CN117812672A

CN117812672A - Network energy consumption optimization method and base station

Info

Publication number: CN117812672A
Application number: CN202211167919.0A
Authority: CN
Inventors: 张化; 许森; 熊尚坤; 信金灿
Original assignee: China Telecom Corp Ltd
Current assignee: China Telecom Corp Ltd
Priority date: 2022-09-23
Filing date: 2022-09-23
Publication date: 2024-04-02

Abstract

The disclosure provides a network energy consumption optimization method and a base station, and belongs to the technical field of communication. The method comprises the following steps: the CU determines model reasoning output based on a network energy consumption optimization model, feature input data of DU statistics and UE measurement report data, and optimizes the network energy consumption of the UE in the network energy consumption optimization range based on the model reasoning output; the CU sends a feedback request message to the DU, wherein the feedback request message is used for requesting the network performance after the network energy consumption is optimized, and the feedback request message comprises a requested measurement object; the DU sends a feedback reply message to the CU, wherein the feedback reply message comprises a measurement object which can be provided; the DU sends a feedback data update message to the CU, the feedback data update message including the network performance feedback data. Based on the technical scheme provided by the embodiment of the disclosure, energy-saving decisions can be provided for the base station on the premise of guaranteeing the network service quality.

Description

Network energy consumption optimization method and base station

Technical Field

The disclosure belongs to the technical field of communication, and particularly relates to a network energy consumption optimization method and a base station.

Background

With the development of wireless communication technology, a 5G (5 th Generation Mobile Communication Technology, fifth generation mobile communication technology) network has characteristics of high rate, low latency, large capacity, and the like, compared with a network of a previous communication age.

Currently, in order to meet the key performance index requirements of the 5G network and the unprecedented growing demands of mobile users, the deployment amount of the base station is increased, so that the problems of high energy consumption, high carbon dioxide emission, high operation expenditure and the like are still the problems that need to be solved by operators at present, so that a way for providing energy-saving decisions for the base station on the premise of guaranteeing the service quality of the network is needed.

Disclosure of Invention

The embodiment of the disclosure aims to provide a network energy consumption optimization method and a base station, which can provide energy-saving decisions for the base station on the premise of guaranteeing network service quality.

In order to solve the above technical problems, the present disclosure is implemented as follows:

in a first aspect, an embodiment of the present disclosure provides a network energy consumption optimization method, which is applied to a first base station, where the first base station includes a CU (Centralized management node) and a DU (Distributed Unit, separate management node), and the method includes: the CU determines model reasoning output based on a network energy consumption optimization model, DU statistical characteristic input data and User Equipment (UE) measurement report data, and performs network energy consumption optimization on the UE in a network energy consumption optimization range based on the model reasoning output; the CU sends a feedback request message to the DU, wherein the feedback request message is used for requesting the network performance after the network energy consumption is optimized, and the feedback request message comprises a requested measurement object; the DU sends a feedback reply message to the CU, wherein the feedback reply message comprises a measurement object which can be provided; the DU sends a feedback data update message to the CU, the feedback data update message including network performance feedback data.

In a second aspect, embodiments of the present disclosure provide a base station, the base station comprising a centralized management node CU and a separate management node DU; the CU is used for: determining model reasoning output based on a network energy consumption optimization model, DU statistical characteristic input data and UE measurement report data, and optimizing network energy consumption of the UE in a network energy consumption optimization range based on the model reasoning output; the CU is further configured to send a feedback request message to the DU, where the feedback request message is used to request network performance after network energy consumption optimization, and the feedback request message includes a requested measurement object; the DU is configured to send a feedback reply message to the CU, where the feedback reply message includes an available measurement object; the DU is further configured to send a feedback data update message to the CU, where the feedback data update message includes network performance feedback data.

In a third aspect, embodiments of the present disclosure provide a base station comprising a processor, a memory and a program or instructions stored on the memory and executable on the processor, which when executed by the processor, implements the steps of the network energy consumption optimization method according to the first aspect.

In a fourth aspect, embodiments of the present disclosure provide a readable storage medium having stored thereon a program or instructions which, when executed by a processor, implement the steps of the network energy consumption optimization method according to the first aspect.

In a fifth aspect, embodiments of the present disclosure provide a chip, where the chip includes a processor and a communication interface, where the communication interface is coupled to the processor, and where the processor is configured to execute a program or instructions to implement the network energy consumption optimization method according to the first aspect.

In a sixth aspect, embodiments of the present disclosure provide a computer program product comprising instructions which, when run on a computer, cause the computer to perform the steps of the network energy consumption optimization method according to the first aspect.

In the embodiment of the disclosure, by deploying the functions of model training and model reasoning for network mobility optimization on the CUs of the 5G separation architecture base station, the CUs of the separation architecture base station can request AI/ML capability to the OAM for energy saving purposes, and the OAM is responsible for collecting and processing measurement statistics data of each node of the network according to preconfigured model indication information to perform offline or online training of the model, and deploying and updating the trained model to each CU. Based on the model indication information of OAM deployment, the base station CU configures measurement to the terminal and sends a data request to the DU so as to acquire data required by the online model training and model reasoning process and conduct model reasoning, and generates reasoning output information containing prediction information and decision information for network execution, so that network energy saving based on the AI/ML technology is realized. For example, the energy efficiency, the radio resource state, the energy consumption state of the base station and the adjacent base station, the UE performance data, the UE position and other information are used, the latest model configuration is updated each time before the network saves energy, the base station energy efficiency prediction and the base station radio resource prediction under different mobile scenes are analyzed according to the latest model, the reasoning output information containing the prediction information and the decision information is generated, the energy consumption and the load state of the base station are predicted by collecting various data in the wireless network, the energy saving decision is optimized, the more flexible energy saving strategy is dynamically configured for the base station in advance on the premise of ensuring the service quality of the network, and the energy consumption, the carbon dioxide emission and the operation expenditure can be reduced under the condition of increasing the deployment quantity of the base station.

Drawings

Fig. 1 is a schematic structural diagram of a separate base station according to an embodiment of the disclosure;

fig. 2 is a schematic diagram of a wireless network intelligent function architecture based on AI/ML technology according to an embodiment of the disclosure;

fig. 3 is a schematic diagram of CU and DU interaction flow of a separate base station according to an embodiment of the present disclosure;

fig. 4 is a schematic diagram of a network architecture for optimizing network energy consumption according to an embodiment of the present disclosure;

fig. 5 is one of flow diagrams of a network energy consumption optimization method according to an embodiment of the present disclosure;

FIG. 6 is a second flow chart of a network energy consumption optimization method according to an embodiment of the disclosure;

FIG. 7 is a third flow chart of a network energy consumption optimization method according to an embodiment of the disclosure;

fig. 8 is a schematic diagram of one possible structure of a base station according to an embodiment of the disclosure;

fig. 9 is a second possible structural diagram of a base station according to an embodiment of the disclosure;

fig. 10 is a schematic diagram of base station hardware according to an embodiment of the present disclosure.

Detailed Description

The following description of the technical solutions in the embodiments of the present disclosure will be made clearly and completely with reference to the accompanying drawings in the embodiments of the present disclosure, and it is apparent that the described embodiments are some embodiments of the present disclosure, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without inventive effort, based on the embodiments in this disclosure are intended to be within the scope of this disclosure.

The terms first, second and the like in the description and in the claims, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged, where appropriate, such that embodiments of the disclosure may be practiced in sequences other than those illustrated and described herein, and that the objects identified by "first," "second," etc. are generally of the same type and are not limited to the number of objects, e.g., the first object may be one or more. Furthermore, in the description and claims, "and/or" means at least one of the connected objects, and the character "/", generally means that the associated object is an "or" relationship.

