Key performance indicator (KPI) assurance for energy saving

Abstract

A method and an electronic device for KPI assurance for energy saving. A method at a network node (210) for facilitating a RAN (105) to manage its energy consumption comprises: receiving (S310, S410) one or more first target values for one or more first PIs of the RAN (105); receiving (S340, S420) one or more first current values of one or more first PIs of the RAN (105); and determining (S350, S430) one or more second target values for one or more second PIs based at least on the one or more first current values and the one or more first target values, where the one or more second target values are used for one or more operations for managing energy consumption.

Description

Technical Field

[0001] The present disclosure relates to the field of telecommunications, and in particular, to a network node, a radio access network (RAN) node, and a method for ensuring key performance indicators (KPIs) for energy saving. Background Art

[0002] With the development of electronics and telecommunications technologies, mobile devices (e.g., mobile phones, smart phones, laptop computers, tablet computers, in-vehicle devices) have become an important part of our daily lives. To support a large number of mobile devices, an efficient radio access network (RAN), such as a 5G new radio (NR) RAN, will be required.

[0003] Carriers have been concerned about energy efficiency for several years, but 5G will make this issue a top consideration as it will consume more energy than 4G. Some carriers spend on average 5% to 6% (excluding depreciation and amortization) of their operating expenses on energy costs, and this proportion is expected to increase with the transition from 4G to 5G.

[0004] A typical 5G base station consumes up to twice or more the power of a 4G base station, and due to the need for more antennas and a denser small cell layer, energy costs may increase even more at higher frequencies. Edge computing facilities required to support local processing and new Internet of Things (IoT) services will also increase the overall network power consumption.

[0005] Based on data on remote radio unit (RRU) / baseband unit (BBU) requirements per site, the power demand of a typical 5G site exceeds 11.5 kilowatts, nearly 70% more than a base station that deploys 2G, 3G, and 4G radios simultaneously. 5G macro base stations may require several new types of high-power-consuming components, including microwave or millimeter-wave transceivers, field programmable gate arrays (FPGAs), faster data converters, high-power / low-noise amplifiers, and integrated MIMO antennas.

[0006] The increasing power demand of 5G base stations may give rise to several problems:

[0007] - Insufficient AC power supply;

[0008] - Insufficient battery capacity: More backup battery capacity is needed, while traditional lead-acid batteries have a low energy density and it is difficult to expand their capacity;

[0009] - Unable to support high-power long-distance transmission: In 5G scenarios where high power needs to be supplied to remote active antenna units (AAUs), voltage drop means that the transmission distance is limited. Summary of the Invention

[0010] There are many methods for RAN energy saving. For example, machine learning (ML) / artificial intelligence (AI) can be utilized to optimize energy saving decisions by leveraging the data collected in the RAN network. However, in some cases, the ML / AI-based decision maker may make decisions that result in unacceptable network performance (which can be reflected by one or more degraded KPIs). For example, even if there are annual or quadrennial sports events (e.g., the Olympics) in a stadium, the ML / AI-based decision maker will still recommend aggressive energy saving parameters to the base stations serving the stadium because, based on historical data prediction, there should be no traffic as the decision maker is unaware of the sports event. In this case, the remaining active units / RU components will be overloaded and the network performance will deteriorate significantly.

[0011] Therefore, ML / AI can only improve the accuracy of KPI guarantee from a statistical perspective, while some KPIs need to consider almost the worst-case scenarios. Relying solely on ML / AI is not sufficient to achieve energy saving KPI guarantee.

[0012] Therefore, to solve or at least mitigate the above problems, some embodiments of the present disclosure are provided.

[0013] According to a first aspect of the present disclosure, a method for facilitating a RAN to manage its energy consumption at a network node is provided. The method includes: receiving one or more first target values of one or more first performance indicators (PIs) of the RAN; receiving one or more first current values of one or more first PIs of the RAN; and determining one or more second target values of one or more second PIs based at least on the one or more first current values and the one or more first target values, wherein the one or more second target values are for one or more operations of managing energy consumption.

[0014] In some embodiments, the one or more first PIs indicate one or more performance metrics that can be perceived by end users in the RAN. In some embodiments, the one or more first PIs indicate at least one of the following: performance metrics in terms of accessibility; performance metrics in terms of retainability; performance metrics in terms of integrity; performance metrics in terms of mobility; and performance metrics in terms of availability. In some embodiments, the one or more second PIs indicate one or more performance metrics that cannot be perceived by end users in the state of the RAN. In some embodiments, the one or more second PIs indicate at least one of the following: maximum path loss; media access control (MAC) layer scheduling delay; schedulable physical layer channel capacity; and schedulable sessions per transmission time interval (TTI).

[0015] In some embodiments, the step of receiving one or more first target values includes: receiving one or more first target values from an operator of the RAN. In some embodiments, the method further includes at least one of the following: determining whether one or more first target values are feasible; and when there are multiple first target values, determining whether the first target values conflict with each other. In some embodiments, the method further includes: providing an alert to the operator of the RAN in response to determining at least one of the following: at least one of the first target values is not feasible; and at least two of the first target values conflict with each other. In some embodiments, the method further includes: determining a baseline value for one or more first PIs, each baseline value indicating a performance metric in the case where no operations for energy saving are performed, wherein the step of determining whether one or more first target values are feasible includes at least one of the following: determining that at least one of the first target values is not feasible in response to determining that at least one of the first target values indicates a requirement higher than that indicated by at least one corresponding baseline value; and determining that all of the first target values are feasible in response to determining that each of the first target values indicates a requirement lower than or equal to that indicated by the corresponding baseline value.

[0016] In some embodiments, the step of determining whether first target values conflict with each other when there are multiple first target values includes at least one of the following: determining that the multiple first target values do not conflict with each other in response to determining that none of the multiple first target values indicates a requirement conflicting with the requirement indicated by any other of the multiple first target values; and determining that at least two of the multiple first target values conflict with each other in response to determining that at least two of the multiple first target values indicate conflicting requirements. In some embodiments, the step of receiving one or more first current values includes: monitoring current values of one or more first PIs of the RAN.

[0017] In some embodiments, the step of determining one or more second target values includes: performing a closed-loop control process to determine one or more second target values based at least on one or more first current values and one or more first target values, wherein the one or more first current values and the one or more first target values are inputs of the closed-loop control process, the one or more second target values are outputs of the closed-loop control process, and the control objective of the closed-loop control process is at least one of the following: making the one or more first current values satisfy the one or more first target values; respectively minimizing the difference between the one or more first current values and the one or more first target values; minimizing the difference between at least one of the first current values and its corresponding first target value; and making the one or more first current values satisfy the one or more first target values to the greatest extent. In some embodiments, the closed-loop control process is a deviation-based control process. In some embodiments, the second target value of the second PI is determined as the sum of the second target value of the second PI in the previous cycle and an adjustment value, wherein the adjustment value is determined as the product of a constant and the difference between the first current value of the first PI and the first target value of the first PI.

