI / ML model to be trained for model reasoning, and output prediction adjustment strategies and historical model prediction performance parameters of multiple distributed base station entities;

[0177] S406, receiving historical performance feedback data after the distributed base station entity and the terminal device execute the prediction adjustment strategy;

[0178] S408, determining whether a preset training stop condition is met;

[0179] S410. When the preset training stop condition is not met, adjust the model parameters of the AI / ML model according to the historical model prediction performance parameters and the historical performance feedback data, and input the new training sample into the adjusted AI / ML model until the preset training stop condition is met;

[0180] S412: When the preset training stop condition is met, a trained AI / ML model is obtained.

[0181] When training the AI / ML model, historical measurement data is used as training samples. The training samples are divided into training sets and test sets according to a certain ratio. The training set is used to train and verify the AI / ML model, so that when the AI / ML model is used to predict the data in the test set, the obtained adjustment strategy is more in line with the actual situation, thereby improving the accuracy of the AI / ML model.

[0182] When training the model, historical measurement data in the training set is selected as the model feature input, and the prediction adjustment strategy and historical model prediction performance parameters are used as output, so that the historical performance feedback data obtained after the distributed base station entity executes the prediction adjustment strategy is closer to the historical model prediction performance parameters.

[0183] Input the training set into the AI / ML model to be trained, calculate the F1score, precision, accuracy, and confusion matrix of the AI / ML model to be trained to measure the quality of the AI / ML model to be trained.

[0184] The training stop conditions are set as precision greater than the preset precision, accuracy greater than the preset accuracy, and F1score greater than the preset F1 score. The model parameters of the AI / ML model are determined to obtain the trained AI / ML model. Then, the test set is input into the trained AI / ML model to verify the model.

[0185] In order to verify the accuracy of the trained AI / ML model, when training the AI / ML model, 70% of the data in the training set is randomly selected as training data, and the remaining 30% is used as test data. The model is trained using the training data. After the AI / ML model is trained, the remaining 30% of the test data is input into the above model for verification. If the verification result passes, the trained AI / ML model is obtained. If the verification result fails, the model parameters of the AI / ML model are adjusted until the training stop conditions are met to obtain the trained AI / ML model.

[0186] It should be noted that after the AI / ML model is trained, it is packaged and generated into an application for calling. The upgraded AI / ML model is saved in OAM so that after the aggregated base station entity sends an analysis subscription request message, OAM deploys the corresponding AI / ML model to the aggregated base station entity and the distributed base station entity according to the analysis subscription request message, and then performs model inference.

[0187] The distributed wireless network upgrade method provided by the embodiment of the present disclosure realizes data sharing and model sharing among multiple network entities by coordinating the distributed base station entities through the aggregation of base station entities, thereby improving the accuracy, robustness and confidence of the model, and enhancing the performance of the intelligent network. It solves the problems of scarce training samples, inaccurate training models, and difficulties in realizing network intelligent use cases caused by limited data storage and insufficient computing resources for some base station entities. It performs model sharing for distributed base station entities with the participation of multiple parties, realizes lightweight model training and model reasoning for distributed base station entities, enhances the performance of the entire network by improving the intelligent performance of the local entity network, and realizes distributed wireless network machine learning.

[0188] The following is combined with Figure 5A distributed wireless network upgrade method is described.

[0189] The evolution of wireless network independent AI architecture to distributed AI architecture is an inevitable trend. Based on the differences in data volume, computing power and algorithms of wireless network devices, this disclosure uses the aggregated base station entity with more data and computing resources as the main entity to perform wide-area collection and processing of measurement data of multiple distributed base station entities, and then optimizes and deploys distributed machine learning algorithm models, enabling the implementation of network distributed AI algorithm-related use cases. Figure 5 As shown, the specific process is as follows:

[0190] S501, the aggregation base station entity sends an analysis subscription request message to the OAM, notifying the OAM to perform model training on the AI / ML model of the relevant subscription use case planned to be implemented;

[0191] S502. OAM sends an analysis subscription response message to the aggregation base station entity for deployment of the AI / ML model and related instructions for feature input required for model training and model reasoning;

[0192] S503a. OAM performs AI / ML model training to obtain a trained AI / ML model.

