Technique for configuring network energy savings

Abstract

A technique for configuring a network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, is described. As to a method aspect of the technique, the network node (100; 1000; 1212) connected to a power grid network (500) obtains (304) one or more target values (910) of one or more performance indicators, PIs (901), of the network node (100; 1000; 1212). The one or more PI target values (910) are dependent on at least one parameter (802) of the power grid network (500). The network node (100; 1000; 1212) configures (310) at least one NES feature of the network node (100; 1000; 1212) based on the one or more obtained (304) PI target values (910).

Description

[0001] Technique for configuring network energy savings

[0002] Technical Field

[0003] The present disclosure relates to a technique for configuring a radio access network node for network energy savings. More specifically, and without limitation, a network entity configuring a network node and a network node being configured as well as corresponding methods are provided.

[0004] Background

[0005] Power consumption makes up the largest fraction of the network operational excellence (OpEx) of a cellular network, whereof the power consumption of the radio access network nodes (e.g., radio base stations such as a next generation Node B or gNB) may be assessed to make up 50 % to 60 %. Without improvements in network equipment efficiency, power consumption could increase even further in the future as radio access technology evolves, e.g. towards distributed MIMO (D-MIMO).

[0006] The power consumption of the network node has a significant impact on the power grid network, especially in view of the more volatile availability of electrical power from renewable power sources. Also on a national scale, cellular networks consume a non-negligible fraction of the total produced electricity (e.g. on the order of 1-2%) and its network energy saving (NES) policies may affect the stability of the power grid, e.g. its instantaneous frequency.

[0007] This concern highlights an often overlooked responsibility of cellular networks to harmonize their energy consumption with the delicate balance of a renewable-powered grid, ensuring that their pursuit of technological advancement does not come at the cost of energy availability and stability.

[0008] Summary

[0009] Accordingly, there is a need for a technique that aligns network energy savings of a radio access networks with the fluctuating availability of renewable energy without or unnoticeably compromising on performance.

[0010] As to a first method aspect, a method of configuring a network node of a radio access network (RAN) for network energy savings (NESs) is provided. The method is performed by the network node connected to a power grid network. The method comprises obtaining one or more target values of one or more performance indicators (Pls) of the network node. The one or more PI target values are dependent on at least one parameter of the power grid network. The method further comprises configuring at least one NES feature of the network node based on the one or more obtained PI target values.

[0011] The one or more performance indicators (Pls) may be key performance indicators (KPIs) of the RAN or the network node of the RAN. The one or more target values of the one or more Pls are briefly referred to as the one or more PI target values (or symbolically by KPI t). A technically achievable value of any of the one or more Pls that is achievable by the network node may be referred to as a PI capability value of the network node. The network node may report its one or more PI capability values to a network entity from which the one or more PI target values are obtained. Alternatively or in addition, a service-guaranteed or service-required value of any of the one or more Pls may be referred to as PI requirement value. The PI requirement value may be guaranteed by RAN, e.g. for a specific one of the radio devices (e.g., UEs) or any radio device served by the RAN. For example, the PI requirement value may be required by a service (e.g., an application) associated with the RAN and / or the served radio device. A current value (e.g., an actual and / or measured value) of any of the one or more Pls is briefly referred to as PI current value. Moreover, the at least one parameter of the power grid network is briefly referred to as a power grid parameter.

[0012] By configuring the at least one NES feature of the network node based on the one or more obtained PI target values, embodiments of the method can control an operation of the RAN (particularly, an operation of the network node of the RAN) in a one-dimensional or multidimensional PI space spanned by the one or more Pls. The PI space may comprise the one or more Pls for which PI target values have been obtained and / or the one or more Pls affected by the at least one configured NES feature.

[0013] By configuring the at least one NES feature, the operation of the network node is controlled by the one or more obtained PI target values, which in turn are dependent on the at least one power grid parameter. Hence, embodiments of the method can control a power consumption of the network node responsive to the at least one power grid parameter.

[0014] The at least one power grid parameter may be obtained from another RAN node or network entity, which may be internal or external to the RAN, e.g., an operations, administration and management / maintenance (0AM) function of a telecommunication network comprising the RAN or from an advanced metering infrastructure (AMI) or an energy management system (EMS) of the power grid network.

[0015] A performance increase indicated by the one or more PI target values may require a reduction (e.g., deactivation) of the at least one NES feature and / or may correlate with (e.g., correspond to) an increase in the power consumed by the network node. Hence, embodiments can increase the power consumption of the network node responsive to an increased availability of power as indicated by the at least one power grid parameter. Alternatively or in addition, the one or more PI target values (e.g., a performance decrease indicated by the one or more PI target values) may decrease (e.g., limit) one or more PI current values. For example, based on the performance decrease indicated by the one or more PI target values, the configuring of the at least one NES feature may encompass activating or expanding the at least one NES feature, which entails a reduction in the power taken up by the network node from the power grid network. Hence, embodiments can decrease the power consumption of the network node responsive to a decreased availability of power as indicated by the at least one power grid parameter.

[0016] In any embodiment, when the at least one NES feature is deactivated (e.g., not configured), a baseline mode of operation may achieve one or more PI current values as high-performing as technically possible for the RAN or the network node (i.e., PI capability values), optionally as a best effort mode of operation in the sense that there are no (or low) guaranteed values of the one or more Pls (i.e., PI requirement values) but high average values of the one or more Pls). Alternatively or in addition, when the at least one NES feature is deactivated, the one or more PI current values of the network node may be closer to the one or more PI capacity values than to the corresponding one or more PI requirement values. In contrast, the one or more PI target values may set or imply an upper bound (in terms of performance) on the respective one or more PI current values achieved by the operation under the configured NES.

[0017] The one or more PI target values may be obtained internally (e.g., determined) in the network node or obtained externally (e.g., received) from another node of the RAN or outside of the RAN.

[0018] Configuring the at least one NES feature of the network node may encompass configuring the network node to apply the at least one NES feature. Configuring the at least one NES feature of the network node may encompass performing or applying or setting the at least one NES feature at the network node. Alternatively or in addition, configuring the at least one NES feature of the network node may encompass configuring a distributed unit (DU) of the network node (e.g., from a central unit, CU, of the network node), optionally via an Fl interface.

[0019] The network node may be a network node of the RAN. The network node may a base station or a cell or a beam or a radio unit (RU), e.g. an RU of distributed multiple input-multiple output (D-MIMO) network. The RAN and / or the network node may provide a variety of services (e.g., radio access) to at least one radio device, e.g. a user equipment (UE).

[0020] Embodiments of the method may autonomously obtain PI current values and / or the PI target values of the one or more Pls (e.g., by measuring the one or more Pls and / or the at least one power grid parameter for determining the PI target values) and / or may autonomously perform the configuration of the at least one NES feature.

[0021] Different services of the RAN may be associated with different Pls (e.g. different numbers of Pls and / or different types of Pls). The fulfilment of a service may correspond to the fulfilment of all of the lower performance limits of all Pls of the service. As the number of different Quality of Service (QoS) requirements and / or services of the RAN increases, the number of the one or more Pls may increase, for which PI target values are obtained.

[0022] The PI space for the NES may be growing as the RAN (e.g., the network node) becomes more powerful (e.g., to serve burst-like services) and / or may become more complex (e.g., multidimensional) as the number of services and associated Pls increases. Embodiments performing the method in a distributed manner can handle this growing complexity.

[0023] In an embodiment, the one or more obtained PI target values may set a lower performance limit on the one or more Pls of the network node. For example, the at least one NES feature of the network node may be configured to operate closer to each of the one or more PI target values than a corresponding PI capability value of the network node. The at least one NES feature of the network node may be configured to operate within a margin of the lower performance limit. The lower performance limit also referred to as lower limit or lower threshold value.

[0024] In one variant of any embodiment, the network node may operate according to the at least one configured NES feature to meet the one or more PI target values (in the PI space), i.e. to operate in a range between the PI requirement value (e.g. as a lower performance bound) and the PI capability value for each of the one of the one or more Pls. By operating closer to the PI requirement value than to the PI capability value, the configuring of the at least one NES feature can reduce a power consumption of the RAN (particularly, a power consumption of the network node of the RAN) responsive to the at least one power grid parameter.