[0018] In some embodiments, the mapping between one or more first PIs and one or more second PIs is predetermined or configured, wherein the mapping indicates which first PI or which first PIs among the first PIs are involved in determining the second target value of the second PI. In some embodiments, the method further includes: triggering the RAN to perform one or more operations for managing energy consumption based at least on the one or more second target values. In some embodiments, the step of triggering the RAN to perform one or more operations includes: sending the one or more second target values to one or more RAN nodes in the RAN, so that the one or more RAN nodes can determine one or more operations to be performed based at least on the one or more second target values of the one or more second PIs and the one or more second current values.

[0019] In some embodiments, the one or more first current values and the one or more second target values are determined periodically. In some embodiments, the network node includes an Operations and Maintenance (O&M) node.

[0020] According to a second aspect of the present disclosure, there is provided a network node. The network node includes: a processor; a memory storing instructions that, when executed by the processor, cause the processor to execute any method of the first aspect.

[0021] According to a third aspect of the present disclosure, there is provided a network node for facilitating a RAN to manage its power consumption. The network node includes: a first receiving module configured to receive one or more first target values of one or more first PIs of the RAN; a second receiving module configured to receive one or more first current values of one or more first PIs of the RAN; and a determining module configured to determine one or more second target values of one or more second PIs based at least on the one or more first current values and the one or more first target values, wherein the one or more second target values are for one or more operations of managing power consumption. Further, the network node includes one or more additional modules, each module performing any step of any method of the first aspect.

[0022] According to a fourth aspect of the present disclosure, there is provided a method for managing power consumption at a RAN node. The method includes: receiving, from a network node, one or more second target values of one or more second PIs, the one or more second target values being determined based at least on one or more first target values and one or more first current values of one or more first PIs; and determining one or more operations to be performed for managing its power consumption based at least on the one or more second target values.

[0023] In some embodiments, the one or more first PIs indicate one or more performance metrics that can be perceived by an end user. In some embodiments, the one or more first PIs indicate at least one of the following: a performance metric in terms of accessibility; a performance metric in terms of retainability; a performance metric in terms of integrity; a performance metric in terms of mobility; and a performance metric in terms of availability. In some embodiments, the one or more second PIs indicate one or more performance metrics that cannot be perceived by an end user. In some embodiments, the one or more second PIs indicate at least one of the following: a maximum path loss; a MAC layer scheduling delay; a schedulable physical layer channel capacity; and a schedulable session per TTI. In some embodiments, the step of determining the one or more operations includes: determining, by a machine learning (ML) / artificial intelligence (AI) assisted decision-making module, the one or more operations based at least on one or more of the following: the one or more second target values; a measurement report; a load estimate; and an energy estimate. In some embodiments, the one or more second target values are received periodically, and the one or more operations to be performed for managing its power consumption are determined periodically. In some embodiments, the method further includes: performing the one or more operations.

[0024] According to a fifth aspect of the present disclosure, there is provided a RAN node. The RAN node includes: a processor; and a memory storing instructions that, when executed by the processor, cause the processor to perform any method of the fourth aspect.

[0025] According to a sixth aspect of the present disclosure, a RAN node for managing its energy consumption is provided. The RAN node includes: a receiving module configured to receive one or more second target values of one or more second PIs from a network node, the one or more second target values being determined based at least on one or more first target values and one or more first current values of one or more first PIs; and a determining module configured to determine one or more operations to be performed for managing its energy consumption based at least on the one or more second target values. Further, the RAN node includes one or more additional modules, each module performing any step of any method of the fourth aspect.

[0026] According to a seventh aspect of the present disclosure, a computer program including instructions is provided. When executed by at least one processor, the instructions cause the at least one processor to perform any method of the first aspect or the fourth aspect.

[0027] According to an eighth aspect of the present disclosure, a carrier contains the computer program of the seventh aspect. In some embodiments, the carrier is one of an electrical signal, an optical signal, a radio signal, or a computer-readable storage medium.

[0028] According to a ninth aspect of the present disclosure, a telecommunications network is provided. The telecommunications network includes: a network node; and a RAN including one or more RAN nodes, wherein the network node is configured to: receive one or more first target values of one or more first PIs of the RAN; receive one or more first current values of one or more first PIs of the RAN; and determine one or more second target values of one or more second PIs based at least on the one or more first current values and the one or more first target values; and send the one or more second target values to one or more RAN nodes, wherein each of the one or more RAN nodes is configured to: receive the one or more second target values from the network node; and determine one or more operations to be performed for managing the energy consumption of the RAN node based at least on the one or more second target values.

[0029] In some embodiments, the network node is the network node of the second aspect or the third aspect. In some embodiments, the one or more RAN nodes are the RAN nodes of the fifth aspect or the sixth aspect.

[0030] Through some embodiments of the present disclosure, a closed-loop controller and an ML / AI-based solution can be combined to avoid KPI degradation while still achieving energy savings. Description of the Drawings

[0031] Figure 1 is a diagram showing an exemplary system for energy saving according to an embodiment of the present disclosure.

[0032] Figure 2 FIG. is a diagram showing an exemplary system for KPI assurance for energy saving according to an embodiment of the present disclosure.

[0033] Figure 3 FIG. is a flowchart showing an exemplary method for KPI assurance for energy saving according to an embodiment of the present disclosure.

[0034] Figure 4 FIG. is a flowchart showing an exemplary method at a network node for facilitating the RAN to manage its energy consumption according to an embodiment of the present disclosure.

[0035] Figure 5 FIG. is a flowchart showing an exemplary method at a RAN node for managing its energy consumption according to an embodiment of the present disclosure.

[0036] Figure 6 FIG. schematically shows an embodiment of an arrangement that can be used in a network node or a RAN node according to an embodiment of the present disclosure.

[0037] Figure 7 FIG. is a block diagram of an exemplary network node according to an embodiment of the present disclosure.

[0038] Figure 8 FIG. is a block diagram of an exemplary RAN node according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0039] Hereinafter, the present disclosure will be described with reference to the embodiments shown in the accompanying drawings. However, it should be understood that these descriptions are provided for illustrative purposes only and do not limit the present disclosure. In addition, descriptions of known structures and technologies are omitted hereinafter to avoid unnecessarily obscuring the concept of the present disclosure.

[0040] Those skilled in the art will understand that the term "exemplary" is used herein to mean "illustrative" or "serving as an example", and is not intended to imply that a particular embodiment is superior to another embodiment or that a particular feature is essential. Similarly, unless the context clearly indicates otherwise, the terms "first", "second", "third", and "fourth" and similar terms are only used to distinguish one particular instance of an item or feature from another particular instance, and do not indicate a particular order or arrangement. In addition, as used herein, the term "step" is intended to be synonymous with "operation" or "action". Unless the context or details of the described operation clearly indicate otherwise, any description of the sequence of steps herein does not mean that these operations must be performed in a particular order, or even that these operations are performed in any order.

[0041] Unless otherwise expressly defined herein or understood from the context, conditional language used herein (e.g., "can," "should," "may," "for example," etc.) is generally intended to convey that some embodiments include some features, elements, and / or steps while other embodiments do not include the said features, elements, and / or steps. Thus, such conditional language is generally not intended to imply that the features, elements, and / or steps are necessary in any case for one or more embodiments, or that one or more embodiments must include logic circuitry to determine, with or without author input or permission, whether to include such features, elements, and / or steps in any particular embodiment or to perform the said features, elements, and / or steps in any particular embodiment. Additionally, the term "or" is used in an inclusive sense (and not an exclusive sense), such that when used, for example, to connect a list of elements, the term "or" means one, some, or all of the elements in the list. Further, in addition to having its ordinary meaning, the term "each" as used herein can mean any subset of the set of elements to which the term "each" is applied.