[0193] S503b, OAM configures the trained AI / ML model to the aggregated base station entity;

[0194] S504a-S504c, the aggregation base station entity sends model input data requests to the distributed base station entities respectively, instructing each distributed base station and terminal device to measure relevant data for model reasoning analysis;

[0195] Among them, the distributed base station entities include multiple, such as Figure 5 The distributed base station entities 1, 2, ..., N in the aggregation base station entity send a model input data request to the distributed base station entities 1, 2, ..., N;

[0196] S505. Multiple distributed base station entities perform MDT measurement configuration on terminal devices in local cells to obtain measurement information, where the measurement information includes but is not limited to measurement trigger period, measurement area configuration information, etc.;

[0197] S506: The terminal device reports an MDT measurement report to the corresponding distributed base station entity.

[0198] The measurement report includes measurement data obtained according to the measurement information, and the MDT measurement report is reported using the RRM mechanism;

[0199] S507a-S507c, the distributed base station entities 1, 2, ..., N send their respective measurement data and the measurement data of the terminal device to the aggregated base station entity as model feature input;

[0200] S508. The aggregated base station entity performs model reasoning of the AI / ML model and outputs an analysis report, wherein the analysis report includes adjustment strategies to be executed by the distributed base station entity and the terminal device;

[0201] S509, multiple distributed base station entities and terminal devices execute the above adjustment strategy;

[0202] S510. OAM sends a feedback request message to the aggregation base station entity;

[0203] S511a~S511c, the aggregated base station entity sends feedback requests to the distributed base station entities 1, 2, ..., N respectively, to instruct the multiple distributed base station entities to test and send performance feedback data required for the feedback model upgrade according to the feedback request message;

[0204] S512a-S512c, distributed base station entities 1, 2, ..., N send their respective measured performance feedback data to the aggregated base station entity;

[0205] S513, the aggregated base station entity sends the received performance feedback data of the distributed base station entities 1, 2, ..., N, and the model performance feedback information obtained by model reasoning to the OAM;

[0206] S514. OAM optimizes and upgrades model parameters according to the above performance feedback data and model performance feedback information;

[0207] S515 and OAM deploy the upgraded AI / ML model in aggregated base station entities and multiple distributed base station entities to achieve collaboration between base stations for model training and model reasoning.

[0208] Figure 6 The signaling diagram of the distributed wireless network load balancing use case in the embodiment of the present disclosure is shown. Figure 6 As shown, the specific steps for the load balancing use case are as follows:

[0209] S601, the aggregation base station entity sends an analysis subscription request message to the OAM, notifying the OAM to perform load balancing model training for the load balancing use case planned to be implemented;

[0210] S602, OAM sends an analysis subscription response message to the aggregation base station entity for the deployment of the load balancing model and related instructions on the feature input required for model training and model reasoning;

[0211] S603a, OAM performs load balancing model training to obtain a pre-trained load balancing model;

[0212] S603b, OAM configures the pre-trained load balancing model to the aggregation base station entity;

[0213] S604a-S604c, the aggregated base station entity sends input data requests to the distributed base station entities 1, ..., N respectively, instructing each distributed base station and terminal device to measure relevant data for model reasoning analysis;

[0214] S605. Multiple distributed base station entities perform MDT measurement configuration on terminal devices in local cells to obtain measurement information, where the measurement information includes but is not limited to measurement trigger period, measurement area configuration information, etc.;

[0215] S606: The terminal device reports an MDT measurement report to the corresponding distributed base station entity. The MDT measurement report includes measurement data of the terminal device. The MDT measurement report is reported using an RRM mechanism.