It is noted that the techniques described in embodiments of the present disclosure are not limited to LTE (Long Term Evolution )/LTE-a (LTE-Advanced, evolution of LTE) systems, but may also be used in other wireless communication systems, such as CDMA (Code Division Multiple Access ), TDMA (Time Division Multiple Access, time division multiple access), FDMA (Frequency Division Multiple Access ), OFDMA (Orthogonal Frequency Division Multiple Access, orthogonal frequency division multiple access), SC-FDMA (Single-carrier Frequency-Division Multiple Access, single carrier frequency division multiple access), and other systems. The terms "system" and "network" in embodiments of the present application are often used interchangeably, and the techniques described may be used for both the above-mentioned systems and radio technologies, as well as other systems and radio technologies. However, the following description describes an NR system for purposes of example and NR terminology is used in much of the following description, although the techniques may also be applied to applications other than NR system applications, such as 6G (6 th Generation) communication systems.

The network energy consumption optimization method provided by the embodiment of the disclosure is described in detail below through specific embodiments and application scenarios thereof with reference to the accompanying drawings.

Fig. 1 is a schematic structural diagram of a separate base station according to an embodiment of the disclosure. As shown in FIG. 1, gNB 100 may include gNB-CU 101 and at least one gNB-DU 102. Wherein, gNB-CU 101 is connected with each gNB-DU 102 through F1 interface; the F1 interface is used for signaling interaction and data transmission between the gNB-CU 103 and the gNB-DU 104.

It should be noted that, the F1 interface may support separation of the control plane and the user plane, and the radio network layer and the transport network, and support interaction of UE association information and non-UE association information. To satisfy the new services and functions, one gNB-CU and a set of gNB-DUs are visible to other logical nodes, which gNB-CU may be separated in the Control Plane (CP) and the User Plane (UP). The F1 interface also facilitates interconnection between centralized control entities and discrete control entities provided by different vendors.

For convenience of explanation, in the embodiments of the present disclosure, a centralized management entity (or a centralized control entity, or a centralized unit) of a separate base station is denoted by CU, and a separate management entity (or a separate control entity, or a separate unit) of the separate base station is denoted by DU.

Fig. 2 is a schematic diagram of a wireless network intelligent function architecture based on AI (Artificial Intelligence, machine learning)/ML (Machine learning artificial intelligence) technology, where the architecture includes four main functions of a data collection function 201, a model training function 202, a model reasoning function 203, and an execution function (Actor) 204. Wherein the data collection functionality 201 provides input data to the model training functionality 202 and the model reasoning functionality 203; the input data contains measurements from the UE or different network entities, feedback from the executive function 204, and output from the AI/ML model. Model training function 202 performs AI/ML model training, validation and testing functions and is capable of generating model performance metrics as part of a model testing process. Model training function 202 is also responsible for data preprocessing, cleaning, formatting, and converting based on training data provided by the data collection function. The model inference function 203 provides AI/ML model inference output based on the trained model, which contains prediction and decision information for each node of the network to judge and execute. The model inference function 203 also has data processing capability. The executive function 204 receives the output of the model inference function 203 and triggers or performs a corresponding action.

Fig. 3 is a schematic diagram of CU and DU interaction flow of a separate base station provided in an embodiment of the present disclosure, where, as shown in fig. 3, S301 and S303 include the following:

s301, sending a feedback request message to the gNB-DU by the gNB-CU.

The feedback request message is used for requesting network performance after network energy consumption optimization, and the feedback request message comprises a measurement object requested by the gNB-CU.

S302, the gNB-DU sends a feedback response message to the gNB-CU.

The feedback response message may also be referred to as a feedback reply message, which includes a measurement object that may be provided by the gNB-DU.

S303, the gNB-DU sends a feedback data update message to the gNB-CU.

Wherein the feedback data update message includes network performance feedback data.

Fig. 4 is a schematic diagram of a network architecture for optimizing network energy consumption, which is provided in an embodiment of the present disclosure, and as shown in fig. 4, the network architecture includes OAM (Operation Administration and Maintenance, network management equipment) 400, a base station 401, a base station 402, and UEs accessed in coverage areas of the respective base stations; base station 402 is a neighbor base station of base station 401. The base station 401 is a separate base station, the CU of the base station 401 may request the capability of network energy consumption optimization to the OAM 400, based on the capability reply message sent by the OAM 400, trigger the DU of the base station 401 to acquire data of model training, the DU of the base station 401 may send feature input data of DU statistics to the CU and trigger the UE to measure and report, after the UE sends measurement report data to each DU of the base station 401, each DU of the base station 401 may send measurement report data to the CU, and the CU of the base station 401 may send training data request message to each DU to request training data required by DU report, after the CU acquires training data update message of each DU of the base station 401, the CU of the base station 401 may perform model training to obtain a model of network energy consumption optimization, then determine model reasoning output of the model of network energy consumption optimization based on the feature input data of the model and each statistic feature input data of the base station 401, UE measurement report data, and then perform network energy consumption optimization based on model reasoning output, after optimization, the CU of each DU of the base station 401 may send feedback request message to each DU of the base station 401, so as to determine network energy consumption optimization parameters of the network energy consumption optimization after optimization, thereby determining performance parameters of the network energy consumption optimization.

Fig. 5 is a flow chart of a network energy consumption optimization method according to an embodiment of the present disclosure, as shown in fig. 5, the method includes the following steps S501 to S504:

s501, the CU determines model reasoning output based on a network energy consumption optimization model, feature input information of DU statistics and UE measurement report data, and optimizes the network energy consumption of the UE in the network energy consumption optimization range based on the model reasoning output.

S502, the CU sends a feedback request message to the DU.

The feedback request message is used for requesting network performance after network energy optimization is performed on the UE in the network energy optimization range based on model reasoning output. The feedback request message may include the requested measurement object.

S503, the DU sends a feedback reply message to the CU.

Wherein the feedback reply message may indicate a measurement object that the DU may provide to the UE for indicating network performance.

The feedback request message indicates the IP address and the port address of the CU; the feedback reply message indicates the IP address and port address of the DU; the feedback request message and the feedback reply message are used for indicating the nodes of the separated base station to transmit mass model data through the established user plane channel.

S504, the DU sends a feedback data update message to the CU.

The embodiment of the disclosure provides a network energy consumption optimization method, after a CU determines model reasoning output based on a network energy consumption optimization model, characteristic input data of DU statistics and UE measurement report data, and performs network energy consumption optimization on UE in a network energy consumption optimization range based on the model reasoning output, the CU can request a feedback network performance after the network energy consumption optimization to a DU so as to determine whether an effect of the network energy consumption optimization meets requirements, specifically, the CU can send a feedback request message to the DU to request a measurement object indicating the network performance, and the DU can send a feedback reply message to the CU after receiving the feedback request message to indicate the available measurement object indicating the network performance; after that, the DU may send a feedback data update message including the network performance feedback data to the CU, so that the CU may determine an effect of optimizing network energy consumption, thereby providing an energy saving decision for the base station on the premise of guaranteeing network service quality, and reducing energy consumption, carbon dioxide emission, and operating expenditure under the condition that the deployment amount of the base station is suddenly increased.

Fig. 6 is an interactive flow chart of a network energy consumption optimization method provided by an embodiment of the present disclosure, as shown in fig. 3, may include the following S601 to S620:

optionally, the network energy consumption optimization method may include the following S601 and S602:

s601, the CU sends a capability request message to the OAM.