[0042] The term "based on" shall be construed to mean "at least partially based on". The terms "one embodiment" and "embodiment" shall be construed to mean "at least one embodiment". The term "another embodiment" shall be construed to mean "at least one other embodiment". Other definitions, which may be explicit or implicit, may be included below. Additionally, unless expressly defined otherwise herein, statements such as "at least one of X, Y, and Z" shall be understood in context to generally mean that items, terms, etc. can be X, Y, or Z or combinations thereof.

[0043] The terms used herein are for the purpose of describing particular embodiments only and are not intended to limit the example embodiments. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when used herein, the words "comprising," "having," "including" specify the presence of the stated features, elements, and / or components, etc., but do not preclude the presence or addition of one or more other features, elements, components, and / or combinations thereof. It will also be understood that unless expressly stated to the contrary, when used herein, the terms "connected," "connected with," "connected to," etc. only mean that there is an electrical connection or communication connection between two elements and they can be connected directly or indirectly.

[0044] Of course, without departing from the scope and essential characteristics of the present disclosure, the present disclosure may be implemented in other specific ways different from those described herein. One or more of the specific processes discussed below may be executed in any electronic device including one or more appropriately configured processing circuits, which in some embodiments may be embodied in one or more application specific integrated circuits (ASICs). In some embodiments, these processing circuits may include one or more microprocessors, microcontrollers, and / or digital signal processors programmed with appropriate software and / or firmware to implement one or more of the above operations and their variations. In some embodiments, these processing circuits may include custom hardware that performs one or more of the above functions. The proposed embodiments should therefore be considered illustrative rather than restrictive in all respects.

[0045] Although multiple embodiments of the present disclosure will be shown in the drawings and described in the following detailed description, it should be understood that the present disclosure is not limited to the disclosed embodiments, but is also capable of various rearrangements, modifications, and substitutions without departing from the present disclosure as set forth and defined in the appended claims.

[0046] In addition, although the following description of some embodiments of the present disclosure is given in the context of 4G LTE (Long Term Evolution), the present disclosure is not limited thereto. In fact, as long as the KPI guarantee for energy saving is involved, the inventive concept of the present disclosure can be applied to any suitable communication architecture, for example, Global System for Mobile Communications (GSM) / General Packet Radio Service (GPRS), Enhanced Data Rates for GSM Evolution (EDGE), Code Division Multiple Access (CDMA), Wideband CDMA (WCDMA), Time Division-Synchronous CDMA (TD-SCDMA), CDMA2000, Worldwide Interoperability for Microwave Access (WiMAX), Wireless Fidelity (Wi-Fi), Long Term Evolution (LTE), LTE-Advanced (LTE-A), or 5G New Radio (NR), etc. Therefore, those skilled in the art can easily understand that the terms used herein may also refer to their equivalents in any other infrastructure. For example, the term "user equipment" or "UE" used herein may refer to a terminal device, a mobile device, a mobile terminal, a mobile station, a user equipment, a user terminal, a wireless device, a wireless terminal, or any other equivalent. Another example is that the term "network node" used herein may refer to a network function, a network element, a RAN node, an OAM node, a test network function, a transmission and reception point (TRP), a base station, a base station transceiver station, an access point, a hotspot, a NodeB, an evolved NodeB (eNB), a gNB, or any other equivalent. In addition, the term "electronic device" used herein may refer to any one of the devices listed above.

[0047] In addition, the following 3rd Generation Partnership Project (3GPP) documents are incorporated herein by reference in their entirety:

[0048] - 3GPP TR 37.817 V17.0.0 (2022-04), Technical Report (TR), 3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Evolved Universal Terrestrial Radio Access (E-UTRA) and NR; Study on Enhancements for Data Collection for NR and EN-DC (Release 17).

[0049] As mentioned above, from the perspectives of cost and environment, RAN network energy consumption is one of the greatest challenges faced by the RAN industry. How to reduce energy consumption and carbon emissions while meeting the demand for the expected massive growth of data services is becoming a hot topic. For telecom providers, energy saving / efficiency will be a key competitive advantage at all levels. For example, the GSMA Intelligent Network Transformation Survey indicates that over 90% of operators prioritize energy efficiency and sustainability.

[0050] Efficient energy consumption can also be achieved by means such as load reduction, coverage modification, or other RAN configuration adjustments. The best energy-saving decisions depend on various factors, including the load conditions at different RAN nodes, the capabilities of RAN nodes, KPI / Quality of Service (QoS) requirements, the number of active User Equipments (UEs), and UE mobility, cell utilization, etc.

[0051] However, it is not easy to identify actions aimed at improving energy efficiency. Incorrectly shutting down cell / Radio Unit (RU) components for energy saving may seriously degrade network performance because the remaining active cell / RU components still need to serve the traffic. Incorrect energy-saving actions may lead to the deterioration of network KPIs, which are prohibited by operators.

[0052] To address the above problems, Machine Learning (ML) / Artificial Intelligence (AI) can be utilized to optimize energy-saving decisions by leveraging the data collected in the RAN network. In 3GPP Technical Report (TR) 37.817, V17.0.0, some recommendations for ML implementation are listed (section 5.1.2).

[0053] Specifically, ML / AI algorithms can provide:

[0054] 1. Prediction: ML / AI predicts the energy consumption and / or load status in the next cycle, which can be used to make better energy-saving decisions.

[0055] 2. Decision-making: Based on the predicted load, ML / AI can configure energy-saving strategies to maintain the balance between system performance and energy efficiency.

[0056] Figure 1FIG. is a diagram showing an exemplary system 10 for energy saving according to an embodiment of the present disclosure. As Figure 1 shown, the system 10 may include one or more RAN nodes 105, an ML / AI-assisted prediction module 110, and an ML / AI-assisted decision-making module 120. As described above and as Figure 1 shown, the ML / AI-assisted prediction module 110 may predict or estimate the load status / energy consumption in the next cycle based on the measurement reports collected from the RAN nodes 105. As described above and as Figure 1 shown, the ML / AI-assisted decision-making module 120 may make decisions on which energy-saving actions / operations the RAN nodes 105 should perform based on, for example, the predicted load / energy information from the ML / AI-assisted prediction module 110, the measurement reports collected from the RAN nodes 105, and / or the KPI requirements input by the operator 100 of the RAN nodes. Once these decisions are made, the ML / AI-assisted decision-making module 120 may trigger the RAN nodes 105 to perform corresponding actions / operations for energy saving.

[0057] There are many prediction algorithms, such as LSTM (Long / Short-Term Memory), DNN (Deep Neural Network), decision tree, etc. Two types of principles behind these algorithms are briefly introduced below.

[0058] - Causality-based prediction

[0059] Causality (which refers to the causal relationship between variables) has received increasing attention due to its effectiveness in enhancing the interpretability and robustness of predictable models.