[0216] S607a-S607c, the distributed base station entities 1, ..., N send their own measurement data and the measurement data of the terminal device to the aggregated base station entity as model feature input;

[0217] S608, the aggregated base station entity performs load balancing model reasoning and outputs an analysis report, wherein the analysis report includes a load balancing adjustment strategy to be executed by the distributed base station entity and the terminal device;

[0218] S609, multiple distributed base station entities and terminal devices execute the above load balancing adjustment strategy;

[0219] S610. OAM sends a feedback request message to the aggregation base station entity;

[0220] S611a-S611c, the aggregated base station entity sends feedback requests to the distributed base station entities 1, ..., N respectively, to instruct the multiple distributed base station entities to test and send performance feedback data required for feedback model upgrade;

[0221] S612a-S612c, distributed base station entities 1, ..., N send their respective measured performance feedback data to the aggregated base station entity;

[0222] S613, the aggregated base station entity sends the received performance feedback data of the distributed base station entities 1, 2, ..., N, and the model performance feedback information obtained by model reasoning to the OAM;

[0223] S614. OAM optimizes and upgrades model parameters according to the above performance feedback data and model performance feedback information;

[0224] S615, OAM deploys the upgraded load balancing model in the aggregated base station entity and the distributed base station entity to achieve collaboration between base stations for model training and model reasoning;

[0225] S616, the aggregation base station entity runs the upgraded load balancing model;

[0226] S616a-S616b, distributed base station entities 1, ..., N run the upgraded load balancing model;

[0227] Repeat the above steps to achieve load balancing in the distributed wireless network system.

[0228] Based on the same inventive concept, the embodiment of the present disclosure also provides a distributed wireless network system, as described in the following embodiments. Since the principle of solving the problem in the system embodiment is similar to that in the above method embodiment, the implementation of the system embodiment can refer to the implementation of the above method embodiment, and the repeated parts will not be repeated.

[0229] Figure 7 A schematic diagram of a distributed wireless network system in an embodiment of the present disclosure is shown. Figure 7 As shown, the distributed wireless network system includes an aggregated base station entity 101, multiple distributed base station entities 102, a terminal device 103 and an OAM 701, wherein:

[0230] The aggregated base station entity 101 is used to send a model input data request to multiple distributed base station entities 102, where the model input data request is used to perform model reasoning on the AI / ML model of the subscribed use case, and the AI / ML model is obtained by model training, testing and verification in the aggregated base station entity 101 or OAM 701; receive measurement data obtained by measuring multiple distributed base station entities 102 and terminal devices 103 according to the model input data request as model feature input; input the model feature input into the AI / ML model for model reasoning, and output adjustment strategies and model performance feedback information of multiple distributed base station entities 102 and terminal devices 103; receive performance feedback data after multiple distributed base station entities 102 and terminal devices 103 execute the adjustment strategy, and send the performance feedback data and model performance feedback information to OAM 701;

[0231] Multiple distributed base station entities 102 and terminal devices 103 are used to receive the model input data request sent by the aggregate base station entity 101, perform measurements according to the model input data request, obtain measurement data and send it to the aggregate base station entity 101; and execute the adjustment strategy obtained by the aggregate base station entity 101 through model reasoning, and send performance feedback data after executing the adjustment strategy to the aggregate base station entity 101;

[0232] OAM 701 is used to upgrade the AI / ML model according to the performance feedback data and the model performance feedback information, and deploy the upgraded AI / ML model on the aggregated base station entity 101 and multiple distributed base station entities 102.

[0233] In one embodiment of the present disclosure, the aggregate base station entity 101 is configured to send an analysis subscription request message to the OAM 701 before the aggregate base station entity 101 sends a model input data request to the plurality of distributed base station entities 102, so as to perform model training on the AI / ML model of the subscription use case according to the analysis subscription request message;

[0234] OAM 701 is used to send an analysis subscription response message to the aggregation base station entity 101 to indicate the model feature input quantity required for AI / ML model deployment and model training.

[0235] In one embodiment of the present disclosure, the analysis subscription request message includes at least one of a base station node ID, a subscription use case ID, an analysis report target, and a parameter type contained in the analysis report, wherein:

[0236] The subscription use case ID includes at least one of a network energy saving use case, a load balancing use case, a mobility management use case, and a slice management use case;

[0237] The analysis report target includes at least one of OAM, a specific aggregated base station entity, a specific group of aggregated base station entities, a specific terminal device, and a group of terminal devices;

[0238] The parameter types contained in the analysis report include at least one of the following:

[0239] Base station temperature, humidity, voltage, and current;

[0240] Base station physical radio resource utilization measurement, cell physical resource block PRB utilization;

[0241] Measurement of the average number of radio resource control (RRC) connections in a cell;

[0242] gNB-DU downlink packet loss rate measurement;

[0243] Downlink Packet Data Convergence Protocol PDCP SDU packet loss rate of gNB-CU-UP;

[0244] Delay measurement;

[0245] Minimize drive test (MDT) data, wherein the MDT data includes at least one of the reference signal received power (RSRP), the reference signal received quality (RSRQ), the signal to interference plus noise ratio (SINR), and the terminal device location information related to UE measurement;

[0246] New radio NR cell ID;

[0247] Timestamp.