[0025] For example, the one or more obtained PI target values may be formally a lower limit for the one or more Pls, while the configuration of the at least one NES feature itself may ensure that the one or more PI current values are not significantly exceeded, i.e., the network node may end up operating at one or more PI current values slightly above the corresponding one or more obtained PI target values.

[0026] At least one or each of the one or more Pls may be associated with a PI requirement value that is indicative of a minimum performance (e.g., a lower bound in terms of performance) or that defines a minimum service (e.g., a lower bound in terms of functionality) provided by the network node to radio devices (e.g., UEs). For example, the PI target value may be greater than or equal to the PI requirement value or may not fall below the PI requirement value.

[0027] While the technique has been described primarily assuming that a greater value of a PI corresponds to better service (e.g., better performance or more functionality of the RAN), each disclosure applies as well to a PI were a greater value means worse service, such as latency as an example of the one or more Pls. For example, an inverse value of such a PI may be considered, or the PI requirement value of such a PI may be indicative of a maximum value that defines the minimum performance or minimum service provided to the radio devices by the network node. By way of example, for some services, the latency as one example of the Pls may not exceed a maximum value as the PI requirement value.

[0028] In an embodiment, the one or more obtained PI target values may set an upper performance limit on the one or more Pls of the network node. Alternatively or in addition, the one or more PI target values may be greater than one or more corresponding PI requirement values of a Quality of Service (QoS) requirement.

[0029] The PI current value resulting from the configured NES feature may be closer to the PI target value than to the PI requirement value. This may allow for a dynamic adaptation of the power consumption of the network node responsive to the at least one power grid parameter, e.g. as opposed to an always-on NES configuration that operates closer to the PI requirement value.

[0030] Herein, the expressions bound and limit may be interchanged. The one or more PI target values may define an upper performance limit (also: upper limit or upper threshold value), for example an upper numerical bound (e.g., if a greater value of the PI means better service, such as transmit power as one of the Pls) or a lower numerical bound (e.g., if a greater value of the PI means lesser service, such as latency as one of the Pls).

[0031] The afore-mentioned upper limit and / or lower limit may be a numerical bound strictly applied by the network node according to the configured at least one NES feature. Alternatively or in addition, the numerical bound may be statistically fulfilled (e.g., as a probability requirement for each of the Pls) by the network node configured with the at least one NES feature. For example, the numerical bound may be a quantile of the one or more Pls. That is, the PI current value of one of the Pls may fulfill the numerical bound of the respective PI with a certain probability. For example, the numerical bound may be a 95%-quantile (i.e., the 95th percentile) of the respective Pls.

[0032] Obtaining the one or more PI target values may correspond to setting or increasing or decreasing one or more upper limit values (in terms of performance) on the one or more Pls. As opposed to a Quality of Service (QoS) requirement, which defines lower limit values (in terms of performance, e.g., guaranteed values) of the one or more Pls, the one or more PI target values may correspond to one or more upper limits on the one or more Pls. Method embodiments limiting the one or more Pls according to the obtained one or more Pls can dynamically reduce or increase the power consumption of the network node from the power grid network responsive to the at least one power grid parameter.

[0033] The power grid network (or shortly: grid) may be an electrical grid and / or an interconnected network for electricity delivery from one or more power plants (i.e., producers of electrical power) to electrical loads (i.e., electrical power consumers). Each power plant may correspond to a source of electrical power for the power grid network. The one or more power plants may also be referred to as power suppliers and / or power stations. The one or more power plants may comprise mechanical generators and solar cells with electronic inverters. The RAN and / or the network node may be a load of the power grid network. The grid may be a combined transmission and distribution network that is part of electricity delivery.

[0034] In an embodiment, the method may further comprise obtaining (optionally measuring) the at least one parameter of the power grid network.

[0035] The at least one parameter of the grid (i.e., the at least one power grid parameter) may comprise the frequency of electrical power of the power grid, e.g., an (three-phase) alternating current (AC) frequency.

[0036] A value of the at least one power grid parameter may vary over time. For example, the AC frequency may vary over time.

[0037] The obtaining of the at least one power grid parameter may be performed periodically or aperiodically, e.g., triggered by an event. The event may comprise a change of the at least one power grid parameter (e.g. if the change exceeds a predefined threshold value) or a change in a power plant (e.g., a power supplier) of the power grid network. For example, the changing from one power plant to another power plant may comprise switching from a solar power plant to a wind power plant (e.g., during nighttime). The at least one power grid parameter may be obtained by the network node, optionally by measuring a state (e.g., the frequency) of the power grid network, preferably locally at the network node (e.g., at a power inlet of the network node).

[0038] In an embodiment, the method may further comprise determining the at least one NES feature of the network node based on the one or more obtained PI target values.

[0039] Determining the at least one NES feature may comprise determining a configuration of the at least one NES feature, which is applied by configuring the at least one NES feature. Alternatively, or in addition, determining the at least one NES feature may comprise selecting the at least one NES feature (e.g., the configuration of the at least one NES feature) from a set of NES features (e.g., preconfigured at the network node).

[0040] The network node may determine the at least one NES feature locally and / or independently from other nodes, e.g. solely based on the one or more obtained PI target values and / or a NES capability of the network node. Alternatively or in addition, the network node may report its NES capability (e.g., the configurable NES features at the network node or the Pls which form the PI space controllable by the at least one NES feature) to the network entity.

[0041] Since each network node may have different (e.g., hardware) NES capabilities, the at least one NES feature may be node-specific. For example, different network nodes may select the at least one NES feature from different sets of NES features.

[0042] The NES features may comprise at least one of: a number of active radio antennas, a number of active radio stages or radio frequency chains, a transmit power level in a (e.g., linear) power range of a power amplifier, and operation modes of a radio access technology (RAT). The network node may use the determined at least one NES features of the network node for the configuration of the at least one NES features based on the one or more obtained PI target values.

[0043] In an embodiment, the configuring of the at least one NES feature of the network node may comprise activating or deactivating the NES, e.g., an activation or a deactivation of the at least one NES feature. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise an activation or a deactivation of one or more cells served by the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise changing a modulation scheme of the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise changing a transmission power of the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise changing a bandwidth of the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise changing a number of antennas used by the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node comprises may activate or deactivating a multiple input-multiple output (MIMO) operation of the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise changing a rank of a MIMO transmission or MIMO reception of the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise switch between a single-user MEMO (SU- MEMO) operation and multi-user MEMO (MU-MIMO) operation of the network node. Alternatively or in addition, the configuring of the at least one NES feature of the network node may comprise increasing or reducing a clock rate of one or more processors of the network node.

[0044] Deactivation of the NES may be associated with completely switching off the NES of the network node and / or deactivating all of the NES features, e.g. allowing the network node to operate without any power or energy restrictions. Deactivation of the NES may enable the network node to operate without power or energy constraints, e.g., may enable the network node to perform all the services in its best technologically possible configuration. Activation of the NES may be associated with switching from the deactivated state to the activated state of the NES.

[0045] A MEMO operation can be associated with an increased power consumption, e.g., compared to a rank-1 operation (SIMO or MISO). In the MIMO operation, the network node may consume more power because the MIMO operation requires multiple antennas to simultaneously transmit different data streams. This requires more signal processing units, radio frequency (RF) chains and / or can increase the required transmission power.

[0046] A SU-MIMO operation can be associated with an enhanced power efficiency. Despite the increased power usage, SU-MIMO can lead to better power efficiency. This may be because, by using multiple antennas, SU-MIMO can deliver higher data rates, which means the required tasks can be completed quicker, leading to a lower energy usage per bit of data.

[0047] In the SU-MIMO operation, the network node may communicate with one radio device (e.g., UE) at a time. In this mode, the network node may be focusing its entire signal strength and beamforming capabilities on that one radio device. As a result, the power consumption may be less per active time. Thus, SU-MEMO may be used if the number of served radio device is one or fulfills a SU-MIMO switching criterion (e.g., below a threshold number).