[0060] Causality-based prediction has become more powerful because big data technology is accumulating a large amount of massive data across different fields, which makes it possible to discover the causal mechanisms in the system. Then, the discovered causal relationships enable the prediction of the system's behavior to automate its control.

[0061] An example of causality-based prediction is the load estimation of a railway cell. Based on big data, ML / AI learns that for some cells, most of the traffic comes from passengers on the train, and the train travels along a specific trajectory, such as moving from cell A to cell B and then to cell C. Therefore, if cell A has a high traffic load, based on causality-based prediction, cell C may have a high traffic load in the next few minutes.

[0062] - Correlation-based prediction

[0063] It utilizes the correlation between services and other factors, such as temporal correlation, i.e., the statistical relationship (temporal correlation) between the current service and historical services in the same cell. This method is effective in predicting the regular component of services. For example, the cell covering the office always has a higher traffic volume from Monday to Friday, but a lower traffic volume on weekends.

[0064] In addition, there is an Energy Performance Optimizer (EPO) that uses ML / AI technologies to make decisions while ensuring KPI requirements. Based on the predicted future load, measurement values, and time / date information, the EPO can use reinforcement learning to directly determine energy-saving actions. The EPO can automatically generate energy-saving parameters for the gNB. It attempts to meet the operator's KPI requirements and sets the KPI requirements as the reward in reinforcement learning.

[0065] Although ML / AI-based solutions contribute significantly to the accuracy of service prediction and decision-making and outperform traditional solutions without ML / AI, however, due to some limitations, it can only ensure good performance with a certain probability.

[0066] The limitations of different ML / AI-based solutions can include but are not limited to:

[0067] - Limitations of causality-based prediction

[0068] When there are lectures, sports games, or frequent flights, school buildings, stadiums, or airports respectively generate more data traffic. In other words, lectures, sports games, and frequent flights are the main causes of traffic surges in the above areas. In addition, the time from the start to the end of the event, popularity, date (holiday or working day), weather, and traffic congestion can also affect traffic changes. These factors are coupled with each other via complex relationships. Unfortunately, it is difficult (too costly or legally prohibited) for operators to obtain such social information, so the RAN can only obtain partial effective social information or even no effective social information. This will limit the accuracy of traffic prediction based on causal relationships.

[0069] Limitations of correlation-based prediction

[0070] In office buildings, the data traffic volume is higher on working days. In most cases, the weekly temporal correlation can predict the trend of traffic load (the traffic load on Monday morning is highly correlated with the traffic load on another Monday morning).

[0071] Considering the data collection cost, some annual events (e.g., public holidays) are difficult to be captured by ML / AI. Another example is that the Chinese New Year occurs according to the lunar calendar, so it is difficult for ML / AI to learn / predict through simple temporal correlations of weekly data or even annual data.

[0072] Limitations of ML / AI-based decision making

[0073] There are two key inputs to ML / AI-based decision making:

[0074] 1. Prediction of future load / energy performance, e.g. Figure 1 As indicated by the arrow from the ML / AI-assisted prediction module 110 to the ML / AI-assisted decision-making module 120 .

[0075] 2. Time / date, which can be, for example, Figure 1 The ML / AI-assisted decision making module 120 is shown as being obtained or otherwise determined locally.

[0076] For example, if it is midnight now and the predictor (e.g., ML / AI assisted prediction module 110) informs the ML / AI based decision maker (e.g., ML / AI assisted decision making module 120) that the future load consumption is high, the decision maker will still recommend aggressive parameters because it is midnight and based on historical data, the traffic burst in the near future will disappear quickly without affecting the KPI.

[0077] The performance limitations are obvious:

[0078] 1. Its performance depends on ML / AI-based predictions, which have limitations (e.g., those listed above).

[0079] 2. The decision making itself uses time / date as input, and this is similar to the limitation of correlation-based prediction, which is not robust enough for some social events (e.g., concerts, public holidays, etc.).

[0080] Since predictions and decision making based on correlation or causality have limitations, ML / AI can only ensure the accuracy of predictions on business load, KPI impact, etc. with a “high probability”.

[0081] Unfortunately, due to the high KPI ambitions of operators, many KPIs cannot tolerate too many "possibilities". For example, some operators may have a requirement for call drop rate: less than 1.5%, and many cells have already reached 1% to 1.2% even without energy saving. There is limited room for energy saving by reducing KPIs.

[0082] For example, if ML / AI cannot predict the heavy traffic load of an upcoming concert in 4 hours, the energy-saving solution may result in an unacceptable call drop rate. Therefore, ML / AI can only improve the accuracy of KPI guarantee from a statistical perspective, while some KPIs need to consider almost the worst case scenario. Relying solely on ML / AI is not enough to achieve energy-saving KPI guarantee.

[0083] To address the "high performance but low determinacy" issue of ML / AI, some embodiments of the present disclosure propose two key creative steps:

[0084] - Combine a traditional closed-loop controller with "low performance but high determinacy" and an ML / AI-based controller with "high performance but low determinacy" to ensure KPI requirements.

[0085] - To avoid oscillating commanders from ML / AI and the traditional closed-loop controller (i.e., two commanders for one target), some embodiments of the present disclosure use the closed-loop controller to ensure that KPI requirements are met, and use ML / AI to ensure that PI requirements are met. The closed-loop controller can be responsible for converting KPI requirements into PI requirements.

[0086] In some embodiments, a traditional closed-loop controller and an ML / AI-based solution can be combined. In some embodiments, the closed-loop controller can focus on meeting KPI requirements, and ML / AI can focus on meeting PI requirements. In some embodiments, the closed-loop controller can be responsible for converting KPI requirements into PI requirements.

[0087] Figure 2 is a diagram showing an exemplary system 20 for KPI assurance for energy saving according to an embodiment of the present disclosure. As Figure 2 shown, the system 20 may include one or more RAN nodes 105, an ML / AI-assisted prediction module 110, an ML / AI-assisted decision-making module 220, and a closed-loop controller 210. In some embodiments, the ML / AI-assisted decision-making module 220 may be part of the RAN node 105 and / or executed by the RAN node 105. In some embodiments, the closed-loop controller 210 may be part of the O&M node and / or executed by the O&M node.

[0088] As Figure 2 shown, the ML / AI-assisted prediction module 110 may predict or estimate the load state / energy consumption in the next period based on the measurement reports collected from the RAN node 105, in a manner similar to, for example, the manner shown Figure 1 shown.

[0089] As Figure 2 shown, the closed-loop controller 210 may convert one or more KPI requirements received from the operator 100 into one or more PI requirements based at least on the current KPI status collected from the RAN node 105. As Figure 2As shown, the ML / AI-assisted decision-making module 220 can make decisions on which energy-saving actions / operations the RAN node 105 should perform based on, for example, the predicted load / energy information from the ML / AI-assisted prediction module 110, the measurement reports collected from the RAN node 105, and / or the PI requirements determined by the closed-loop controller 210. Once these decisions are made, the ML / AI-assisted decision-making module 220 can trigger the RAN node 105 to perform the corresponding actions / operations for energy saving.