[0248] It should be noted that the model feature input includes at least one of the following:

[0249] New radio NR cell ID;

[0250] IP address information and port information used for input data interaction;

[0251] Base station physical radio resource utilization measurement, cell PRB utilization;

[0252] Measurement of the average number of RRC connections in a cell;

[0253] Downlink packet loss rate measurement;

[0254] MDT data;

[0255] Service experience statistics and service usage forecasts for each terminal device.

[0256] In one embodiment of the present disclosure, the distributed base station entity 102 is configured to configure MDT measurement information for the terminal device 103 according to the model input data request, wherein the measurement information includes at least one of the following: measurement trigger period, measurement area configuration information;

[0257] The terminal device 103 is used to report the measurement data of the terminal device 103 obtained according to the measurement information to the distributed base station entity 102.

[0258] In one embodiment of the present disclosure, the distributed base station entity 102 is used to send measurement data of the distributed base station entity 102 and the terminal device 103 to the aggregated base station entity 101.

[0259] It should be noted that the measurement data of the terminal device 103 includes at least one of the following:

[0260] Terminal device ID;

[0261] Serving cell ID;

[0262] Terminal device location information;

[0263] Terminal equipment measures related RSRP, RSRQ, and SINR;

[0264] RSRP and RSRQ of the serving cell;

[0265] RSRP and RSRQ of neighboring cells.

[0266] In one embodiment of the present disclosure, the aggregation base station entity 101 is configured to output an analysis report, wherein the analysis report includes at least one of the following:

[0267] Predicted wireless resource information;

[0268] Forecast range;

[0269] List of congested cells within the prediction range;

[0270] A list of target cells carrying handover traffic within the predicted range;

[0271] List of congestion relief cells within the forecast range;

[0272] Prediction confidence;

[0273] List of services;

[0274] Root causes of energy consumption in residential areas;

[0275] Current energy-saving cell status;

[0276] Predicting cell energy efficiency;

[0277] Recommended energy-saving decisions.

[0278] In one embodiment of the present disclosure, a plurality of distributed base station entities 102 and terminal devices 103 are used to execute corresponding adjustment strategies according to the analysis report, wherein the adjustment strategies include at least one of UE switching and cell execution of energy-saving strategies.

[0279] In one embodiment of the present disclosure, the OAM 701 is configured to send feedback request information to the aggregated base station entity 101 to indicate the type of feedback data required before the aggregated base station entity 101 receives performance feedback data after the multiple distributed base station entities 102 and the terminal device 103 execute the adjustment strategy;

[0280] The aggregate base station entity 101 is configured to send a feedback request to multiple distributed base station entities 102 according to the feedback request information;

[0281] The multiple distributed base station entities 102 are used to perform measurements according to the feedback request, obtain performance feedback data after executing the adjustment strategy, and send the performance feedback data to the aggregated base station entity 101.

[0282] It should be noted that the performance feedback data after executing the adjustment strategy includes at least one of the following:

[0283] Cell PRB utilization;

[0284] Measurement of the average number of RRC connections in a cell;

[0285] Cell packet loss rate measurement;

[0286] Average delay;

[0287] MDT data;

[0288] Service experience analysis for each terminal device.