Wherein the capability request message indicates the function and model reasoning type requested by the first base station.

Illustratively, the capability request message may include at least one of an AI capability request message and an ML capability request message; the AI capability request message is used for triggering OAM to perform AI model offline training to acquire an AI function, and the ML capability request message is used for triggering OAM to perform ML model offline training to acquire an ML function.

Optionally, in an embodiment of the present disclosure, the capability request message may include at least one of the following 1-1 to 1-7:

1-1: a CU identification;

1-2: model indication information;

wherein the model indication information indicates an enabled model or function.

For example, the models referred to in the embodiments of the present disclosure may include at least one of an AL model and an ML model, and for example, the target capability request message may include at least one of AI model indication information and ML model indication information.

Optionally, the enabled models or functions include at least one of the following associated models: base station energy efficiency prediction and base station radio resource prediction.

For example, a model request message may include at least model indication information of a related model. The model indication information may be enumerated, for example, the model request message may include integer values 0, 1, 2, …, which respectively indicate models associated with base station energy efficiency prediction, base station radio resource prediction, and the like.

1-3: training request indication information;

specifically, the model request message may carry a training request indication identifier, which may indicate information required for achieving the expected performance by the sender of the model request message.

Wherein the training request indication information indicates at least one of the following A1 to A3:

a1, inputting information of model features;

for example, the model request message may carry a model feature input information identifier or model feature input information indication information to indicate the required model feature input information.

A2, training mode;

wherein the training pattern is indicative of a model algorithm.

For example, the model request message may carry a training mode identifier to indicate a required training mode, which may be an enumeration type.

Illustratively, the training pattern may include at least one of: linear regression, logistic regression, decision trees, support vector machines, random forests, and the like.

A3, model performance.

Optionally, the model properties include at least one of the following a11 to a 15:

a11, model parameters; wherein the model parameters indicate configuration variables inside the model, the configuration variables comprising at least one of: weights, biases, and learning rates, etc.

a12, reasoning the output name; the name of the reasoning output indicates the name of the model functional body to perform the reasoning output, and the name can be enumerated. Illustratively, the inferential output comprises at least one of: network load analysis, network performance analysis, and UE-related analysis. For example, the model functionality may be an AI model functionality or an ML model functionality.

a13, performance indexes; wherein the performance index indicates a performance index of the evaluation model function. Illustratively, the performance index includes at least one of accuracy and precision.

a14, performance score; wherein the performance score indicates a performance score of the model functionality when reasoning about a particular dataset. Illustratively, the allowable value of the performance score is 0-100. Depending on the nature of the model, the performance indicators of different types of models may differ. For example, models, metrics for digital prediction may include accuracy; the model of user classification, the metrics may include a combination of precision and recall.

a15, confidence score. Wherein the confidence score indicates a value of at least one of reliability and quality of a given decision generated by the model functionality. Illustratively, the allowable value of the performance score is 0-100. The lowest value identifies the lowest reliability level of the decision.

For example, a model performance identifier may be carried in the capability request message to indicate the required model performance. For example, the capability request message may carry a model parameter identifier, an inference output name identifier, a performance index identifier, a performance score identifier, and a confidence score identifier.

1-4: training data source indication information;

specifically, the training data source indication information indicates address information of the training data source requested by the acquisition target capability request message sender (i.e., CU or first base station).

Illustratively, the source of the training data may be indicated and distinguished by address information of the training data source; the data sources may be different network functions, including base stations, core networks, etc., as well as operators, etc. The detailed training data format may be a vendor specific data format.

1-5: model function context;

optionally, the model functionality context includes at least one of: a training context, an expected runtime context, and a first runtime context.

Wherein the training context indicates a state and condition associated with training; the expected run-time context indicates the context of the model functionality of the expected application, including the particular time conditions under which model training is performed, etc.; the first runtime context indicates a context of the application model.

1-6: request status;

the request state indicates the state of the training request and may be an enumeration type.

Illustratively, the training request state includes at least one of: "NOT STARTED", "in training" (TRAINING IN PROGRESS), "SUSPENDED", "completed" (fixed), and "cancelled" (CANCELLED).

1-7: cancel the request indication information;

wherein the cancellation request indication information indicates whether the sender cancels the training request.

Illustratively, if the attribute of the cancellation request indication information is set to "TRUE", the cancellation response training request; the cancel request indicates that the attribute of the information is set to "FALSE" and indicates that a response to the training request is required. Default to "FALSE".

1-8: pause request indication information.

Wherein the suspension request indication information indicates whether the sender suspends the training request.

For example, when the request state is not the "fixed" state, suspension may be indicated, i.e., the attribute of the suspension request indication information may be set to "TRUE"; if no suspension is required, the attribute of the suspension request indication information may be set to "FALSE".

S602, the OAM determines the range of the network energy consumption optimization area of the first base station according to the capability request message, performs offline model training by taking the stored historical data as training data according to the capability request message, and sends a capability reply message to the CU.

Optionally, in an embodiment of the present disclosure, the capability reply message may include at least one of the following 2-1 to 2-8:

2-1: the first base station indicates information;

2-2: model indication information;

2-3: priority indication information;

wherein the priority indication information indicates a priority of the training process.

Specifically, the training process of the AI model, the ML model may be arranged using priorities.

For example, a lower value may be used to indicate a higher priority and a higher value may be used to indicate a lower priority.

2-4: training request indication information;

wherein the training request indication information indicates at least one of: model feature input information, training patterns, and model performance.

Optionally, in an embodiment of the present disclosure, the model feature input information may include at least one of the following B1 to B3:

b1, inputting information by UE characteristics;

Optionally, the UE feature input information includes at least one of the following b11 to b 13:

b11, UE position information;

for example, the UE location information may include at least one of: UE coordinates, serving cell ID (identity) of the UE, and moving speed.

b12, UE performance measurement information;

for example, the UE performance measurement information includes at least one of: RSRP (Reference Signal Receiving Power, reference signal received power), RSRQ (Reference Signal Receiving Quality, reference signal received quality), SINR (Signal to Interference plus Noise Ratio ), uplink and downlink throughput, packet delay, and packet loss rate.

B2: the first base station characteristic input information;

that is, the OAM may indicate that the base station (or CU) that received the capability reply message needs to report the base station characteristic input information of the base station.

Optionally, the first base station characteristic input information includes at least one of: current energy efficiency, predicted energy efficiency, current radio resource status, predicted radio resource status, and UE mobility prediction.

Illustratively, the radio resource status includes at least one of: cell PRB (Physical Resource Block ) utilization, average RRC (Radio Resource Control ) connection number, and packet loss rate.

B3: the neighbor base station characteristics of the first base station input information.

That is, the OAM may indicate that the base station (or CU) that receives the model capability message needs to report the base station characteristic input information of the neighboring base stations of the base station.

Optionally, the neighbor base station characteristic input information of the first base station includes at least one of: current energy efficiency, predicted energy efficiency, current radio resource status, predicted radio resource status, and current energy consumption status.

Illustratively, the energy consumption state may include: active, high, low, inactive.

2-5: a training data provider list;

wherein the training data provider list indicates entities that have provided or should provide model-related data. Such as an entity for training or reasoning.

2-6: training data source indication information.

Optionally, after S602 described above, the following S603, S604, and S605 may also be included:

s603, CU indicates DU measurement and statistics of feature input data to be provided by the first base station and the neighbor base station of the first base station.

Specifically, after receiving the capability reply message of the OAM, the CU may send a training data request message to each DU corresponding to the first base station according to the feature input information indicated in the capability reply message, so as to trigger a measurement, statistics and recording process local to the first base station.