[0090] Thus, through some embodiments of the present disclosure, ML / AI applications (e.g., the ML / AI-assisted decision-making module 220) are enabled to achieve some strict energy-saving KPI guarantees.

[0091] Figure 3 is a flowchart showing an exemplary method 300 for KPI guarantee for energy saving according to an embodiment of the present disclosure. In some embodiments, the method 300 can be performed by Figure 2 the system 20 shown. The method 300 can include steps S310 to S360. However, the present disclosure is not limited thereto. In some other embodiments, the method 300 can include more steps, fewer steps, different steps, or any combination thereof. Additionally, when there are multiple steps, the steps of the method 300 can be performed in an order different from the order described herein. Further, in some embodiments, the steps in the method 300 can be split into multiple sub-steps and performed by different entities, and / or multiple steps in the method 300 can be combined into a single step. Next, the method 300 will be described in detail with reference to Figure 2 to.

[0092] As Figure 3 shown, the method 300 can start at step S310, where the operator 100 can input its required KPI requirements.

[0093] At step S320, the system 20 (or the closed-loop controller 210) can check the feasibility of the KPI requirements, e.g., whether it is feasible and / or whether any requirements conflict with each other. If the check fails, the system 20 (or the closed-loop controller 210) can issue an alert to the operator 100.

[0094] At step S330, the closed-loop controller 210 can initially convert the KPI requirements into PI requirements.

[0095] Execute a loop from step S340 to step S360:

[0096] - At step S340, the current KPI status can be monitored. In some embodiments, the current KPI status can be continuously monitored. In some embodiments, continuously monitoring the current KPI status can include (but is not limited to): periodically monitoring, event-based (aperiodic) monitoring, or both. In some embodiments, the monitoring can include at least one of the following: polling the RAN 105 through the closed-loop controller 210, and receiving reports from the RAN 105.

[0097] - At step S350, the closed-loop controller 210 can generate new PI requirements based on the current KPI measurements from the RAN node 105 and the KPI requirements from the operator 100:

[0098] - Input: Current KPI status and KPI requirements;

[0099] - Output: PI requirements;

[0100] - Control objective: The KPI meets the KPI requirements.

[0101] - At step S360, the ML / AI-based energy-saving solution (e.g., the ML / AI-assisted decision-making module 220) can attempt to ensure the new PI requirements from the closed-loop controller 210.

[0102] When this loop is continuously executed, a closed-loop control is formed.

[0103] In some embodiments, the term "KPI" can represent the end-user's perception of the network at a macro level. Operators can use KPI statistics to compare networks with each other and / or detect problems and errors.

[0104] For example, the KPI can include (but is not limited to) at least one of the following:

[0105] - Accessibility, including connection establishment success rate, random access rate, etc.

[0106] - Retainability, including session time normalized loss rate, etc.

[0107] - Integrity, including average UE delay, UE throughput, packet loss rate, etc.

[0108] - Mobility, including handover success rate, etc.

[0109] - Availability, including cell availability, etc.

[0110] In some embodiments, the operator may have specific requirements for the KPI. For example, the mobility KPI handover success rate > 98%. This requirement can be set based on the operator's own business considerations.

[0111] In some embodiments, energy-saving features often affect KPIs, and the tolerance thresholds (KPI requirements) should be set manually by the operator. However, manual operation means potential unreasonable inputs. The RAN should check the operator-entered KPI requirements regarding the following aspects:

[0112] - Feasibility: For example, a cell without any energy-saving enabled can only achieve a call drop rate of 2%, while the operator wants the energy-saving function to ensure a call drop rate lower than 1%. This is infeasible, and the network should notify the operator that this KPI requirement may be infeasible. In some embodiments, to implement the feasibility check of KPI requirements, the energy-saving function should record the KPI performance in history when the energy-saving function is "off".

[0113] - Conflict-free: The requirements of KPI A should not conflict with the requirements of KPI B. For example, the integrity (e.g., latency) requirements for high-QoS UEs are worse than those for low-QoS UEs.

[0114] In some embodiments, the term "PI" can represent system-level information that explains the KPI results. In some embodiments, PI can be information that cannot be perceived by end users. In some embodiments, PI can be information that can only be perceived by the RAN and / or its operator. Many PIs can be used for root cause analysis. PI can also be in the form of a metric that shows what a specific part of the system can perform. PI usually has an impact on KPIs.

[0115] The PIs used for energy-saving ML / AI decision-making can include (but are not limited to) at least one of the following items:

[0116] - The maximum path loss, which indicates the maximum path loss that supports the KPI (including accessibility, retainability, and mobility) requirements.

[0117] - The MAC layer scheduling latency: Indicates the maximum MAC layer scheduling latency that supports the integrity (latency) requirements.

[0118] - The schedulable physical layer channel capacity.

[0119] - The schedulable sessions per TTI.

[0120] Return reference Figure 2 , there are many solutions for the implementation of the closed-loop controller 210. One of them will be described here, namely the solution based on the deviation controller:

[0121]

[0122] where, is a constant that indicates the sensitivity of the PI to the difference between the KPI and the KPI requirement.

[0123] Next, a specific example of the deviation-based controller process at the closed-loop controller 210 will be described in detail. Assume an initial maximum path loss threshold = -120 dB (i.e., the PI requirement), D is 100, and the KPI requirement is that the call drop rate <= 2%.

[0124] 1 (e.g., step S340). Through the ML / AI decision-making 220, the current KPI status is: the call drop rate = 3%.

[0125] 2 (e.g., step S350). The closed-loop controller 210 will adjust the PI requirement according to the updated (or latest) KPI status. The new maximum path loss threshold = 100 * (3% - 2%) + (-120) = -119 dB.

[0126] 3 (e.g., step S360). The ML / AI 220 will make a decision based on the following condition: the maximum path loss threshold = -119 dB. The ML / AI decision-making 220 will be more conservative in terms of energy conservation.

[0127] 4 (e.g., step S340 in a new cycle). The KPI is improved a little. Assume the current KPI status is: the call drop rate = 2.1%.

[0128] 5 (e.g., step S350 in a new cycle). The RAN 105 updates the new KPI status, and then the closed-loop controller 210 will update the maximum path loss threshold = 100 * (2.1% - 2%) + (-119) = -118.9 dB;

[0129] 6 (e.g., step S360 in a new cycle). The ML / AI 220 will make a decision based on the following condition: the maximum path loss threshold = -118.9 dB.

[0130] 7.... Loop as in 1 to 6.

[0131] Generally speaking, the closed-loop controller 210 can act as a safety belt for the ML / AI 220 to ensure that the KPI can be well controlled to meet the requirements even if the ML / AI makes mistakes occasionally.

[0132] Although some embodiments with only a single PI and a single KPI are described above, the present disclosure is not limited thereto. In some embodiments, a single PI can be determined based on multiple KPIs. In some embodiments, multiple PIs can be determined based on a single KPI. In some embodiments, multiple PIs can be determined based on multiple KPIs. In other words, in different embodiments, or even in different cycles, there can be a one-to-one, one-to-many, many-to-one, or many-to-many mapping between the KPI and the PI.