[0048] However, if several radio devices are to be served in a tight schedule, it could result in longer transmission time, and thus overall power spent might increase. The MU-MEMO operation, on the other hand, allows the network node to communicate with multiple radio devices (e.g., UEs) simultaneously. This may need complex processing to separate the signals destined to different radio devices and combine the transmissions from them, which could increase the power consumption due to the computational load by the signal processing. Yet, this can also lead to a more efficient utilization of the spectrum, so the overall active time can be reduced, reducing the total power consumption in some scenarios if the number of served radio devices fulfills a MU-MIMO switching criterion (e.g., above the threshold number).

[0049] In an embodiment, the at least one parameter of the power grid network may comprise a power demand in (i.e., taken from) the power grid network. Alternatively or in addition, the at least one parameter of the power grid network may comprise a power supplied to the power grid network. Alternatively or in addition, the at least one parameter of the power grid network may comprise a type of power plant (or a combination of such types) supplying the power grid network. Alternatively or in addition, the at least one parameter of the power grid network may comprise an alternating current frequency (AC frequency) of the power grid network. Alternatively or in addition, the at least one parameter of the power grid network may comprise a voltage of the power grid network. Alternatively or in addition, the at least one parameter of the power grid network may comprise an amount of carbon dioxide emitted into the atmosphere per energy unit provided by the power grid network. Alternatively or in addition, the at least one parameter of the power grid network may comprise a price per energy unit of the power grid network.

[0050] The power demand may be the total load in the power grid network or total power output from the power grid network. The power supplied to the power grid network may be the total power generation in the power grid network or the total power supplied or input in the power grid network.

[0051] The grid may be a synchronous grid. For example, all producers connected to the grid may be run at the same frequency (e.g., with a very small deviation), and may stay very close in phase with each other and the grid. Generation and consumption may be balanced across the entire grid, because energy may be consumed as it is produced. An imbalance of the generation and the consumption may lead to deviations of the frequency from a nominal frequency. Small deviations from the nominal frequency may be important in regulating individual producers, e.g. as the deviation is used as a way of assessing the equilibrium of the grid as a whole. When the grid is lightly loaded, the grid frequency may run above the nominal frequency. This may be taken as an indication (e.g., by Automatic Generation Control, AGC, systems) across the grid that producers should reduce their output. Conversely, when the grid is heavily loaded, the frequency naturally slows. Governors may adjust their producers so that more power is output (e.g., according to a droop speed control). When producers have identical droop speed control settings, the deviation of the frequency may ensure that multiple parallel producers with the same settings share the load of the grid in proportion to their rating. Alternatively or in addition, the grid may comprise a central control, which may change the (e.g., configuration) parameters of the AGC systems, e.g. over timescales of a minute or longer, to further adjust the regional network flows and the operating frequency of the grid. For timekeeping purposes, the nominal frequency may be allowed to vary in the short term, but is adjusted to prevent line-operated clocks from gaining or losing significant time over the course of a whole 24 hour period.

[0052] The power grid network may comprise an entire synchronous grid, which runs at the same frequency. Alternatively, the power grid network may comprise one or more neighboring power grids not synchronized, optionally even if they share the same nominal frequency. The neighboring power grids may be interconnected for power transfer by an asynchronous or direct current (DC) tie line.

[0053] Failures in the grid may be associated with producers or power transmission lines tripping circuit breakers due to faults leading to a loss of generation capacity for loads, or excess demand. This will often cause the frequency to reduce, and the remaining producers can react and together attempt to stabilize the frequency of the grid above a minimum frequency. By configuring the at least one NES feature based on the one or more PI target values, embodiments of the method can control the network nodes of the RAN, e.g. as a major load of the grid, to perform and deliver their best service to the radio devices when the grid is in low demand and / or to perform a minimum guaranteed service to the radio devices when the grid is in high demand. For example the network node may sense (e.g., measure) a frequency of the received power electricity over time and determine (e.g., estimate) whether the grid is in high demand or low demand. The network node may therefore take advantage of variable pricing in the grid and contribute to the stability of the grid.

[0054] The at least one parameter of the power grid network may comprise the measured state of the power grid network and / or an energy metric information. The energy metric information may implicitly comprise one or more of a current spot price, a future spot price, a short-term penalty, a short-term reward, and an average daily spot price.

[0055] In an embodiment, the configuring of the at least one NES feature may comprises activating the at least one NES feature or configuring the at least one NES feature with stricter limits on power consumption. Alternatively or in addition, the configuring of the at least one NES feature may comprise deactivating the at least one NES feature or configuring the at least one NES feature with higher limits on power consumption.

[0056] "Stricter limits" may mean stricter or lower limits as compared to the limits applied before the configuring step. "Higher limits" may mean higher or greater limits as compared to the limits applied before the configuring step.

[0057] The network node may use a variety of techniques for the NESs, i.e. for configuring the at least one NES feature. The NES techniques may reduce the power consumption (e.g., energy per service step) of the network node by 50% or more, or by a factor in the range of 2 to 10 or more. Therefore, obtaining the one or more PI target values of the RAN based on the at least one parameter of the power grid network and configuring the at least one NES feature (e.g. by activating or deactivating the at least one NES feature) can support a stable operation of the grid. For example, when the power grid network is in high demand and / or on the threshold of brownout or blackout, the network node may activate and / or configure the at least one NES feature with stricter limits on power consumption. On the other hand, when the power grid network is in low demand, the at least one NES feature may be reconfigured or even deactivated to give the network node maximum capability (e.g., transmit power) to provide its best service.

[0058] Activation of the at least one NES feature and / or different configurations of the at least one NES feature may not only reduce the power consumption by the network node, but may also reduce some service quality experiences at the side of the one or more radio devices (e.g., UEs).

[0059] Changes in the quality of service provided may be measured by different types of the Pls for each service. Therefore, defining the one or more PI target values based on the at least one parameter of the power grid network (e.g., a currently measured state of the power grid network) can enable the network node to determine the best configuration of the at least one NES feature for the time being. This can be achieved by the configuration of the at least one NES feature based on the obtained one or more PI target values, which in turn are based on the at least one parameter of the power grid network.

[0060] In an embodiment, the power grid network may comprise two or more power grids.

[0061] The different power grids may correspond to different power plants (e.g., different power suppliers), e.g. with independently controlled AC frequencies (i.e., grid frequencies). The network node may measure a difference (e.g., a deviation) between the different power grids (e.g., as one of the at least one parameter of the power grid network) and / or may measure the change in the power grid network as the power grid network switches from one power grid to another.

[0062] A change in the power grids constituting the power grid network may trigger the configuration of the at least one NES feature of the network node.

[0063] In an embodiment, the method may further comprise measuring a current power consumption of the network node. Alternatively or in addition, the method may further comprise measuring one or more current values of the one or more Pls of the network node.

[0064] The network node may measure its own current energy consumption in real time and / or after applying the configuration of the at least one NES feature and / or periodically. The network node may internally obtain (e.g., determine and / or measure) its own current energy consumption. The one or more PI current values of the one or more Pls may correspond to actual real-time values of the one or more Pls of the RAN.

[0065] The network node may further receive a message from the one or more radio devices (e.g., UEs) served by the network node and / or in the RAN. The message may be indicative of signal measurements by the one or more radio devices. The received message from the one or more radio devices may be used for measuring the one or more PI current values of the one or more Pls of the network node.

[0066] The network node may use the measured (e.g., obtained) energy consumption and / or the one or more PI current values of the one or more Pls of the network node to evaluate the efficiency of the applied configuration of the at least one NES feature of the network node. Alternatively or in addition, the network node may use the one or more PI current values of the one or more Pls of the network node to monitor the performance of the network node. For example, the network node may control its operation (e.g., by means of the configuring of the at least one NES feature) so that the one or more PI current values of the one or more Pls do not exceed the one or more upper limits set by the one or more PI target values and / or are not less than the one or more lower limit values (e.g., guaranteed values) of the one or more Pls (e.g., set by service requirements).