S604, the DU transmits feature input data of the DU statistics to the CU.

Alternatively, in the embodiment of the present disclosure, S603 and S604 described above are not sequentially executed with S605 to S608 described below, and may be executed simultaneously, or S603 and S604 may be executed first, or S605 to S608 may be executed first, which is not particularly limited in the embodiment of the present disclosure.

S605, the CU sends a measurement configuration to the DU.

Illustratively, in embodiments of the present disclosure, the measurement configuration described above may include at least one of the following 3-1 and 3-2:

3-1: RRM (Radio Resource Management ) measurement configuration;

wherein the RRM measurement configuration comprises at least one of: and carrying out periodic measurement triggering information and UE wireless measurement information.

Illustratively, the information for performing the periodic measurement trigger includes a trigger period and a recording period.

Illustratively, the UE wireless measurement information includes at least one of: RSRP, RSRQ, and SINR.

3-2: MDT (Minimization of Drive-Tests, minimization of drive Tests) measurement configuration.

Illustratively, the MDT measurement configuration includes at least one of: information supporting periodic measurement triggers, UE location information, and UE movement speed.

S606, the DU sends a measurement trigger message to the UE in the coverage area.

Wherein the measurement trigger message includes a measurement configuration.

Illustratively, the measurement trigger message includes a UE tracking ID.

Specifically, after receiving the measurement configuration sent by the CU, the DU may select an appropriate UE in the coverage area of the first base station to perform measurement triggering.

Alternatively, if the UE is not in the designated area or the PLMN (Public Land Mobile Network, serving public land mobile network) is not based on the managed PLMN list, the UE does not report the measurement data.

S607, the UE performs measurement based on the measurement configuration and reports the UE measurement report data to the DU.

Specifically, the UE performs corresponding measurement according to the measurement configuration sent by the DU, collects measurement information indicated by the DU, for example, measurement values of UE location information, RSRP, RSRQ, SINR of the UE, and if MDT measurement configuration is indicated in the measurement configuration, the UE completes collection and recording of corresponding measurement quantities according to the MDT measurement configuration.

Specifically, according to measurement configuration, the UE periodically transmits the collected relevant measurement information for model training to the DU, where the UE measurement report message includes UE measurement report data. The UE measurement report data can be used for model training and model reasoning. After receiving the report data sent by the UE based on the measurement configuration, the DU may input information according to the model feature indicated in the capability reply message, and send measurement statistics data of the relevant information to the CU as UE measurement report data.

Optionally, in an embodiment of the present disclosure, the UE measurement report data includes at least one of the following 4-1 to 4-5:

4-1: an ID of the UE;

4-2: UE location information;

4-3: UE movement speed;

4-4: UE performance measurement information;

4-5: a time stamp.

S608, the DU sends UE measurement report data to the CU.

Specifically, the DU transmits current UE performance information, predicted UE performance information, current UE trajectory information, predicted UE trajectory information, associated with the ID of the UE, to the CU.

It should be noted that, the above S607 and S608 may be performed periodically, and optionally, after the above S608, the following S609 may also be included:

s609, the CU transmits a training data request message to the DU.

Specifically, after receiving the UE measurement report data and the feature input data of the DU statistics, the CU sends an input data request message to the DU according to the model feature input information indicated by the capability reply message, so as to obtain information required by AI model network energy saving or obtain information required by ML model energy saving.

Optionally, in an embodiment of the present disclosure, the training data request message may include at least one of the following 5-1 to 5-4:

5-1: a CU identification;

5-2: the IP address (Internet Protocol Address ) and port address of the CU;

Specifically, if the training data request message includes the IP address and the port address of the CU, it indicates that massive AI model data or ML model data is to be transmitted through the established user plane channel.

5-3: reporting a cell list;

the report cell list comprises cell report information and a report period.

Illustratively, the cell reporting information includes at least one of: cell ID, SSB (Synchronization Signal and PBCH Block, synchronization signal and physical broadcast channel block) report list, and SSB index; the reporting period indicates the average window length of all measurement objects.

5-4: training the input information.

Wherein the training input information indicates data reported by measurements provided by at least one cell under the DU.

Optionally, the data reported by the measurement provided by each cell of the at least one cell under DU includes at least one of the following C1 and C2:

c1: inputting information by UE characteristics;

wherein the UE feature input information includes at least one of: UE ID, UE location information, UE movement speed, UE performance measurement information.

C2: the first base station characteristic input information;

wherein the first base station characteristic input information may include at least one of: current network energy efficiency, predicted network energy efficiency, current radio network resource status, predicted radio network resource status, and UE trajectory prediction.

5-5: a registration request for training input information.

If the registration request indicates to start, the DU (i.e. the receiving base station node) starts measurement according to the indication in the feature input information list; or,

if the registration request indicates to stop, the DU stops measuring and reporting; or,

if the registration request indicates addition, adding the measurement quantity or the predicted value corresponding to the reasoning input information with the added identification indication to a given training input information reporting list.

The reasoning input information is the newly added characteristic input information required by model reasoning indicated by the CU to the DU.

Optionally, after the DU receives the training data request message sent by the CU, if the DU may provide some or all of the measurement objects requested by the training data request message, the following S610a to S611 may be performed, and if the DU cannot provide any of the measurement objects requested by the training data request message, the following S610b may be performed.

S610a, DU sends a training data reply message to the CU.

Wherein, in the disclosed embodiment, the training data reply message includes at least one of the following 6-1 to 6-3:

6-1: DU identification;

6-2: IP address and port address of the DU;

where the training data reply message includes the IP address and port address of the DU, the IP address and port address of the DU indicate that model data transmission between base station nodes (i.e., between the CU and the DU) is to be performed through the established user plane channel.

6-3: a first threshold diagnostic identifier.

Wherein the first critical diagnostic identifier indicates information which is not understood or lost in the information received by the first DU.

S611, the DU sends a training data update message to the CU.

The training data update message comprises input information reporting data requested by the CU, wherein the input information reporting data comprises measurement reporting data and a time stamp provided by at least one cell under the DU.

Illustratively, the input information reporting data includes measurement reporting data provided by each of the at least one cell: UE characteristic input information and base station characteristic input information; the UE feature input information includes at least one of: UE ID, UE location information, UE movement speed, and UE performance measurement information; the base station characteristic input information comprises at least one of the following: UE trajectory prediction, current network energy efficiency, predicted network energy efficiency, current network radio resource status, and predicted network radio resource status.

S610b, if the DU cannot provide any measurement object requested by the training data request message, the DU sends a training data failure message to the CU.

Wherein, in the disclosed embodiment, the training data failure message includes at least one of the following 7-1 to 7-4:

7-1: DU identification;

7-2: the reason for failure;

illustratively, the failure cause includes at least one of: radio network layer reasons, transport layer reasons, and protocol reasons.

7-3: a second critical diagnostic identifier;

wherein the second critical diagnostic identifier indicates a message that is not understood, lost or contains a logical error in the messages received by the DU;

7-4: second waiting for a retransmission time;

wherein the second waiting retransmission time indicates a time at which the CU re-initiates the training data request message.

And S612, the CU performs model training according to the training data updating message sent by the DU to obtain a network energy consumption optimization model.

The UE periodically reports the UE measurement report data required by the model training and the model reasoning to the DU according to the measurement configuration.

Optionally, the above S603 may be repeatedly performed to S612, so that a network energy consumption optimization model with better performance may be trained.