[0133] Figure 4 It is a flowchart of an exemplary method 400 for a network node to facilitate the RAN in managing its energy consumption according to an embodiment of the present disclosure. The method 400 may be executed at a network node (e.g., the closed-loop controller 210). The method 400 may include steps S410, S420, and S430. However, the present disclosure is not limited thereto. In some other embodiments, the method 400 may include more steps, fewer steps, different steps, or any combination thereof. Additionally, when there are multiple steps, the steps of the method 400 may be executed in an order different from the order described herein. Further, in some embodiments, the steps in the method 400 may be split into multiple sub-steps and executed by different entities, and / or multiple steps in the method 400 may be combined into a single step.

[0134] The method 400 may start at step S410, where one or more first target values of one or more first PIs of the RAN may be received.

[0135] At step S420, one or more first current values of one or more first PIs of the RAN may be received.

[0136] At step S430, one or more second target values of one or more second PIs may be determined based at least on one or more first current values and one or more first target values, where the one or more second target values may be used for one or more operations of managing energy consumption.

[0137] In some embodiments, one or more first PIs may indicate one or more performance metrics in the RAN that can be perceived by end-users. In some embodiments, one or more first PIs may indicate at least one of the following: performance metrics in terms of accessibility; performance metrics in terms of retainability; performance metrics in terms of integrity; performance metrics in terms of mobility; and performance metrics in terms of availability. In some embodiments, one or more second PIs may indicate one or more performance metrics in the state of the RAN that cannot be perceived by end-users. In some embodiments, one or more second PIs may indicate at least one of the following: maximum path loss; MAC layer scheduling delay; schedulable physical layer channel capacity; and schedulable sessions per TTI.

[0138] In some embodiments, the step of receiving one or more first target values may include: receiving one or more first target values from an operator of the RAN. In some embodiments, method 400 may further include at least one of the following: determining whether one or more first target values are feasible; and when there are multiple first target values, determining whether the first target values conflict with each other. In some embodiments, method 400 may further include: providing an alert to the operator of the RAN in response to determining at least one of the following: at least one of the first target values is not feasible; and at least two of the first target values conflict with each other. In some embodiments, method 400 may further include: determining a reference value for one or more first PIs, each reference value indicating a performance metric in a case where no operation for energy saving is performed, wherein the step of determining whether one or more first target values are feasible may include at least one of the following: determining that at least one of the first target values is not feasible in response to determining that at least one of the first target values indicates a requirement higher than the requirement indicated by at least one corresponding reference value; and determining that all of the first target values are feasible in response to determining that each of the first target values indicates a requirement lower than or equal to the requirement indicated by the corresponding reference value.

[0139] In some embodiments, the step of determining whether first target values conflict with each other when there are multiple first target values may include at least one of the following: determining that the multiple first target values do not conflict with each other in response to determining that none of the first target values among the multiple first target values indicates a requirement conflicting with the requirement indicated by any other first target value among the multiple first target values; and determining that at least two of the first target values conflict with each other in response to determining that at least two of the first target values among the multiple first target values indicate conflicting requirements. In some embodiments, the step of receiving one or more first current values may include: monitoring current values of one or more first PIs of the RAN.

[0140] In some embodiments, the step of determining one or more second target values may include: performing a closed-loop control process to determine one or more second target values based at least on one or more first current values and one or more first target values, wherein the one or more first current values and the one or more first target values may be inputs of the closed-loop control process, the one or more second target values may be outputs of the closed-loop control process, and the control objective of the closed-loop control process may be at least one of the following: making the one or more first current values satisfy the one or more first target values; respectively minimizing the difference between the one or more first current values and the one or more first target values; minimizing the difference between at least one of the first current values in the first current values and its corresponding first target value; and making the one or more first current values satisfy the one or more first target values to the greatest extent. In some embodiments, the closed-loop control process may be a deviation-based control process. In some embodiments, the second target value of the second PI may be determined as the sum of the second target value of the second PI in the previous cycle and an adjustment value, wherein the adjustment value may be determined as the product of a constant and the difference between the first current value of the first PI and the first target value of the first PI.

[0141] In some embodiments, the mapping between one or more first PIs and one or more second PIs may be predetermined or configured, wherein the mapping may indicate which first PI or which first PIs in the first PIs are involved in determining the second target value of the second PI. In some embodiments, method 400 may further include: triggering the RAN to perform one or more operations for managing energy consumption based at least on one or more second target values. In some embodiments, the step of triggering the RAN to perform one or more operations may include: sending one or more second target values to one or more RAN nodes in the RAN, so that the one or more RAN nodes can determine one or more operations to be performed based at least on the one or more second target values of the one or more second PIs and one or more second current values.

[0142] In some embodiments, one or more first current values and one or more second target values may be determined periodically. In some embodiments, the network node may include an Operations and Maintenance (O&M) node.

[0143] Figure 5It is a flowchart of an exemplary method 500 for managing its energy consumption at a RAN node according to an embodiment of the present disclosure. Method 500 may be executed at a RAN node (e.g., the ML / AI assisted decision-making module 220). Method 500 may include steps S510 and S520. However, the present disclosure is not limited thereto. In some other embodiments, method 500 may include more steps, fewer steps, different steps, or any combination thereof. Additionally, when there are multiple steps, the steps of method 500 may be executed in an order different from the order described herein. Further, in some embodiments, the steps in method 500 may be split into multiple sub-steps and executed by different entities, and / or multiple steps in method 500 may be combined into a single step.

[0144] Method 500 may start at step S510, where one or more second target values of one or more second PIs may be received from a network node, and the one or more second target values are determined based at least on one or more first target values and one or more first current values of one or more first PIs.

[0145] At step S520, one or more operations for managing its energy consumption to be executed may be determined based at least on the one or more second target values.

[0146] In some embodiments, one or more first PIs may indicate one or more performance metrics that can be perceived by an end user. In some embodiments, one or more first PIs may indicate at least one of the following: performance metrics in terms of accessibility; performance metrics in terms of retainability; performance metrics in terms of integrity; performance metrics in terms of mobility; and performance metrics in terms of availability. In some embodiments, one or more second PIs may indicate one or more performance metrics that cannot be perceived by an end user. In some embodiments, one or more second PIs may indicate at least one of the following: maximum path loss; MAC layer scheduling delay; schedulable physical layer channel capacity; and schedulable sessions per TTI. In some embodiments, the step of determining one or more operations may include: the ML / AI assisted decision-making module determining the one or more operations based on at least one of the following: the one or more second target values; measurement reports; load estimates; and energy estimates. In some embodiments, the one or more second target values may be received periodically, and the one or more operations for managing its energy consumption to be executed may be determined periodically. In some embodiments, method 500 may further include: executing the one or more operations.

[0147] Figure 6An embodiment of an arrangement 600 that can be used in a network node (e.g., a closed-loop controller 210) or a RAN node (e.g., an ML / AI assisted decision-making module 220) according to an embodiment of the present disclosure is schematically illustrated. The arrangement 600 includes a processing unit 606, e.g., having a digital signal processor (DSP) or a central processing unit (CPU). The processing unit 606 can be a single unit or multiple units for performing different actions of the processes described herein. The arrangement 600 can also include an input unit 602 for receiving signals from other entities, and an output unit 604 for providing signals to other entities. The input unit 602 and the output unit 604 can be arranged as an integrated entity or independent entities.