[0067] In an embodiment, the configuration of or for the at least one NES feature of the network node may control or influence the measured power consumption of the network node. Alternatively or in addition, the configuration of or for the at least one NES feature of the network node may control or influence the one or more measured current values of the one or more Pls of the network node. Alternatively or in addition, one or more configurations of the at least one NES feature of the network node may have no impact on the PI current values of the one or more Pls of the RAN. For example, lowering the number of antennas (as an example of the configuration of the at least one NES feature) may not have an impact on the latency (as an example of the PI target value). Alternatively or in addition, one or more configurations of the at least one NES feature of the network node may increase or decrease the PI current values of the one or more Pls of the RAN.

[0068] In an embodiment, the obtaining of the one or more PI target values and / or the configuring of the at least one NES feature of the network node may be performed using an artificial intelligence (Al) model, optionally according to intent-based networking (IBN).

[0069] Due to the complexity of the services to be provided by the network node and their corresponding Pls, any one of the steps of an embodiment of the method may be performed by an Al model, e.g. a neural network system trained by reinforcement learning or supervised learning. The measured current energy consumption of the network node, and / or the measured one or more PI current values of the one or more Pls of the network node, and / or the at least one parameters of the power grid network, and / or the determined at least one NES feature of the network node may be used by the Al model, e.g., for training the Al model.

[0070] The Al model may be part of an IBN system. For example, the intent may specify to selectively reduce and increase the power consumption of the network node based on the at least one parameter of the power grid network.

[0071] In an embodiment, the method may further comprise adjusting the configured at least one NES feature of the network node based on the measured current power consumption of the network node and / or the measured current value of the one or more Pls of the network node.

[0072] Adjusting (e.g., reconfiguring) the configured at least one NES features of the network node may correspond to a feedback control loop for the configuring of the at least one NES feature of the network node. The adjusting may be a substep of the configuring of the at least one NES feature.

[0073] Furthermore, the NES may have more than one NES feature. Therefore, the configuration of the at least one NES feature may lead to multiple possible configurations. Each configuration may have its own impact on the one or more PI current values of the one or more Pls. The adjusting may be performed by the Al model.

[0074] In an embodiment, the method may further comprise transmitting a feedback message to a network entity. The feedback message may be indicative of a measured current power consumption of the network node. Alternatively or in addition, the feedback message may be indicative of one or more measured PI current values of the one or more Pls of the RAN. Alternatively or in addition, the feedback message may be indicative of the obtained at least one parameter of the power grid network. Alternatively or in addition, the feedback message may be indicative of the determined at least one NES feature of the network node. Alternatively or in addition, the feedback message may be indicative of the configuration of at least one NES features of the network node based on the obtained one or more PI target values. Alternatively or in addition, the feedback message may be indicative of the obtained one or more PI target values. The feedback message transmitted by the network node may enable the network entity to have comprehensive information of the status of the RAN, the individual network node, and / or the power grid network (e.g., as measured at multiples sites of the network nodes). The comprehensive information may enable the network entity to provide an effective configuration of the at least one NES feature of the respective network node of the RAN and / or the one or more PI target values.

[0075] The network entity may be part of a core node and / or a central unit and / or a cloud and / or open RAN (0-RAN). The network entity may be located at any part of the RAN, e.g., in a network node.

[0076] In an embodiment, the method may further comprise receiving a control message from the network entity. The control message may be indicative of the at least one parameter of the power grid network. Alternatively or in addition, the control message may be indicative of the one or more PI target values of the network node. Alternatively or in addition, the control message may be indicative of the configuration of the at least one NES feature of the network node depending on the at least one parameter of the power grid network.

[0077] The control message (or multiple control messages) may be received in at least one of the step of obtaining the one or more PI target values and the step of obtaining the at least one power grid parameter.

[0078] The network node may configure the at least one NES feature of the network node based on the obtained (e.g., received) one or more target values of the one or more Pls of the RAN. The network node may adopt (e.g., use) the received configuration of the at least one NES feature of the network node to perform (e.g., apply) the configuration of the at least one NES feature of the network node.

[0079] In an embodiment, the one or more Pls of the network node may comprise an access delay (e.g., a time duration) measured in one or more radio devices. Alternatively or in addition, the one or more Pls of the network node may comprise a data serving latency measured in one or more radio devices. Alternatively or in addition, the one or more Pls of the network node may comprise a system transmission power (TP) of the network node. Alternatively or in addition, the one or more Pls of the network node may comprise a data rate of a radio connection to one or more radio devices. Alternatively or in addition, the one or more Pls of the network node may comprise a data throughput of one or more radio devices.

[0080] The access delay may also be referred to as access latency. The access delay may comprise the timer required by the one or more radio devices for performing a random access procedure to the network node, e.g. until a connection setup is complete.

[0081] The data rate may refer to the amount of data (e.g., a channel capacity or according to a modulation scheme) that can be transferred (e.g., radio transmitted) per second. The data throughput may refer to the actual amount of data that is delivered to the radio device in a specified amount of time.

[0082] In an embodiment, the obtained one or more PI target values may be further based on a number or load of one or more radio device served by the network node. Alternatively or in addition, the obtained one or more PI target values may be further based on a service currently provided by the network node. Alternatively or in addition, the obtained one or more PI target values may be further based on a quality requirement of radio devices served by the RAN. Alternatively or in addition, the obtained one or more PI target values may be further based on a future service to be provided by the network node. Alternatively or in addition, the obtained one or more PI target values may be further based on a capability of one or more UEs served by the network node.

[0083] In an embodiment, the configuring of the at least one NES feature of the network node may be performed at a first rate and at a second rate that is slower than the first rate.

[0084] The configuration of at least one NES feature may be varied at a first (e.g., fast) rate based on short-term composite variations of the at least one parameter of the power grid network by adapting a first set of NES features (fast to reconfigure). Alternatively or in addition, the configuration of the at least one NES feature may be varied at a second (e.g., slow) rate based on medium-term composite variations of the grid parameters by adapting a second set of NES features (fast or slow to reconfigure). The slow rate may correspond with a change in the NES feature configuration for every hour and / or a few hours and / or a day and / or a week or longer. The fast rate may correspond to a change in the configuration of the at least one NES feature for each hour and / or minute or shorter.

[0085] In an embodiment, the obtaining of the one or more PI target values may be performed periodically and / or triggered, e.g. based on the configuration of the at least one NES feature of the network node. Alternatively or in addition, the obtaining of the one or more PI target values may be performed periodically and / or triggered based on a change in the at least one parameter of the power grid network. Alternatively or in addition, the obtaining of the one or more PI target values may be performed periodically and / or triggered based on a change in the measured current power consumption of the network node. Alternatively or in addition, the obtaining of the one or more PI target values may be performed periodically and / or triggered based on a control message received from the network entity.

[0086] In an embodiment, the one or more PI target values may fulfil a performance criterion of a QoS Class Identifier (QCI) or a 5G QoS Identifier (5QI) associated with the network node. The performance criterion may correspond to a numerical value, e.g., a (upper or lower) threshold value, of the lower performance limit.

[0087] The one or more PI target values may be equal to or greater than the threshold value, if greater values of the respective one or more Pls correspond to higher performance and / or higher power consumption (e.g. data rate). Alternatively or in addition, the one or more PI target values may be equal to or less than the threshold value (e.g., an inverse performance threshold value), if smaller values of the respective one or more Pls correspond to higher performance and / or higher power consumption (e.g. latency).

[0088] The threshold value may correspond to as a baseline target value of the one or more Pls. For example, the one or more Pls may be subject to not falling below the threshold value as a PI lower bound value in terms of performance. Alternatively or in addition, the PI lower bound value may correspond to a value that the PI current value (i.e., an actual value of the one or more Pls) may fall below with a probability of 10 %, or 5 %, or less. The threshold value may be service-specific and / or may be a QoS requirement.

[0089] As to a second method aspect, a method of configuring at least one network node of a radio access network (RAN) for network energy savings (NESs) is provided. The method is performed by a network entity of the RAN. The network entity and / or the RAN is connected to a power grid network. The method comprises obtaining at least one parameter of the power grid network. Furthermore, the method comprises transmitting a control message to the least one network node. The control message is indicative of one or more target values of one or more performance indicators (Pls) of the RAN (e.g., of the at least one network node) for configuring at least one NES feature of the respective network node (e.g., the respective network node that is receiving the control message). The one or more PI target values are dependent on the obtained at least one parameter of the power grid network. Alternatively or in addition, the control message is indicative of a configuration of at least one NES feature of the respective network node, the configuration may be dependent on the obtained at least one parameter of the power grid network.