S613, the CU determines model reasoning output based on the network energy consumption optimization model, the feature input data of DU statistics and the UE measurement report data, and optimizes the network energy consumption of the UE in the network energy consumption optimization range based on the model reasoning output.

Specifically, the CU may perform a model inference function based on feature input data of the base station itself (i.e., DU) measurement statistics, collected UE measurement report data, and generate AI model inference output or ML model inference output including prediction information and decision information, so as to allow the network to analyze or perform a corresponding operation.

The CU, for example, performs model reasoning according to the indication information output by the functional entity, for example, the recommended energy-saving cell list, where each energy-saving cell enters a corresponding energy-saving state, and the UE in the energy-saving cell switches to a cell with a corresponding bearer flow according to the recommended candidate cell ID list, so as to achieve network energy saving on the premise of guaranteeing UE user experience.

S614, the CU sends a feedback request message to the DU.

The feedback request message is used for requesting network performance after network energy consumption optimization, and the feedback request message comprises a requested measurement object.

Specifically, the CU sends feedback request messages to DUs to obtain network performance feedback data to optimize model performance.

Optionally, in an embodiment of the disclosure, the feedback request message includes at least one of:

8-1: a CU identification;

8-2: the IP address and port address of the CU;

if the feedback request message includes the IP address and the port address of the CU, the feedback request message indicates that massive AI model data or ML model data is to be transmitted through the established user plane channel.

8-3: reporting a cell list;

8-4: and feeding back indication information.

Wherein, the feedback indication information indicates measurement report data provided by at least one cell under the DU.

Specifically, the measurement report data provided by each cell of the at least one cell indicated in the feedback indication information includes at least one of: radio resource status, energy efficiency, handover UE performance measurement information, and system KPI (Key Performance Indicator ) where the first base station is located.

The method comprises the steps that UE performance measurement information is switched, wherein the UE performance measurement information after switching indicates the UE performance measurement information after switching; the KPI of the system where the first base station is located comprises at least one of the following: throughput, delay, current RLF (Radio Link Failure, radio link failure rate) of the system in which the first base station is located, and RLF of neighbor systems of the system in which the first base station is located.

Optionally, after the DU receives the feedback request message sent by the CU, if the DU may provide a part or all of the measurement objects requested by the feedback request message, the DU may initiate corresponding measurement, and S615a and S616 described below may be performed, and if the DU cannot provide any of the measurement objects requested by the feedback request message, S615b described below may be performed.

S615a, DU sends a feedback reply message to the CU.

Wherein the feedback reply message includes the available measurement object.

Optionally, the feedback request message and the feedback reply message are used for indicating that mass model data transmission is performed between nodes of the split base station through the established user plane channel.

Optionally, in an embodiment of the present disclosure, the feedback reply message includes at least one of the following 9-1 to 9-3:

9-1: DU identification;

9-2: IP address and port address of the DU;

wherein, the IP address and the port address of the DU are used for establishing a channel for transmitting data with the CU.

9-3: a first critical diagnostic identifier;

wherein the first critical diagnostic identifier indicates an unintelligible or missing message among the messages received by the CU.

S616, the DU sends a feedback data update message to the CU.

Optionally, in an embodiment of the present disclosure, the feedback data update message includes at least one of the following 10-1 and 10-2:

10-1: a cell report list;

10-2: reporting data by the feedback information;

the feedback information reporting data comprises measurement reporting data and a time stamp provided by at least one cell under the DU.

It may be understood that the measurement report data provided by at least one cell in the feedback information report data may refer to the measurement report data provided by at least one cell requested in the feedback request message, which is not described herein.

S616b, if the DU cannot provide any measurement object requested by the feedback request message, the DU sends a feedback data failure message to the CU.

Specifically, if any requested feedback data measurement object cannot be provided, the DU may send a feedback data failure message with the appropriate cause value to the CU.

Optionally, in an embodiment of the present disclosure, the feedback data failure message includes at least one of:

11-1: DU identification;

11-2: the reason for failure;

for example, the failure cause in the feedback data failure message may refer to the failure cause in the training data failure message, which is not described herein.

11-3: a second critical diagnostic identifier;

wherein the second critical diagnostic identifier indicates a message that is not understood, lost or contains a logical error in the messages received by the DU.

11-4: a first waiting retransmission time;

wherein the first waiting retransmission time indicates a time when the CU re-initiates the feedback request message.

It can be understood that in the embodiment of the disclosure, a signaling flow related to AI/ML data interaction in the energy saving process of the separation architecture base station is designed, wherein the request message and the reply message carry the IP address and the port address of the base station node and are respectively used for starting the measurement of the required data and reporting the measurement condition of the data. The data update message is used for realizing massive AI/ML data transmission, and enables the AI/ML technology to truly enable the wireless network. Furthermore, the patent realizes a network energy-saving scheme based on federal learning of the base station with a separated architecture, and a base station centralized management entity is used as a network aggregation function body and is responsible for network data collection, model training, model updating and issuing, model reasoning and feedback data information collection processes. The feedback data indicate the relevant data used for monitoring the performance of the network or the terminal after the network or the terminal executes the reasoning output operation, the centralized management entity can realize the model optimization function through the signaling flow, the energy saving effect of the network is improved, and the realization of the in-network intelligence in the future is promoted by one progress.

Generally, in the related art, the wireless network energy conservation depends on the mode of centralized statistics of coverage information and load information of each base station, so that the OAM needs to collect and process measurement data from each base station node and terminals under the coverage area thereof to form a large amount of data transmission signaling, thereby affecting network communication and calculation efficiency and being unfavorable for realizing efficient intelligent network energy conservation.

The network energy consumption optimization method provided by the embodiment of the disclosure can collect and process the network data information of each separated architecture base station based on the computing capability and the data storage capability of the centralized management entity of the 5G separated architecture base station, and is more beneficial to realizing the federal learning mode facing the network AI under the 5G eMBB large connection scene. Specifically, by disposing the model training and model reasoning functional body in the centralized management entity, the centralized management entity can measure and collect data required by model training or model reasoning based on the model issued by the OAM, and execute the model training and model reasoning process to generate reasoning output information containing prediction information and decision information, thereby realizing intelligent network energy saving based on the federal learning technology.

The embodiment of the disclosure provides a network energy consumption optimization method, by deploying a model training and model reasoning function body of network mobility optimization on a CU of a 5G separation architecture base station, the CU of the separation architecture base station can request AI/ML capability to OAM for energy saving purpose, the OAM is responsible for collecting and processing measurement statistical data of each node of a network according to preconfigured model indication information so as to perform offline or online training of the model, and deploying and updating the trained model to each CU. Based on the model indication information of OAM deployment, the base station CU configures measurement to the terminal and sends a data request to the DU so as to acquire data required by the online model training and model reasoning process and conduct model reasoning, and generates reasoning output information containing prediction information and decision information for network execution, so that network energy saving based on the AI/ML technology is realized. For example, the energy efficiency, the radio resource state, the energy consumption state of the base station and the adjacent base station, the UE performance data, the UE position and other information are used, the latest model configuration is updated each time before the network saves energy, the base station energy efficiency prediction and the base station radio resource prediction under different mobile scenes are analyzed according to the latest model, the reasoning output information containing the prediction information and the decision information is generated, the energy consumption and the load state of the base station are predicted by collecting various data in the wireless network, the energy saving decision is optimized, the more flexible energy saving strategy is dynamically configured for the base station in advance on the premise of ensuring the service quality of the network, and the energy consumption, the carbon dioxide emission and the operation expenditure can be reduced under the condition of increasing the deployment quantity of the base station.