[0289] In one embodiment of the present disclosure, the aggregate base station entity 101 or OAM 701 is used to train an AI / ML model. Specifically, the aggregate base station entity 101 or OAM 701 is used to obtain historical measurement data of multiple distributed base station entities 102 and terminal devices 103 as training samples;

[0290] Input the training samples into the AI / ML model to be trained for model inference, and output the prediction adjustment strategies and historical model prediction performance parameters of multiple distributed base station entities 102 and terminal devices 103;

[0291] Receiving historical performance feedback data after the distributed base station entity 102 and the terminal device 103 execute the prediction adjustment strategy;

[0292] Determine whether the preset training stop condition is met;

[0293] When the preset training stop condition is not met, the model parameters of the AI / ML model are adjusted according to the historical model prediction performance parameters and the historical performance feedback data, and the new training samples are input into the adjusted AI / ML model until the preset training stop condition is met;

[0294] When the preset training stop conditions are met, the trained AI / ML model is obtained.

[0295] In one embodiment of the present disclosure, OAM 701 is used to store the upgraded AI / ML model.

[0296] It should be noted that AI / ML models include but are not limited to network energy-saving models, load balancing models, mobility management models, and slice management models.

[0297] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as a system, method or program product. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to herein as a "circuit", "module" or "system".

[0298] The distributed wireless network upgrade method and system provided in this embodiment realizes data sharing and model sharing among multiple network entities by coordinating distributed base station entities through aggregation base station entities, thereby improving the accuracy, robustness and confidence of the model, and enhancing the performance of the intelligent network. Model sharing is performed for distributed base station entities with the participation of multiple parties, thereby realizing lightweight model training and model reasoning of distributed base station entities, and enhancing the performance of the entire network by improving the intelligent performance of the local entity network.

[0299] Refer to the following Figure 8 An electronic device 800 according to this embodiment of the present invention will be described. Figure 8 The electronic device 800 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0300] like Figure 8 As shown, the electronic device 800 is in the form of a general computing device. The components of the electronic device 800 may include but are not limited to: at least one processing unit 810, at least one storage unit 820, and a bus 830 connecting different system components (including the storage unit 820 and the processing unit 810).

[0301] The storage unit stores program codes, which can be executed by the processing unit 810, so that the processing unit 810 performs the steps according to various exemplary embodiments of the present invention described in the above “Exemplary Method” section of this specification. For example, the processing unit 810 can perform the following steps: Figure 2 The distributed wireless network shown in the embodiment includes an aggregated base station entity and multiple distributed base station entities communicating with the aggregated base station entity. The method includes: deploying a pre-trained, verified and tested artificial intelligence / machine learning AI / ML model on the aggregated base station entity and the distributed base station entity according to a subscription use case, obtaining a model input data request for model reasoning, and the AI / ML model is obtained by model training, testing and verification in the aggregated base station entity or operation maintenance management OAM; the aggregated base station entity receives measurement data obtained by multiple distributed base station entities and terminal devices according to the model input data request as a model feature input; the model feature input is input into the pre-trained, verified and tested AI / ML model for model reasoning, and adjustment strategies and model performance feedback information of multiple distributed base station entities are output; the aggregated base station entity receives performance feedback data after multiple distributed base station entities and terminal devices execute the adjustment strategy, and sends the performance feedback data and model performance feedback information to OAM; OAM upgrades the AI / ML model according to the performance feedback data and the model performance feedback information, and deploys the upgraded AI / ML model on the aggregated base station entity and multiple distributed base station entities.

[0302] The storage unit 820 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 8201 and / or a cache memory unit 8202 , and may further include a read-only memory unit (ROM) 8203 .

[0303] The storage unit 820 may also include a program / utility 8204 having a set (at least one) of program modules 8205, such program modules 8205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0304] Bus 830 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0305] The electronic device 800 may also communicate with one or more external devices 840 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the system 800, and / or communicate with any device that enables the electronic device 800 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 850. In addition, the system 800 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 860. As shown, the network adapter 860 communicates with other modules of the electronic device 800 via a bus 830. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 800, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0306] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0307] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the above method of the present specification is stored. In some possible implementations, various aspects of the present invention can also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps according to various exemplary embodiments of the present invention described in the above "Exemplary Method" section of the present specification.