[0090] The network entity may be in communication with the at least one network node (e.g., for exchanging the control message and / or a feedback message). The network entity may be a part or function of a core network (CN) serving the RAN or an operations support system (OSS) or a system for operations, administration, and maintenance (0AM).

[0091] The network entity may be "connected to" the power grid in that the power grid network supplies power to the network entity and / or in that the network entity is in data communication with any functionality (central control system) or sensor (e.g., smart meter) of the power grid network. Optionally, based on any connectivity to the power grid network, the network entity may obtain the at least one parameter of the power grid network (i.e., the at least one power grid parameter).

[0092] Transmitting the control message indicative of the configuration may encompass or trigger configuring the at least one NES features of the respective network node based on the one or more obtained PI target values of the respective network node and / or based on the obtained at least one parameter of the power grid network. Alternatively or in addition, the control message may be broadcast to all network nodes of the RAN or to all network nodes in a tracking area or a registration area of the RAN.

[0093] In an embodiment, the method may further comprise determining the one or more target values of the one or more Pls (of the RAN) for the at least one network node. For example, the one or more PI target values may be based on the obtained at least one parameter of the power grid network. The configuration of the at least one NES feature of the respective network node may be dependent on the obtained at least one parameter of the power grid network by determining the configuration based on the determined one or more PI target values. Alternatively or in addition, the method may further comprise determining the configuration based on the obtained at least one parameter of the power grid network. By transmitting the control message indicative of the determined target value of the one or more Pls for the respective at least one network node, the network entity may control the operation of the RAN, e.g. the operation of the at least one network node.

[0094] The network entity may be a core network node and / or a part of a core network node and / or a function of the core network node and / or a network exposure function (NEF) and / or an operation and administration management (0AM) and / or an operational support system (OSS) and / or a network management and / or a coordinating (e.g., steering) node of the RAN. Alternatively or in addition, the network entity may be in communication with more than one network node. The network entity may be connected to the power grid and / or sense the power grid (e.g., to obtain the at least one power grid parameter). The network entity may monitor a plurality of network nodes of the RAN. Alternatively or in addition, the measured one or more PI current values for the at least one network node may be (e.g., pairwisely) different from the one or more obtained (e.g., received) current PI values by the network node. Alternatively or in addition, in presence of the difference, the measuring of the one or more PI current values may replace the obtained one or more PI current values from the network node to solve the discrepancy and / or the one or more network nodes of the RAN may communicate with each other to determine one or more consistent PI current value.

[0095] In an embodiment, the determining of the one or more PI target values for the at least one network node may be further based on the feedback message received from the at least one network node. Alternatively or in addition, the determining of the one or more PI target values for the at least one network node may be further based on the operational information of the RAN.

[0096] In an embodiment, the one or more PI target values may be dynamically determined in real-time based on a predictive model that is trained by a machine learning technique to anticipates fluctuations in the at least one power grid parameter using historical data (of the at least one power grid parameter) in conjunction with current sensor data (e.g., indicative of the at least one power grid parameter). Alternatively or in addition, the current sensor data may comprise location data of locations of a plurality of radio devices served by the RAN. Alternatively or in addition, the predictive model may be trained to foresee short-term and longterm trends in power availability (or power shortage), optionally allowing the network entity and / or the network node to proactively configure the at least one NES feature of the network node, optionally to maintain a service quality within predefined limits.

[0097] In an embodiment, the method may further comprise receiving a feedback message from the at least one network node. The feedback message may be indicative of a measured current power consumption of the network node. Alternatively or in addition, the feedback message may be indicative of one or more measured current values of the one or more Pls. Alternatively or in addition, the feedback message may be indicative of the obtained at least one parameter of the power grid network, optionally measured at the network node. Alternatively or in addition, the feedback message may be indicative of the at least one NES feature of the network node determined by the network node based on the one or more transmitted PI target values. Alternatively or in addition, the feedback message may be indicative of the configuration of the at least one NES features of the network node based on the one or more transmitted PI target values.

[0098] In an embodiment, at least one of the transmitted control message and the received feedback message may be communicated between the at least one network node (e.g., the respective one network node in case of dedicated control message signaling) and the network entity through a midhaul interface, optionally an Fl interface as specified by 3GPP. Alternatively or in addition, the network node and the network entity may be functionally split into a distributed unit (DU) and a central unit (CU), respectively.

[0099] As to another aspect, a computer program product is provided. The computer program product comprises program code portions for performing any one of the steps of the first method aspect and / or second method aspect disclosed herein when the computer program product is executed by one or more computing devices. The computer program product may be stored on a computer-readable recording medium. The computer program product may also be provided for download, e.g., via the radio network, the RAN, the Internet and / or the host computer. Alternatively, or in addition, any one of the method aspects may be encoded in a Field- Programmable Gate Array (FPGA) and / or an Application-Specific Integrated Circuit (ASIC), or the functionality may be provided for download by means of a hardware description language.

[0100] As to a first device aspect, a device (e.g., a network node) of a radio access network (RAN) for network energy savings (NESs) is provided. The network node comprises memory operable to store instructions and processing circuitry operable to execute the instructions, such that the network node is operable to obtain one or more target values of one or more performance indicators (Pls) of the network node. The one or more PI target values may be dependent on at least one parameter of the power grid network. The network node is further operable to configure at least one NES feature of the network node based on the one or more obtained PI target values.

[0101] In an embodiment, the network node may further comprise the features or may be further operable to perform the steps of any one of the embodiments disclosed herein in the context of the first method aspect. Alternatively or in addition, the network node may further comprise corresponding features or may be further operable to perform corresponding steps that mirror any one of the embodiments disclosed herein in the context of the second method aspect.

[0102] As to another first device aspect, a device (e.g., a network node) of a radio access network (RAN) for network energy savings (NESs) is provided. The network node is configured to obtain one or more target values of one or more performance indicators (Pls) of the network node. The one or more PI target values are dependent on at least one parameter of the power grid network. The network node is further configured to configure (e.g., apply) at least one NES feature of the network node based on the one or more obtained PI target values.

[0103] In an embodiment, the network node may further comprise the features or may be further configured to perform the steps of any one of the embodiments disclosed herein in the context of the first method aspect. Alternatively or in addition, the network node may further comprise corresponding features or may be further configured to perform corresponding steps that mirror any one of the embodiments disclosed herein in the context of the second method aspect.

[0104] As to a second device aspect, a device (e.g., a network entity) for configuring at least one network node of a radio access network (RAN) for network energy savings (NESs) is provided. The network entity comprises memory operable to store instructions and processing circuitry operable to execute the instructions, such that the network node is operable to obtain at least one parameter of the power grid network. The network entity is further operable to transmit a control message to the least one network node. The control message is indicative of one or more target values of one or more performance indicators (Pls) of the RAN (e.g., of the respective radio devices) for configuring at least one NES feature of the respective network node. The one or more PI target values are dependent on the obtained at least one parameter of the power grid network. Alternatively or in addition, the control message is indicative of a configuration of at least one NES feature of the respective network node (e.g., the receiving one network node). The configuration is dependent on the obtained at least one parameter of the power grid network.

[0105] In an embodiment, the network entity may further comprise the features or may be further operable to perform the steps of any one of the embodiments disclosed herein in the context of the second method aspect. Alternatively or in addition, the network entity may further comprise corresponding features or may be further operable to perform corresponding steps that mirror any one of the embodiments disclosed herein in the context of the first method aspect.

[0106] According to another second device aspect, a network entity for configuring at least one network node of a radio access network (RAN) for network energy savings (NESs) is provided. The network entity is configured to obtain at least one parameter of the power grid network. Alternatively or in addition, the network entity is configured to transmit a control message to the least one network node. The control message is indicative of one or more target values of one or more performance indicators (Pls) of the RAN for configuring at least one NES feature of the respective network node, the one or more PI target values being dependent on the obtained at least one parameter of the power grid network. Alternatively or in addition, the control message is indicative of a configuration of at least one NES feature of the respective network node, the configuration being dependent on the obtained at least one parameter of the power grid network.