Examples:

fig. 7 is an interactive flow chart of a network energy consumption optimization method provided by an embodiment of the present disclosure, where, as shown in fig. 7, the following S701 to S718 may be included:

s701, CU 1 sends an AI/ML capability request message to OAM.

The capability request message is used for triggering OAM to perform AI/ML offline model training so as to obtain related AI/ML functions. The AI/ML function of the request and the type of reasoning of the object model are indicated in the capability request message.

Specifically, after the OAM receives the capability request message, the network energy consumption optimization scope is determined and the first base station is set as the anchor base station. And the OAM takes the stored historical data as training data according to the capability request message indication information to perform offline model training.

S702, the OAM sends an AI/ML capability response message to the CU 1.

Wherein, after CU 1 receives the capability response message, CU 1 triggers measurement and collection of base station characteristic input information and neighbor base station characteristic input information according to the model characteristic input information related indication information,

s703, CU 1 sends a measurement configuration to CU 1 to collect UE feature input information.

The UE measurement configuration may include RRM measurement and MDT measurement, among others.

S704, DU 1 receives measurement configuration, and selects proper UE to perform measurement triggering.

It should be noted that if the UE is not in the designated area or the serving PLMN is not in the administrative-based PLMN list, such UE will not report measurement data.

And S705, the UE performs corresponding measurement according to the measurement configuration, collects the indicated measurement information, and completes recording and collection of corresponding measurement quantity according to MDT measurement.

For example, the measurement information includes UE location information, UE RSRP, RSRQ, SINR measurements, etc., while at the same time.

S706, the UE reports the UE measurement report message to the DU 1.

The UE measurement report message includes relevant measurement information for model training.

S707, the DU 1 receives the UE measurement report data, and inputs information indication information according to the model characteristics in the capability reply message, and sends UE information to the CU 1.

S708, CU 1 transmits a training data request message to DU 1.

S709a, DU 1 transmits a training data reply message to CU 1.

The measurement can be successfully started and corresponding measurement object information of the measurement information can be provided by indicating the measurement object for the request through the training data reply message.

S709b, DU 1 transmits a training data failure message to CU 1.

S710, DU 1 transmits a training input data update message to CU 1.

The DU 1 sends a training input data update message to report measurement information required for model training requested by the CU 1.

The CU 1 receives training input data information from a base station and training input information of the UE in the energy network consumption optimization area range, and processes and divides the data set into a training data set, a verification data set and a test data set. OAM carries out online model training based on the data set, and parameter adjustment and feature optimization are carried out through model training, verification and testing processes so as to obtain a more accurate model.

S711, CU 1 obtains an optimal model through online model training.

Wherein, CU 1 sends the optimized model related indication information to DU 1 through the model update message.

S712, the UE reports the UE measurement to the DU 1.

S713, DU 1 sends UE measurement report information to CU 1

Specifically, CU 1 sends an inference input data request message to DU 1, instructing DU 1 to provide measurement or prediction information required for model inference added to it for CU 1 to perform model inference. The DU 1 receives the reasoning data request message, indicates a measurement object for the request through the reasoning data reply message, and can successfully start measurement and provide corresponding measurement object information of the measurement information. The DU 1 transmits an inference input data update message to the CU 1 to report measurement information required for model inference requested by the CU 1.

S714, CU 1 executes model reasoning function based on the feature input information of DU 1 measurement statistics and the UE measurement report data collected by DU 1, generates AI/ML model reasoning output containing prediction information and decision information, and is used for the network to analyze or execute operation.

S715, the first base station performs network energy saving according to the indication information output by the model reasoning functional body.

For example, the model reasoning function body output comprises a recommended energy-saving cell ID list, each energy-saving cell enters a corresponding energy-saving state, UE in the energy-saving cell is switched to a cell with corresponding bearing flow according to the recommended candidate cell ID list, and network energy saving is realized on the premise of guaranteeing UE user experience.

S716, CU 1 transmits a feedback request message to DU 1.

In particular, model performance may be optimized based on the user plane channels already established for transmitting AI/ML data to obtain network performance feedback data.

Specifically, the DU 1 receives the feedback request message, and starts corresponding measurement according to the request indication information to perform feedback reply. If DU 1 can start some or all of the measurement objects in the feedback request message, S717a described below may be executed to indicate the measurement that can be started to CU 1, and feedback data may be transmitted to CU 1 through S718 described below. If the DU 1 cannot start any measurement object in the feedback request message, S717b described below may be performed.

S717a, DU 1 transmits a feedback response message to CU 1.

S717b, DU 1 sends a feedback failure message to CU1

S718, DU 1 sends a feedback update message to CU 1.

It should be noted that, in the network energy consumption optimizing method provided by the embodiment of the present disclosure, the execution body may also be a network energy consumption optimizing device, or a control module in the network energy consumption optimizing device for executing the network energy consumption optimizing method. In the embodiment of the present disclosure, a method for executing network energy consumption optimization by using a network energy consumption optimization device is taken as an example, and the network energy consumption optimization device provided by the embodiment of the present disclosure is described.

Fig. 8 is a schematic structural diagram of a base station according to an embodiment of the present disclosure, as shown in fig. 8, a base station 800 includes: CU 801 and DU 802; CU 801 is for: determining model reasoning output based on a network energy consumption optimization model, DU statistical characteristic input data and UE measurement report data, and optimizing network energy consumption of the UE in a network energy consumption optimization range based on the model reasoning output; CU 802, further configured to send a feedback request message to the DU, where the feedback request message is used to request network performance after network energy consumption optimization, and the feedback request message includes a requested measurement object; a DU 802 configured to send a feedback reply message to the CU, where the feedback reply message includes an available measurement object; DU 802 is further configured to send a feedback data update message to the CU, where the feedback data update message includes network performance feedback data.

The base station 800 provided in the embodiment of the present disclosure can implement each process implemented by the embodiments of the methods of fig. 1 to 7, and in order to avoid repetition, a description is omitted here.

Optionally, as shown in fig. 9, the embodiment of the present disclosure further provides a base station 900, including a processor 901, a memory 902, and a program or an instruction stored in the memory 902 and capable of being executed on the processor 901, where the program or the instruction implements each process of the embodiment of the network energy consumption optimization method described above and can achieve the same technical effect, and for avoiding repetition, a description is omitted herein.

It should be noted that, the network entity or the base station 1000 shown in fig. 10 is only an example, and should not impose any limitation on the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 10, the base station 1000 includes a central processing unit (Central Processing Unit, CPU) 1001 which can perform various appropriate actions and processes according to a program stored in a ROM (Read Only Memory) 1002 or a program loaded from a storage section 1008 into a RAM (Random Access Memory ) 1003. In the RAM 1003, various programs and data required for system operation are also stored. The CPU 1001, ROM 1002, and RAM 1003 are connected to each other by a bus 1004. An I/O (Input/Output) interface 1005 is also connected to bus 1004.

The following components are connected to the I/O interface 1005: an input section 1006 including a keyboard, a mouse, and the like; an output portion 1007 including a CRT (Cathode Ray Tube), an LCD (Liquid Crystal Display ), and the like, and a speaker, and the like; a storage portion 1008 including a hard disk or the like; and a communication section 1009 including a network interface card such as a LAN (Local Area Network, wireless network) card, a modem, or the like. The communication section 1009 performs communication processing via a network such as the internet. The drive 1010 is also connected to the I/O interface 1005 as needed. A removable medium 1011, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like, is installed as needed in the drive 1010, so that a computer program read out therefrom is installed as needed in the storage section 1008.