[0308] A program product for implementing the above method according to an embodiment of the present invention is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto, and in this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0309] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0310] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, in which readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Readable signal media may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0311] The program code embodied on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0312] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0313] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0314] In addition, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0315] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0316] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

Claims

1. A distributed wireless network upgrade method, applied to a distributed wireless network system, wherein the distributed wireless network system includes an aggregated base station entity, multiple distributed base station entities, terminal equipment, and operation, maintenance, and management (OAM), characterized in that: The method comprises: The aggregated base station entity sends a model input data request to the multiple distributed base station entities, where the model input data request is used to perform model reasoning on an AI / ML model of a subscribed use case, and the AI / ML model is obtained by model training, testing, and verification at the aggregated base station entity or the OAM; The aggregated base station entity receives measurement data obtained by the multiple distributed base station entities and terminal devices according to the model input data request as a model feature input quantity; The aggregated base station entity inputs the model feature input into the AI / ML model for model inference, and outputs adjustment strategies and model performance feedback information for the multiple distributed base station entities and the terminal devices; The aggregated base station entity receives performance feedback data after the multiple distributed base station entities and the terminal equipment execute the adjustment strategy, and sends the performance feedback data and model performance feedback information to the OAM; The OAM upgrades the AI / ML model according to the performance feedback data and the model performance feedback information, and deploys the upgraded AI / ML model to the aggregated base station entity and the multiple distributed base station entities, and the upgraded AI / ML model is stored in the OAM.

2. The method according to claim 1, characterized in that: Before the aggregated base station entity sends a model input data request to the multiple distributed base station entities, the method further includes: The aggregation base station entity sends an analysis subscription request message to the OAM to perform model training on the AI / ML model of the subscription use case according to the analysis subscription request message; The OAM sends an analysis subscription response message to the aggregate base station entity to indicate the AI / ML model deployment and the model feature input quantity required for model training.

3. The method according to claim 2, characterized in that: The analysis subscription request message includes at least one of a base station node ID, a subscription case ID, an analysis report target, and a parameter type contained in the analysis report, wherein: The subscription use case ID includes at least one of a network energy saving use case, a load balancing use case, a mobility management use case, and a slice management use case; The analysis report target includes at least one of OAM, a specific aggregated base station entity, a specific group of aggregated base station entities, a specific terminal device, and a group of terminal devices; The parameter types contained in the analysis report include at least one of the following: Base station temperature, humidity, voltage, and current; Base station physical radio resource utilization measurement, cell physical resource block PRB utilization; Measurement of the average number of radio resource control (RRC) connections in a cell; gNB-DU downlink packet loss rate measurement; Downlink Packet Data Convergence Protocol PDCP SDU packet loss rate of gNB-CU-UP; Delay measurement; Minimization of drive test (MDT) data, wherein the MDT data includes at least one of a reference signal received power (RSRP), a reference signal received quality (RSRQ), a signal to interference plus noise ratio (SINR), and a terminal device location information related to UE measurement; New radio NR cell ID; Timestamp.

4. The method according to claim 1, characterized in that: The model feature input includes at least one of the following: New radio NR cell ID; IP address information and port information used for input data interaction; Base station physical radio resource utilization measurement, cell PRB utilization; Measurement of the average number of RRC connections in a cell; Downlink packet loss rate measurement; MDT data; Service experience statistics and service usage forecasts for each terminal device.

5. The method according to claim 1, characterized in that: The aggregated base station entity receives measurement data obtained by the multiple distributed base station entities and the terminal devices according to the model input data request as a model feature input quantity, including: The distributed base station entity configures MDT measurement information for the terminal device according to the model input data request, wherein the measurement information includes at least one of the following: measurement trigger period, measurement area configuration information; The terminal device reports the measurement data of the terminal device obtained according to the measurement information to the distributed base station entity.

6. The method according to claim 5, characterized in that: The method further comprises: The distributed base station entity sends measurement data of the distributed base station entity and the terminal device to the aggregate base station entity.

7. The method according to claim 5, characterized in that: The measurement data of the terminal device includes at least one of the following: Terminal device ID; Serving cell ID; Terminal device location information; Terminal equipment measures related RSRP, RSRQ, and SINR; RSRP and RSRQ of the serving cell; RSRP and RSRQ of neighboring cells.