[0107] In an embodiment, the network entity may further comprise the features or may be further configured to perform the steps of any one of the embodiments disclosed herein in the context of the second method aspect. Alternatively or in addition, the network entity may further comprise corresponding features or may be further configured to perform corresponding steps that mirror any one of the embodiments disclosed herein in the context of the first method aspect.

[0108] Without limitation, for example in a 3 GPP implementation, any of the radio devices may be a user equipment (UE) and / or may be configured for radio access to the RAN, e.g., according to a 3 GPP specification.

[0109] The technique may be applied in the context of 3 GPP New Radio (NR), e.g. as a fifth generation radio access technology of the RAN. The technique may be implemented in accordance with a 3GPP specification, e.g., for 3GPP release 17 or 18 or later. The Pls may be key Pls (KPIs), e.g. as defined in the 3GPP Work Item 810023 on Energy Efficiency of 5G and / or may be Pls for collecting performance measurements, e.g., data volume measurements and PEE (Power, Energy and Environment) parameters from 5G base stations (e.g., according to 3GPP document TS 28.310, version 18.3.0).

[0110] The RAN may be any wireless telecommunications infrastructure and / or may comprise one or more base stations as an embodiment of the network node and / or performing the first method aspect.

[0111] Any of the radio devices may be a 3GPP user equipment (UE) or a Wi-Fi station (STA). The radio device may be a mobile or portable station, a device for machine-type communication (MTC), a device for narrowband Internet of Things (NB-IoT) or a combination thereof. Examples for the UE and the mobile station include a mobile phone, a tablet computer and a self-driving vehicle. Examples for the portable station include a laptop computer and a television set. Examples for the MTC device or the NB-IoT device include robots, sensors and / or actuators, e.g., in manufacturing, automotive communication and home automation. The MTC device or the NB-IoT device may be implemented in a manufacturing plant, household appliances and consumer electronics.

[0112] The wireless telecommunications infrastructure (e.g., the RAN) may be implemented by one or more base stations (e.g., network nodes). The base station may encompass any station that is configured to provide radio access to any of the radio devices (UEs). The base stations may also be referred to as cell, transmission and reception point (TRP), radio access node or access point (AP). The base station and / or the relay radio device may provide a data link to a host computer providing the user data to the remote radio device or gathering user data from the remote radio device. Examples for the base stations may include a 3G base station or Node B, 4G base station or eNodeB, a 5G base station or gNodeB, a Wi-Fi AP and a network controller (e.g., according to Bluetooth, ZigBee or Z-Wave).

[0113] The RAN may be implemented according to the Global System for Mobile Communications (GSM), the Universal Mobile Telecommunications System (UMTS), 3GPP Long Term Evolution (LTE) and / or 3GPP New Radio (NR).

[0114] Any one of the devices, network node (e.g., the base station or DU or CU), network entity (e.g., a core network function), the communication system or any node or station for embodying the technique may further include any feature disclosed in the context of the method aspect, and vice versa. Particularly, any one of the units and modules disclosed herein may be configured to perform or initiate one or more of the steps of the method aspect.

[0115] Brief Description of the Drawings

[0116] Further details of embodiments of the technique are described with reference to the enclosed drawings, wherein:

[0117] Fig. 1 shows a schematic block diagram of a first embodiment of a first device aspect for configuring a network node of a RAN for network energy savings; Fig. 2 shows a schematic block diagram of a first embodiment of a second device aspect for configuring at least one network node for network energy savings;

[0118] Fig. 3 shows a flowchart for a first embodiment of a method of a network node being configured for network energy savings;

[0119] Fig. 4 shows a flowchart for a first embodiment of a method of a network entity for configuring at least one network node for network energy savings;

[0120] Fig. 5 schematically illustrates a first example of a communication system with a RAN comprising a second embodiment of the device of Fig. 1, which may include the features and perform the steps of the method of Fig. 3;

[0121] Fig. 6 schematically illustrates a second example of a communication system with a RAN comprising a third embodiment of the device of Fig. 1, which may include the features and perform the steps of the method of Fig. 3;

[0122] Fig. 7 schematically illustrates a third example of a communication system with a RAN comprising a fourth embodiments of the devices of Figs. 1 and 2, which may include the features and perform the steps of the methods of Figs. 3 and 4, respectively;

[0123] Fig. 8A shows a first example of a schematic diagram of an exemplary parameter of the power grid network as a function of time, e.g., according to a second rate or on a short time scale;

[0124] Fig. 8B shows a second example of a schematic diagram of an exemplary parameter of the power grid network as a function of time, e.g., according to a first rate or on a long time scale;

[0125] Fig. 9A shows a first example of a schematic diagram of a performance indicator of at least one network node of the RAN as a function of time responsive to changes in at least one parameter of the power grid network;

[0126] Fig. 9B shows a second example of a schematic diagram of a performance indicator of at least one network node of the RAN as a function of time responsive to changes in at least one parameter of the power grid network;

[0127] Fig. 10 shows a schematic block diagram of a network node embodying the device of Fig. 1; Fig. 11 shows a schematic block diagram of a network entity embodying the device of Fig. 2; and

[0128] Fig. 12 schematically illustrates an example telecommunication network connected via an intermediate network to a host computer.

[0129] Detailed Description

[0130] In the following description, for purposes of explanation and not limitation, specific details are set forth, such as a specific network environment in order to provide a thorough understanding of the technique disclosed herein. It will be apparent to one skilled in the art that the technique may be practiced in other embodiments that depart from these specific details. Moreover, while the following embodiments are primarily described for a New Radio (NR) or 5G implementation, it is readily apparent that the technique described herein may also be implemented for any other radio communication technique, including a Wireless Local Area Network (WLAN) implementation according to the standard family IEEE 802.11, 3GPP LTE (e.g., LTE-Advanced or a related radio access technique such as MulteFire), for Bluetooth according to the Bluetooth Special Interest Group (SIG), particularly Bluetooth Low Energy, Bluetooth Mesh Networking and Bluetooth broadcasting, for Z-Wave according to the Z-Wave Alliance or for ZigBee based on IEEE 802.15.4.

[0131] Moreover, those skilled in the art will appreciate that the functions, steps, units and modules explained herein may be implemented using software functioning in conjunction with a programmed microprocessor, an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), a Digital Signal Processor (DSP) or a general purpose computer, e.g., including an Advanced RISC Machine (ARM). It will also be appreciated that, while the following embodiments are primarily described in context with methods and devices, the invention may also be embodied in a computer program product as well as in a system comprising at least one computer processor and memory coupled to the at least one processor, wherein the memory is encoded with one or more programs that may perform the functions and steps or implement the units and modules disclosed herein.

[0132] Fig. 1 schematically illustrates a block diagram of an embodiment of a device for configuring a network node of a RAN for network energy savings (NESs). The device is generally referred to by the reference sign 100. The device 100 may be further referred to as or embodied by the network node itself, e.g. a base station or distributed unit (DU) of the base station, or by a core node. The device may further be part of the network node or any base station of the RAN or a core node supporting the RAN.

[0133] The device 100 comprises a PI target value obtainment module 104. The PI target value obtainment module 104 obtains (e.g., receives and / or determines) one or more target values of one or more performance indicators (Pls, e.g. key performance indicators or KPIs) of the network node. The one or more PI target values are dependent on at least one parameter of the power grid network.

[0134] The device 100 further comprises a configuration module 110. The configuration module 110 configures at least one NES feature of the network node 100 based on the one or more obtained PI target values.

[0135] Optionally, the above modules or a further (e.g., reception) module of the device 100 obtains at least one parameter of the power grid network 500. The module 102 may further receive the one or more PI target values of one or more Pls of the network node (e.g., the device 100). The receiving module 102 may further receive a message (e.g., a control message) from a network entity.