In particular, according to embodiments of the present disclosure, the processes described below with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flowcharts. In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion 1009, and/or installed from the removable medium 1011. When the computer program is executed by the central processing unit (CPU 1001), various functions defined in the system of the present application are executed.

The embodiment of the present disclosure further provides a readable storage medium, where a program or an instruction is stored, and when the program or the instruction is executed by a processor, the processes of the above-mentioned network energy consumption optimization method embodiment are implemented, and the same technical effects can be achieved, so that repetition is avoided, and no further description is given here.

Wherein the processor is a processor in the electronic device described in the above embodiment. The readable storage medium includes computer readable storage medium such as ROM, RAM, magnetic disk or optical disk.

The embodiment of the disclosure further provides a chip, where the chip includes a processor and a communication interface, where the communication interface is coupled to the processor, and the processor is configured to run a program or an instruction, implement each process of the above-mentioned network energy consumption optimization method embodiment, and achieve the same technical effect, so that repetition is avoided, and no further description is given here.

It should be understood that the chips referred to in the embodiments of the present disclosure may also be referred to as system-on-chip chips, chip systems, or system-on-chip chips, etc.

The embodiments of the present disclosure provide a computer program product including instructions, which when executed on a computer, cause the computer to perform the steps of the network energy consumption optimization method described above, and achieve the same technical effects, and are not repeated herein.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element. Furthermore, it should be noted that the scope of the methods and apparatus in the embodiments of the present disclosure is not limited to performing the functions in the order shown or discussed, but may also include performing the functions in a substantially simultaneous manner or in an opposite order depending on the functions involved, e.g., the described methods may be performed in an order different from that described, and various steps may also be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present disclosure may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk), including several instructions for causing a terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method described in the embodiments of the present disclosure.

The embodiments of the present disclosure have been described above with reference to the accompanying drawings, but the present disclosure is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and many forms may be made by those of ordinary skill in the art without departing from the spirit of the disclosure and the scope of the claims, which are all within the protection of the present disclosure.

Claims

1. A network energy consumption optimization method applied to a first base station, wherein the first base station comprises a centralized management node CU and a separate management node DU, the method comprising:

the CU determines model reasoning output based on a network energy consumption optimization model, feature input data of DU statistics and User Equipment (UE) measurement report data, and optimizes the network energy consumption of the UE in the network energy consumption optimization range based on the model reasoning output;

the CU sends a feedback request message to the DU, wherein the feedback request message is used for requesting the network performance after the network energy consumption is optimized, and the feedback request message comprises a requested measurement object;

the DU sends a feedback reply message to the CU, wherein the feedback reply message comprises a measurement object which can be provided;

The DU sends a feedback data update message to the CU, the feedback data update message including network performance feedback data.

2. The method of claim 1, wherein the feedback request message comprises at least one of:

the CU identification, the IP address and the port address of the CU, the report cell list and the feedback indication information;

the IP address and the port address of the CU are used for establishing a channel for transmitting data with the DU; the report cell list comprises cell report information and a report period; the cell reporting information comprises at least one of the following: cell identification ID, synchronization signal and list on physical broadcast channel block SSB and SSB index; the reporting period indicates the average window length of all the measurement objects; the feedback indication information indicates measurement report data provided by at least one cell under the DU.

3. The method of claim 2, wherein the feedback reply message comprises at least one of:

the DU identification, the IP address and port address of the DU, a first critical diagnostic identification;

the IP address and the port address of the DU are used for establishing a channel for transmitting data with the CU; the first critical diagnostic identifier indicates an unintelligible or missing message among the messages received by the CU.

4. A method according to claim 3, wherein the feedback data update message comprises at least one of: the cell report list and the feedback information report data;

5. The method according to claim 2 or 4, wherein the measurement report data provided by each of the at least one cell comprises at least one of: radio resource state, energy efficiency, switching UE performance measurement information and a system key performance index KPI of a first base station;

the UE performance measurement information indicates UE performance measurement information after handover, and the UE performance measurement information includes at least one of the following: terminal Reference Signal Received Power (RSRP), terminal Reference Signal Received Quality (RSRQ), signal-to-interference plus noise ratio (SNIR), uplink and downlink throughput, packet delay and packet loss rate; the system KPI where the first base station is located comprises at least one of the following: throughput, delay, current radio link failure rate RLF of the system in which the first base station is located, and RLF of a neighbor system of the system in which the first base station is located.

6. The method according to claim 1 or 2, wherein after the CU sends a feedback request message to the DU, the method further comprises:

if the DU cannot provide any measurement object requested by the feedback request message, the DU sends a feedback data failure message to the CU;

wherein the feedback data failure message includes at least one of: the DU identification, the failure cause, the second critical diagnostic identification, and the first waiting retransmission time; the failure cause includes at least one of: radio network layer reasons, transport layer reasons, and protocol reasons; the second critical diagnostic identifier indicates a message that is not understood, lost or contains a logical error in the messages received by the DU; the first waiting retransmission time indicates a time when the CU reinitiates a feedback request message.

7. The method of claim 1, wherein prior to the CU determining a model inference output based on the network energy consumption optimization model, the feature input data for DU statistics, and the UE measurement report data, the method further comprises:

the CU sends a capability request message to the OAM, wherein the capability request message indicates the function and model reasoning type requested by the first base station;

And the CU receives the capability reply message sent by the network management equipment OAM.

8. The method of claim 7, wherein the capability request message comprises at least one of: CU identification, model indication information, training request indication information, training data source indication information, model function context, request status, cancel request indication information, and pause request indication information.

9. The method of claim 8, wherein the step of determining the position of the first electrode is performed,

the model indication information indicates an enabled model or function comprising a model associated with at least one of: base station energy efficiency prediction and base station wireless resource prediction;

the training request indication information indicates at least one of: model feature input information, training modes and model performance;

the training data source indication information indicates address information of a training data source requested by the CU;

the model functionality context includes at least one of: a training context, an expected runtime context, and a first runtime context; the training context indicates states and conditions related to training; the expected run-time context indicates a context in which the model energy is expected to be applied, including a preset time condition for model training; the first runtime context indicates a context of an application model;

The request state indicates a state of a training request, the training request state including at least one of: not started, training, paused, completed and cancelled;

the cancellation request indication information indicates whether the sender cancels the training request;

the suspension request indication information indicates whether the sender suspends the training request.

10. The method of claim 9, wherein the capability reply message comprises at least one of: the first base station indication information, the model indication information, the priority indication information, the training request indication information, the training data provider list, and the training data source indication information;

wherein the priority indication information indicates the priority of the training process; the training data provider list indicates entities that have provided or should provide model-related data.

11. The method according to claim 9 or 10, wherein the training pattern indicates a model algorithm comprising at least one of: linear regression, logistic regression, decision trees, support vector machines, and random forests; the model performance includes at least one of: model parameters, inference output names, performance indicators, performance scores, confidence scores;

Wherein the model parameters are indicative of configuration variables inside the model, the configuration variables comprising at least one of: weight, bias, learning rate; the inference output name indicates the name of the model functional body for performing inference output, and the inference output comprises at least one of the following: network load analysis, network performance analysis, and UE-related analysis; the performance index indicates a performance index of the evaluation model function, the performance index including at least one of accuracy and precision; the performance score indicates a performance score of the model functionality when reasoning about a particular dataset; the confidence score indicates a value of at least one of reliability and quality of a given decision generated by the model functionality.

12. The method of claim 10, wherein the model feature input information comprises at least one of: UE characteristic input information, first base station characteristic input information and neighbor base station characteristic input information of the first base station.