[0148] In addition, the arrangement 600 can include at least one computer program product 608 in the form of a non-volatile or volatile memory, such as an electrically erasable programmable read-only memory (EEPROM), a flash memory, and / or a hard disk drive. The computer program product 608 includes a computer program 610, and the computer program 610 includes code / computer-readable instructions that, when executed by the processing unit 606 in the arrangement 600, cause the arrangement 600 and / or the network node and / or the RAN node included therein to perform actions of the processes described, for example, previously in connection with Figures 2 to 5 or any other variant.

[0149] The computer program 610 can be configured as computer program code structured in computer program modules 610A, 610B, and 610C. Thus, in an exemplary embodiment, when the arrangement 600 is used in a network node for facilitating the RAN to manage its energy consumption, the code in the computer program of the arrangement 600 includes: module 610A, configured to receive one or more first target values of one or more first PIs of the RAN; module 610B, configured to receive one or more first current values of one or more first PIs of the RAN; and module 610C, configured to determine one or more second target values of one or more second PIs based at least on the one or more first current values and the one or more first target values, wherein the one or more second target values are for one or more operations of managing energy consumption.

[0150] Alternatively or additionally, the computer program 610 can also be configured as computer program code embodied in computer program modules 610D and 610E. Thus, in an exemplary embodiment of arrangement 600 used in a RAN node for managing its energy consumption, the code in the computer program of arrangement 600 includes: module 610D, configured to receive one or more second target values of one or more second PIs from a network node, the one or more second target values being determined based at least on one or more first target values and one or more first current values of one or more first PIs; and module 610E, configured to determine one or more operations to be performed for managing its energy consumption based at least on the one or more second target values.

[0151] The computer program modules can essentially perform Figures 3 to 5 the actions of the flow shown to simulate a network node and / or a RAN node. In other words, when different computer program modules are executed in processing unit 606, these computer program modules can correspond to different modules in a network node and / or a RAN node.

[0152] Although the code means in the embodiments disclosed above Figure 6 are implemented as computer program modules which, when executed in a processing unit, cause the apparatus to perform the actions described above in connection with the above-described figures, in alternative embodiments at least one code means can be implemented at least in part as a hardware circuit.

[0153] The processor can be a single CPU (Central Processing Unit), but can also include two or more than two processing units. For example, the processor can include a general microprocessor; an instruction set processor and / or associated chipset and / or a dedicated microprocessor, such as an Application Specific Integrated Circuit (ASIC). The processor can also include on-board memory for caching purposes. The computer program can be carried by a computer program product connected to the processor. The computer program product can include a computer-readable medium on which the computer program is stored. For example, the computer program product can be a flash memory, a Random Access Memory (RAM), a Read Only Memory (ROM) or an EEPROM, and the above-described computer program modules can be distributed on different computer program products in the form of memories within a network node and / or a RAN node in alternative embodiments.

[0154] Corresponding to the above method 400, a network node for facilitating a RAN to manage its energy consumption is provided. Figure 7 is a block diagram of an exemplary network node 700 according to an embodiment of the present disclosure. In some embodiments, the network node 700 can be, for example, a closed-loop controller 210.

[0155] The network node 700 can be configured to perform as described above in connection withFigure 4 The method 400 described. As Figure 7 shown, the network node 700 may include: a first receiving module 710 configured to receive one or more first target values of one or more first PIs of the RAN; a second receiving module 720 configured to receive one or more first current values of one or more first PIs of the RAN; a determining module 730 configured to determine one or more second target values of one or more second PIs based at least on the one or more first current values and the one or more first target values, wherein the one or more second target values are used for one or more operations of managing energy consumption.

[0156] The above modules 710, 720, and / or 730 may be implemented, for example, as a pure hardware solution or a combination of software and hardware by one or more of the following: a processor or microprocessor and appropriate software and a memory for storing the software, a programmable logic device (PLD), or other electronic components or processing circuitry configured to perform the above-described and, for example, the actions shown in Figure 4 In addition, the network node 700 may include one or more additional modules, and each module may perform any step of the method 400 described with reference to Figure 4 Corresponding to the above method 500, a RAN node for managing its energy consumption is provided.

[0157] is a block diagram of an exemplary RAN node 800 according to an embodiment of the present disclosure. In some embodiments, the RAN node 800 may be, for example, the ML / AI-assisted decision-making module 220. Figure 8

[0158] Figure 5 The RAN node 800 may be configured to perform the method 500 described above in connection with Figure 8 As shown, the RAN node 800 may include: a receiving module 810 configured to receive one or more second target values of one or more second PIs from the network node, the one or more second target values being determined based at least on the one or more first target values and the one or more first current values of the one or more first PIs; and a determining module 820 configured to determine one or more operations to be performed for managing its energy consumption based at least on the one or more second target values.

[0159] Figure 5 The above modules 810 and / or 820 may be implemented, for example, as a pure hardware solution or implemented as a combination of software and hardware by one or more of the following: configured to perform the above-described and, for example, in Figure 5A processor or microprocessor for the actions shown, appropriate software, and a memory, PLD, or other electronic component or processing circuit for storing the software. Additionally, the RAN node 800 may include one or more additional modules, each of which may perform any of the steps of the method 500 described with reference to Figure 5 as described.

[0160] The present disclosure has been described above with reference to embodiments of the present disclosure. However, these embodiments are for illustrative purposes only and are not intended to limit the present disclosure. The scope of the present disclosure is defined by the appended claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art can make various changes and modifications, all of which fall within the scope of the present disclosure.

Claims

1. A method (400) at a network node (210) for facilitating a radio access network RAN (105) to manage its energy consumption, the method (400) comprises: receiving (S310, S410) one or more first target values of one or more first performance indicators PI of the RAN (105); receiving (S340, S420) one or more first current values of one or more first PI of the RAN (105); and determining (S350, S430) one or more second target values of one or more second PI at least based on the one or more first current values and the one or more first target values, wherein the one or more second target values are for one or more operations of managing energy consumption.

2. The method (400) according to claim 1, wherein, the one or more first PI indicate one or more performance metrics of the RAN (105) that can be perceived by end users.

3. The method (400) according to claim 1 or 2, wherein, the one or more first PI indicate at least one of the following: - a performance metric in terms of accessibility; - a performance metric in terms of retainability; - a performance metric in terms of integrity; - a performance metric in terms of mobility; and - a performance metric in terms of availability.

4. The method (400) according to any one of claims 1 to 3, wherein, the one or more second PI indicate one or more performance metrics of the state of the RAN (105) that cannot be perceived by end users.

5. The method (400) according to any one of claims 1 to 4, wherein the one or more second PI indicate at least one of the following: - maximum path loss; - media access control MAC layer scheduling delay; - schedulable physical layer channel capacity; and - schedulable sessions per transmission time interval TTI.

6. The method (400) according to any one of claims 1 to 5, wherein, the step of receiving (S310, S410) the one or more first target values comprises: receiving (S310) the one or more first target values from an operator (100) of the RAN (105).