[0136] The configuration module 110 may further determine the at least one NES feature of the network node (e.g., the device 100) based on the one or more obtained PI target values. The reception module and / or the configuration module 110 may further measure or receive at least one of a current power consumption of the network node (e.g., the device 100) and one or more PI current values of the one or more Pls of the network node (e.g., the device 100). The configuration module 110 may further adjust the configured at least one NES feature of the network node (e.g., the device 100) based on the measured current power consumption of the network node (e.g., the device 100) and / or the measured current value of the one or more Pls of the RAN.

[0137] The device 100 may further comprise a transmission module. The transmission module may transmit a feedback message to a network entity. The feedback message may be indicative of at least one of a measured current power consumption of the network node; one or more measured current values of the one or more Pls of the network node (e.g., the device 100); the obtained at least one parameter of the power grid network; the determined at least one NES feature of the network node (e.g., the device 100); the configuration of at least one NES features of the network node (e.g., the device 100) based on the one or more obtained PI target values; and the one or more obtained PI target values.

[0138] The configuration module may be in data communication with the receiving module and the transmitting module.

[0139] The device 100 may be implemented in a node of a wireless telecommunications infrastructure, e.g. a node of the RAN. The node may be the network node that is to be configured for the NES.

[0140] Fig. 2 schematically illustrates a block diagram of an embodiment of a device of configuring at least one network node 100 of a RAN for NESs. The device is generally referred to by the reference sign 200. The device 200 may be embodied by a network entity, e.g. embodied by a core node and / or a central server and / or a central unit and / or a cloud implementation and / or an open RAN interface.

[0141] The device 200 comprises a power grid parameter obtainment module 204 that obtains at least one parameter of the power grid network, which is supplying at least one of the network node, the RAN, the network entity, and the core network.

[0142] The device 200 further comprises a control message transmission module 210. The control message transmission module 210 may transmit a control message to the at least one network node (e.g., the above device 100). In one variant of any embodiment, the control message is indicative of one or more target values of one or more performance indicators (Pls, e.g. key performance indicators or KPIs) of the at least one network node 100 and / or the RAN for configuring at least one NES feature of the respective network node. The one or more PI target values is dependent on the obtained at least one parameter of the power grid network. In another variant of any embodiment, which may be combined with the afore-mentioned variant, the control message is indicative of a configuration of at least one NES feature of the respective network node, the configuration being dependent on the obtained at least one parameter of the power grid network.

[0143] Optionally, the device 200 further comprises a receiving module. The receiving module receives a feedback message from the at least one device 100. The receiving module may further obtain operational information of the RAN (e.g., as a basis for determining the NES configuration in combination with the power grid parameter). The receiving module may further obtain the at least one parameter of the power grid network.

[0144] Alternatively or in addition, the device 200 may further comprise a configuration module. The configuration module may determine the one or more target values of the one or more Pls of the RAN for the at least one network node 100. The one or more PI target values may be based on the obtained at least one parameter of the power grid network. Alternatively or in addition, the configuration module may determine the configuration of the at least one NES feature of the respective network node 100, which is dependent on the obtained at least one parameter of the power grid network by determining the configuration based on the one or more determined PI target values. Alternatively or in addition, the configuration module may further determine the configuration based on the obtained at least one parameter of the power grid network.

[0145] Fig. 3 shows an example flowchart of a method 300 of configuring a network node 100 of a RAN for NESs. The method 300 may be performed by the device 100. For concreteness and without limitation thereto, exemplary performance indicators (Pls) are referred to as key performance indicators (KPIs).

[0146] Optionally in a step 301, the method 300 may receive a control message, e.g. from the network entity 200. The control message may be indicative of at least one of one or more KPI target values of the network node 100 and a configuration of the at least one NES feature of the network node 100.

[0147] Optionally, in a step 302, the method 300 may obtain, e.g., measure, the at least one parameter of the power grid network.

[0148] In the step 304, the method 300 obtains 304 one or more target values of one or more KPIs of the network node 100. The one or more KPI target values may be based on at least one parameter of the power grid network.

[0149] Optionally, in a step 306, the method 300 may determine the at least one NES feature of the network node 100 based on the one or more obtained 304 KPI target values.

[0150] Optionally, in a step 308, the method 300 may measure at least one of a current power consumption of the network node 100 and one or more current values of the one or more KPIs of the network node 100.

[0151] In a step 310, the method 300 configures at least one NES feature of the network node 100 based on the one or more obtained 304 KPI target values.

[0152] Optionally in a step 312, e.g. a substep of the step 310, the method 300 may adjust 312 the configured 310 at least one NES feature of the network node 100 based on the measured 308 current power consumption of the network node 100 and / or the measured 308 current value of the one or more KPIs of the RAN.

[0153] Optionally in step 314, the method 300 may transmit 314 a feedback message to a network entity 200, e.g. the node from which the control message had been received. The feedback message may be indicative of at least one of a measured 308 current power consumption of the network node 100; one or more measured 308 current value of the one or more KPIs of the RAN; the obtained 302 at least one parameter of the power grid network; the determined 306 at least one NES feature of the network node 100; the configuration 310 of at least one NES features of the network node 100 based on the one or more obtained 304 KPI target values; and the one or more obtained 304 KPI target values.

[0154] At least one step of the method 300 may be performed using a machine learning model (also referred to as artificial intelligence, Al). For example, the obtaining 304 of the one or more KPI target values and / or the determining 306 of the configuration for the NES feature of the network node may be the result of a neural network system. A first neural network system (e.g., the module 104) for obtaining the KPI target values may be trained based on training data sets, each data set of the training data sets comprising one or more historical KPI target values and at least one corresponding historical power grid parameter. Alternatively or in addition, a second neural network system (e.g., the module 106) for determining the NES feature configuration may be trained based on training data sets, each data set of the training data sets comprising one or more historical NES feature configurations in conjunction with at least one corresponding historical power grid parameter or one or more historical KPI target values.

[0155] Alternatively or in addition, the RAN may be an intent-based network (IBN). For example the first and / or second neural network system may be simultaneously used for controlling the RAN based on one or more intents (e.g. commands and requirements in text-form).

[0156] Fig. 4 shows an example flowchart of a method 400 of configuring at least one network node 100 of a RAN for NESs. The method 400 may be performed by the device 200. The at least one network node 100 may be an embodiment of the afore-mentioned device 100. For concreteness and without limitation thereto, exemplary performance indicators (Pls) are referred to as key performance indicators (KPIs).

[0157] Optionally, in a step 402, the method 400 may receive a feedback message from the at least one network node 100. The feedback message may be indicative of at least one of a measured 308 current power consumption of the network node 100; one or more measured 308 current values of the one or more KPIs; at least one obtained parameter of the power grid network; at least one NES feature of the network node 100 determined 306 by the network node 100 based on one or more transmitted KPI target values; the configuration 310 of the at least one NES features applied by the at least one network node 100 based on the one or more transmitted KPI target values.

[0158] In a step 404, the method 400 obtains (e.g. by measuring or collecting measurement data of) at least one parameter of the power grid network.

[0159] Optionally, in a step 406, the method 400 may determine the one or more target values of the one or more KPIs of the RAN for the at least one network node 100. The one or more KPI target values are based on the obtained 404 at least one parameter of the power grid network. As a consequence, the configuration of the at least one NES feature of the respective network node 100 may depend on the obtained 404 at least one parameter of the power grid network. The subsequent determining 306 or 408 of the configuration based on the one or more determined 406 KPI target values may be performed by the network node 100 or the network entity 200. Optionally, in a step 408, the method 400 may determine the configuration based on the obtained 404 at least one parameter of the power grid network.

[0160] In a step 410, the method 400 may transmit a control message (or shortly a message) to the least one network node 100. The control message may be indicative of one or more target values of one or more KPIs of the RAN for configuring at least one NES feature of the respective network node 100. The one or more KPI target values may be dependent on the obtained 404 at least one parameter of the power grid network. Alternatively or in addition, the control message may be indicative of a configuration of at least one NES feature of the respective network node 100. The configuration may be dependent on the obtained 404 at least one parameter of the power grid network.