8. The method according to claim 1, characterized in that: The aggregated base station entity inputs the model feature input into the AI / ML model for model inference, and outputs adjustment strategies and model performance feedback information of the multiple distributed base station entities and the terminal devices, including: The aggregation base station entity outputs an analysis report, wherein the analysis report includes at least one of the following: Predicted wireless resource information; Forecast range; List of congested cells within the prediction range; A list of target cells carrying handover traffic within the predicted range; List of congestion relief cells within the forecast range; Prediction confidence; List of services; Root causes of energy consumption in residential areas; Current energy-saving cell status; Predicting cell energy efficiency; Recommended energy-saving decisions.

9. The method according to claim 8, characterized in that: The method further comprises: The multiple distributed base station entities and the terminal devices execute corresponding adjustment strategies according to the analysis report, wherein the adjustment strategies include at least one of UE switching and cell execution of energy-saving strategies.

10. The method according to claim 1, characterized in that: Before the aggregate base station entity receives the performance feedback data after the multiple distributed base station entities and the terminal device execute the adjustment strategy, the method further includes: The OAM sends feedback request information to the aggregation base station entity to indicate the required feedback data type; The aggregated base station entity sends a feedback request to the multiple distributed base station entities according to the feedback request information; The multiple distributed base station entities perform measurements according to the feedback request, obtain performance feedback data after executing the adjustment strategy, and send the performance feedback data to the aggregated base station entity.

11. The method according to claim 10, characterized in that: The performance feedback data after executing the adjustment strategy includes at least one of the following: Cell PRB utilization; Measurement of the average number of RRC connections in a cell; Cell packet loss rate measurement; Average delay; MDT data; Service experience analysis for each terminal device.

12. The method according to claim 1, characterized in that: The AI / ML model is obtained by model training, testing and verification in the aggregated base station entity or the OAM, including: Acquire historical measurement data of the plurality of distributed base station entities and terminal devices as training samples; Input the training samples into the AI / ML model to be trained for model reasoning, and output prediction adjustment strategies and historical model prediction performance parameters of the plurality of distributed base station entities and the terminal devices; Receiving historical performance feedback data after the distributed base station entity and the terminal device execute the prediction adjustment strategy; Determine whether the preset training stop condition is met; When the preset training stop condition is not met, the model parameters of the AI / ML model are adjusted according to the historical model prediction performance parameters and the historical performance feedback data, and the new training samples are input into the adjusted AI / ML model until the preset training stop condition is met; When the preset training stop conditions are met, the trained AI / ML model is obtained.

13. The method according to claim 1, characterized in that: The AI / ML model includes a network energy-saving model, a load balancing model, a mobility management model, and a slice management model.

14. A distributed wireless network system, characterized in that: It includes an aggregated base station entity, multiple distributed base station entities communicating with the aggregated base station entity, terminal equipment and operation maintenance management OAM, wherein: The aggregated base station entity is used to send a model input data request to the multiple distributed base station entities, the model input data request is used to perform model reasoning on the AI / ML model of the subscribed use case, and the AI / ML model is obtained by model training, testing and verification at the aggregated base station entity or the OAM; receiving measurement data measured by the multiple distributed base station entities and terminal devices according to the model input data request as model feature input; inputting the model feature input into the AI / ML model for model reasoning, and outputting adjustment strategies and model performance feedback information for the multiple distributed base station entities and the terminal devices; receiving performance feedback data after the multiple distributed base station entities and the terminal devices execute the adjustment strategy, and sending the performance feedback data and model performance feedback information to the OAM; The multiple distributed base station entities and terminal devices are used to receive the model input data request sent by the aggregated base station entity, perform measurements according to the model input data request, obtain measurement data and send it to the aggregated base station entity; and execute the adjustment strategy obtained by the aggregated base station entity through model reasoning, and send performance feedback data after executing the adjustment strategy to the aggregated base station entity; The OAM is used to upgrade the AI / ML model according to the performance feedback data and the model performance feedback information, and deploy the upgraded AI / ML model to the aggregated base station entity and the multiple distributed base station entities, and the upgraded AI / ML model is stored in the OAM.

15. An electronic device, characterized in that: include: processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the distributed wireless network upgrade method according to any one of claims 1 to 13 by executing the executable instructions.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the distributed wireless network upgrade method according to any one of claims 1 to 13 is implemented.

Images