7. The method (400) according to any one of claims 1 to 6 further comprises at least one of the following: determining (S320) whether the one or more first target values are feasible; and when there are multiple first target values, determining (S320) whether the first target values conflict with each other.

8. The method (400) according to claim 7 further comprises: providing an alert to an operator (100) of the RAN (105) in response to determining at least one of the following: - at least one of the first target values is not feasible; and - at least two of the first target values conflict with each other.

9. The method (400) according to any one of claims 1 to 8 further comprises: Determine a reference value for each of the one or more first PIs, each reference value indicating a performance metric in the absence of performing any operations for energy saving. Wherein, the step of determining (S320) whether the one or more first target values are feasible includes at least one of the following items: In response to determining that at least one of the first target values indicates a requirement higher than the requirement indicated by at least one corresponding reference value, determine that the at least one first target value is not feasible; and In response to determining that each of the first target values indicates a requirement lower than or equal to the requirement indicated by the corresponding reference value, determine that all of the first target values are feasible.

10. The method (400) according to any one of claims 1 to 9, Wherein, The step of determining (S320) whether the first target values conflict with each other when there are multiple first target values includes at least one of the following items: In response to determining that none of the multiple first target values indicates a requirement conflicting with the requirement indicated by any other of the multiple first target values, determine that the multiple first target values do not conflict with each other; And In response to determining that at least two of the multiple first target values indicate conflicting requirements, determine that the at least two first target values conflict with each other.

11. The method (400) according to any one of claims 1 to 10, Wherein, The step of receiving (S340, S420) the one or more first current values includes: Monitoring (S340) the current values of the one or more first PIs of the RAN (105).

12. The method (400) according to any one of claims 1 to 11, Wherein, The step of determining (S350, S430) the one or more second target values includes: Performing (S350) a closed-loop control process to determine the one or more second target values based at least on the one or more first current values and the one or more first target values, Wherein, the one or more first current values and the one or more first target values are inputs of the closed-loop control process, the one or more second target values are outputs of the closed-loop control process, and the control objectives of the closed-loop control process are at least one of the following items: - Making the one or more first current values satisfy the one or more first target values; - Respectively minimizing the difference between the one or more first current values and the one or more first target values; - Minimizing the difference between at least one of the first current values and its corresponding first target value; and - Making the one or more first current values satisfy the one or more first target values to the greatest extent.

13. The method (400) according to claim 12, Wherein, The closed-loop control process is a control process based on deviation.

14. The method (400) according to claim 12 or 13, Wherein, The second target value of the second PI is determined as the sum of the second target value of the second PI in the previous cycle and the adjustment value. Wherein, the adjustment value is determined as the product of a constant and the difference between the first current value of the first PI and the first target value of the first PI.

15. The method (400) according to any one of claims 1 to 14, wherein, the mapping between the one or more first PIs and the one or more second PIs is predetermined or configured, wherein the mapping indicates which first PI or which first PIs among the first PIs are involved in determining the second target value of the second PI.

16. The method (400) according to any one of claims 1 to 15, further comprising: triggering (S360) the RAN (105) to perform one or more operations for managing energy consumption at least based on the one or more second target values.

17. The method (400) according to claim 16, wherein, the step of triggering (S360) the RAN (105) to perform one or more operations includes: sending the one or more second target values to one or more RAN nodes in the RAN (105) so that the one or more RAN nodes can determine one or more operations to be performed at least based on the one or more second target values and the one or more second current values of the one or more second PIs.

18. The method (400) according to any one of claims 1 to 17, wherein, the one or more first current values and the one or more second target values are determined periodically.

19. The method (400) according to any one of claims 1 to 18, wherein, the network node (210) includes an operation and maintenance (O&M) node.

20. A network node (210, 600, 700), comprising: a processor (606); a memory (608) storing instructions which, when executed by the processor (606), cause the processor (606) to execute the method (400) according to any one of claims 1 to 19.

21. A method (500) for managing its energy consumption at a radio access network RAN node (220), the method (500) comprising: receiving (S510) from a network node (210) one or more second target values of one or more second performance indicators PIs, the one or more second target values being determined at least based on one or more first target values and one or more first current values of one or more first PIs; and determining (S520) one or more operations to be performed for managing its energy consumption at least based on the one or more second target values.

22. The method (500) according to claim 21, wherein, the one or more first PIs indicate one or more performance metrics that can be perceived by an end user.

23. The method (500) according to claim 21 or 22, wherein, the one or more first PIs indicate at least one of the following: - performance metrics in terms of accessibility; - performance metrics in terms of retainability; - performance metrics in terms of integrity; - performance metrics in terms of mobility; and - Performance metrics in terms of usability.

24. The method (500) according to any one of claims 21 to 23, wherein, one or more second PIs indicate one or more performance metrics that cannot be perceived by an end user.

25. The method (500) according to any one of claims 21 to 24, wherein one or more second PIs indicate at least one of the following: - Maximum path loss; - Medium Access Control (MAC) layer scheduling delay; - Schedulable physical layer channel capacity; and - Schedulable sessions per Transmission Time Interval (TTI).

26. The method (500) according to any one of claims 21 to 25, wherein, the step of determining (S520) one or more operations includes: determining the one or more operations by a machine learning (ML) / artificial intelligence (AI) assisted decision-making module based on at least one of the following: - The one or more second target values; - Measurement reports; - Load estimation; and - Energy estimation.

27. The method (500) according to any one of claims 21 to 26, wherein, the one or more second target values are received periodically, and the one or more operations to be performed for managing its energy consumption are determined periodically.

28. The method (500) according to any one of claims 21 to 27, further comprises: performing the one or more operations.

29. A Radio Access Network (RAN) node (220, 600, 800), comprising: a processor (606); a memory (608) storing instructions which, when executed by the processor (606), cause the processor (606) to perform the method (500) according to any one of claims 21 to 28.

30. A computer program (610) comprising instructions which, when executed by at least one processor (606), cause the at least one processor (606) to perform the method (400, 500) according to any one of claims 1 to 19 and 21 to 28.

31. A carrier (608) containing the computer program (610) according to claim 30, wherein, the carrier (608) is one of an electrical signal, an optical signal, a radio signal, or a computer-readable storage medium.

32. A telecommunication network (20), comprising: a network node (210); and a Radio Access Network (RAN) (105) comprising one or more RAN nodes (220), wherein the network node (210) is configured to: receive one or more first target values of one or more first Performance Indicators (PIs) of the RAN (105); receive one or more first current values of one or more first PIs of the RAN (105); and determine one or more second target values of one or more second PIs based at least on the one or more first current values and the one or more first target values; and send the one or more second target values to the one or more RAN nodes (220), Each of the one or more RAN nodes (220) is configured to: Receive the one or more second target values from the network node (210); Determine one or more operations to be performed for managing the energy consumption of the RAN node (220) based at least on the one or more second target values.

33. The telecommunications network (20) according to claim 32, wherein, The network node (210) is the network node according to claim 20.

34. The telecommunications network (20) according to claim 32 or 33, wherein, The one or more RAN nodes (220) are the RAN nodes according to claim 29.

Images