13. The method of claim 12, wherein the UE feature input information comprises at least one of: UE location information and UE performance measurement information;

Wherein the UE location information includes at least one of: UE coordinates, serving cell ID of UE, and moving speed; the UE performance measurement information includes at least one of: RSRP, RSRQ, SINR of UE, uplink and downlink throughput, packet delay, and packet loss rate.

14. The method of claim 12, wherein the first base station characteristic input information comprises at least one of: current energy efficiency, predicted energy efficiency, current radio resource state, predicted radio resource state, UE mobility prediction;

wherein the radio resource status includes at least one of: physical resource block, PRB, utilization of a cell, average radio resource control, RRC, connection number, and packet loss rate.

15. The method of claim 12, wherein the neighbor base station characteristic input information of the first base station comprises at least one of: current energy efficiency, predicted energy efficiency, current radio resource status, predicted radio resource status, and current energy consumption status.

16. The method according to claim 7, wherein after the CU receives the capability reply message sent by the OAM of the network management device, the method further comprises:

the CU indicates the DU to count feature input data which needs to be provided by the first base station and the adjacent base station of the first base station, and sends measurement configuration to the DU;

And the DU transmits UE measurement report data and feature input data of the DU statistics to the CU.

17. The method of claim 16, wherein the measurement configuration comprises at least one of: radio resource management, RRM, measurement configuration and minimization of drive tests, MDT, measurement configuration;

wherein the RRM measurement configuration comprises at least one of: information of periodic measurement triggering and UE wireless measurement information are carried out; the MDT measurement configuration includes at least one of: supporting periodic measurement triggering information, UE position information and UE moving speed; the information for performing periodic measurement triggering comprises a triggering period and a recording period; the UE wireless measurement information includes at least one of: RSRP, RSRQ, and SINR.

18. The method of claim 17, wherein the UE does not report measurement data if the UE is not in a designated area or a serving public land mobile network, PLMN, is not on a management-based PLMN list.

19. The method of claim 1, 16 or 17, wherein the UE measurement report data comprises at least one of: UE ID, UE location information, UE movement speed, UE performance measurement information, and time stamp; the UE performance measurement information includes at least one of: RSRP, RSRQ, SINR, uplink and downlink throughput, packet delay, and packet loss rate.

20. The method of claim 1, wherein prior to the CU determining a model inference output based on the network energy consumption optimization model, the feature input data for DU statistics, and the UE measurement report data, the method further comprises:

the CU sends a training data request message to the DU, wherein the training data request message comprises a requested measurement object;

the DU sends a training data reply message to the CU, wherein the training data reply message comprises a measurement object which can be provided;

the DU sends training data update information to the CU;

and the CU performs model training according to the training data updating message sent by the DU to obtain a network energy consumption optimization model.

21. The method of claim 20, wherein the training data request message comprises at least one of: the CU identification, the IP address and the port address of the CU, the report cell list, training input information and a registration request of the training input information;

the report cell list comprises cell report information and a report period; the cell reporting information comprises at least one of the following: cell ID, SSB report list, SSB index; the reporting period indicates the average window length of all the measurement objects; the training input information indicates measurement report data provided by at least one cell under the DU.

22. The method of claim 21, wherein the measurement report data provided by each of the at least one cell comprises at least one of: base station input information of a base station to which a cell belongs and UE related input information of an access cell;

wherein the base station input information includes at least one of: current network energy efficiency, predicted network energy efficiency, current network radio resource status, predicted network radio resource status, and UE trajectory prediction; the UE-related input information includes at least one of: UE ID, UE location information, UE movement speed, and UE performance measurement information; the UE performance measurement information includes at least one of: RSRP, RSRQ, uplink and downlink throughput, packet delay, and packet loss rate of the UE.

23. The method of claim 21, wherein the step of determining the position of the probe is performed,

if the registration request indicates to start, the DU starts measurement according to the indication in the feature input information list; or,

if the registration request indicates stopping, the DU stops measuring and reporting; or,

and if the registration request indicates addition, adding the measurement quantity or the predicted value corresponding to the reasoning input information with the added identification indication to a given training input information reporting list.

24. The method of claim 20, wherein the training data reply message comprises at least one of: the DU identification, the IP address and port address of the DU, a first critical diagnostic identification;

the first critical diagnosis identifier indicates information which is not understood or lost in the information received by the first DU, wherein the IP address and the port address of the DU indicate that model data is transmitted between base station nodes through the established user plane channel.

25. The method of claim 20, wherein after the CU sends a training data request message to the DU, the method further comprises:

if the DU cannot provide any measurement object requested by the training data request message, the DU sends a training data failure message to the CU;

wherein the training data failure message includes at least one of: the DU identification, the failure reason, the second critical diagnosis identification and the second waiting retransmission time; the failure cause includes at least one of: radio network layer reasons, transport layer reasons, and protocol reasons; the second critical diagnostic identifier indicates a message that is not understood, lost or contains a logical error in the messages received by the DU; the second wait for retransmission time indicates a time when the CU reinitiates a training data request message.

26. The method of claim 20, wherein the training data update message comprises input information reporting data requested by a CU, the input information reporting data comprising measurement reporting data and a timestamp provided by at least one cell under the first DU.

27. The method of claim 26, wherein the measurement report data provided by each of the at least one cell comprises: UE characteristic input information and base station characteristic input information;

the UE feature input information includes at least one of: UE ID, UE location information, UE movement speed, and UE performance measurement information; the UE performance measurement information includes at least one of: RSRP, RSRQ, SINR, uplink and downlink throughput, packet delay and packet loss rate;

the base station characteristic input information comprises at least one of the following: UE trajectory prediction, current network energy efficiency, predicted network energy efficiency, current network radio resource status, and predicted network radio resource status.

28. The method according to any one of claims 20, 26 or 27, wherein the CU performs model training according to the training data update message sent by the DU to obtain a network energy consumption optimization model, including:

The CU receives training data from all DUs in the energy consumption optimization area range;

the CU processes and segments the received data set to obtain a training data set, a verification data set and a test data set;

and performing online model training based on the training data set, performing a test based on the test data set at the verification place based on the verification data set to adjust model parameters and optimization characteristics so as to obtain a network energy consumption optimization model.

29. The method of claim 1, wherein the model inference output comprises at least one of predictive information and decision information.

30. The method of claim 29, wherein the optimizing the network energy consumption for the UE within the network energy consumption optimization range based on the model inference output comprises:

and if the model reasoning output comprises a recommended energy-saving cell ID list, indicating each energy-saving cell indicated by the energy-saving cell list to be switched to an energy-saving state, wherein UE in the cell in the energy-saving state reasoning output a recommended candidate cell list based on the model reasoning, and switching to a cell corresponding to the bearing flow.

31. A base station, the base station comprising: a central management node CU and a separation management node DU;

The CU is used for: determining model reasoning output based on a network energy consumption optimization model, DU statistical characteristic input data and UE measurement report data, and optimizing network energy consumption of the UE in a network energy consumption optimization range based on the model reasoning output;

the CU is further configured to send a feedback request message to the DU, where the feedback request message is used to request network performance after network energy consumption optimization, and the feedback request message includes a requested measurement object;

the DU is configured to send a feedback reply message to the CU, where the feedback reply message includes an available measurement object;

the DU is further configured to send a feedback data update message to the CU, where the feedback data update message includes network performance feedback data.

32. A base station comprising a processor, a memory and a program or instruction stored on the memory and executable on the processor, which when executed by the processor implements the steps of the network energy consumption optimisation method according to any one of claims 1 to 30.