[0161] At least one step of the method 400 described above may be performed using an Al at the network entity, optionally an intent based networking (IBN), e.g. as described above in context of the network node 100 with the same or corresponding features implemented at the network entity.

[0162] For concreteness and not limitation, an example gNB is described below as the device 100, with exemplary references to a UE as the radio device served by the gNB 100 and KPIs as the Pls. Furthermore, the KPI target values may be abbreviated to KPI targets. Moreover, power consumption may be referred to as energy consumption.

[0163] A gNB 100 adapts the configuration and activation of NES features (e.g., any gNB features that have a noticeable energy consumption impact) based on energy cost-related criteria on the medium and / or on short time scale. Consistently, the NES feature configurations are selected that lead to minimal power consumption while meeting current performance or KPI targets. The NES features are thus not applied if they would unduly impact the user experience or other predefined user KPIs or network KPIs (NW KPIs, e.g. RAN KPIs).

[0164] In one example, e.g. given that CO2-emmissions are reflected by energy prices, the one or more KPI target values may be adjusted based on energy prices, whereby the gNB 100 is able to steer the NES mode selection towards more effective NES features, sufficient to meet potentially lower KPI targets, when the current energy price is less favorable, and steer it towards reduced NES features, meeting stricter KPI target values, when the energy price is more favorable.

[0165] In other words, the currently acceptable KPI impact or reduction or the minimum acceptable KPI value is appropriately modified or biased, e.g. allowing a larger additional latency or reduced system (or UE) throughput as a result of applying a NES feature if the current energy cost is less favorable, and vice versa. Lower bounds for acceptable KPI or Quality of Experience (QoE) constraints may be additionally applied; those bounds may be UE-, traffic type-, or service / application-specific.

[0166] In case the net price for short-term consumption is negative, the gNB 100 may inactivate NES features in the step 310 and deliberately allow for an increase in the instantaneous power consumption even when the KPIs would be fulfilled with NES features activated. The gNB may have different KPI thresholds for the different price levels. As such the gNB may occasionally exceed the “baseline” (e.g., price level of 50 Hz frequency) KPI requirements / thresholds and instead comply with higher KPI thresholds or alternately not be bound to any KPI requirement / thresholds during such low price periods.

[0167] The possible NES features that may be applied may be divided into medium-term measures (cell deactivation, configuration of the synchronization signal block (SSB) and / or the system information (SI), etc. (e.g., any other NES feature that can be reconfigured via RRC) that are adapted at the longer time scales, e.g. according to spot price fluctuation, and short-term measures (cell-DTX / DRX, port adaptation, etc. that can be reconfigured via LI- and / or L2- signaling and / or L3-signaling) that are adapted according at shorter time scales, e.g. according to instantaneous fluctuations of the AC frequency of the power grid network. For example, configuring 310 the at least one NES feature may comprise controlling via LI- and / or L3- signaling the Cell DTX / DRX. Alternatively or in addition, configuring 310 the at least one NES feature may comprise port- and / or power adaptations via LI- and / or L2-signaling.

[0168] The at least one parameter of the power grid network (power grid parameter) may comprise an energy price or energy cost metric, which may include the current spot price, future predetermined spot price, short-term penalty when instantaneous frequency is below 50 Hz, short-term reward when instantaneous frequency is above 50 Hz, etc. The favorability of the current energy price may be assessed in absolute or relative terms.

[0169] The energy metric assessment and / or corresponding KPI modification may be performed in the network entity, e.g. a core NW function or another dedicated node that may provide relevant operating mode information to one or more network nodes (e.g., radio nodes like gNBs). The operating mode information may comprise levels or classes of applied NES effort, one or more KPI relaxations allowed, etc.

[0170] In one embodiment of the network node 100 as a radio NW aspect or as the corresponding first method aspect, the method 300 in the network node 100 of the RAN (e.g., a RAN node such as a gNB) is a method for power grid-aware (e.g., energy cost-aware) NES activation. Alternatively or in addition, the method 300 comprises obtaining 304 a current target NW KPI information (e.g., as the control message) indicative of the one or more KPI target values (also: "target KPIs"), wherein the target KPIs depend on the current at least one power grid parameter (e.g., a current CO2 emission rate or a current energy cost metric).

[0171] Alternatively or in addition, the method 300 comprises a step of obtaining one or more available NES feature configuration information (e.g., another control message or the aforementioned control message) including estimated KPI impact and / or NES impact for each configuration.

[0172] According to the step 310, the method 300 applies one or more NES configurations that minimizes a metric (e.g., a loss function) of the at least one power grid parameter (e.g., an energy cost metric) subject to meeting the target KPIs.

[0173] The method 300 may be implemented by a step 304 of receiving the current target NW KPI information (i.e., the control message indicative of the KPI target values) from a network entity (e.g., a second NW entity other than the network node). The one or more KPI target values may be received explicitly (e.g., one or more target KPI values being encoded in the control message) and / or implicitly (e.g. an index to a gNB operation mode and / or a regime associated with one or more target KPI values may be included in the control message).

[0174] In any aspect and any embodiment thereof, the KPIs (e.g., the KPI metrics) may comprise one or more of: access latency, data serving latency, system transmit power (TP), and UE throughput. The one or more KPI target values may be determined (e.g., by the network node 100 or the network entity 200) based on at least one of: user equipment (UE) load, current services, and one or more quality requirements (e.g., QoS requirement or QoS class identifier (QCI) or 5G QoS identifier (5QI)) of the UEs (e.g., present UEs, camping UEs, and / or served UEs), services and / or applications, and UE capabilities.

[0175] In the step 310 or 312 of the method 300, the NES configuration may be varied at a first rate (e.g., a rate relatively faster compared to a second rate) based on short-term variations of the power grid parameter (e.g., composite price variations) by adapting a first set of NES features (e.g., those NES features that are fast to reconfigure). Alternatively or in addition, in the step 310 or 312 of the method 300, the at least one NES feature is varied at a second (e.g., slower) rate based on medium-term variations of the power grid parameter (e.g., composite price variations) by adapting a second set of NES features (e.g., those NES features that are fast or slow to reconfigure).

[0176] Any of the steps 310 and 312 of the method 300 may, alternatively to the above or in addition, comprise deactivating the at least one NES feature without considering KPI metrics, or alternately applying maximum possible or infinite KPI bias term, or alternately adopt maximum possible or infinite target KPI threshold and / or KPI target value, if the at least one power grid parameter is indicative of an excess supply of electrical power in the power grid network (e.g., indicated by a composite price being negative).

[0177] The second device aspect, e.g., as a core NW (or other) steering aspect may perform, and / or the corresponding second method aspect may comprise, a method 400 in a network entity 200 (e.g., a second NW entity other than the network node). The network entity 200 may be a core NW node or a core function of the core network supporting the RAN, or a Network Exposure Function (NEF) or an Operations, Administration and Maintenance (0AM) function, or a coordinating and steering node in the RAN for steering power grid-aware NES configuration (e.g., energy cost-aware NES activation).

[0178] The method 400 comprises obtaining 404 a (e.g., current or future) power grid parameter (e.g. an energy cost metric) and / or obtaining operation information (e.g., a status) of an operating mode or regime of the network node 100 or the RAN, which in turn is based on a (e.g., current or future) power grid parameter (e.g. an energy cost metric).

[0179] In a determining step 406, the one or more KPI target values (e.g., a current target NW KPI information) is determined based on the (e.g., current or future) power grid parameter (e.g. current or future energy cost metric) and / or the operation mode information (optionally a KPI lower bound).

[0180] In the step 410, the method 400 signals the one or more KPI target values (e.g., current target NW KPI information) to one or more network nodes 100 (e.g., radio NW nodes). This information may be comprised in a control message and / or may be explicitly signaled (e.g., as one or more target KPI values) or implicitly signaled (e.g. as an index to a gNB operation mode and / or regime associated with one or more target KPI values).

[0181] In any aspect and any embodiment thereof, the at least one power grid parameter (e.g., the energy cost metric information) comprises one or more of: a CO2 emission per energy unit, an energy source mix indicative of the fraction of renewable energy in the power provided by the power grid network, a current spot price, future sport price, short-term penalty, short-term reward, and average daily spot price. Optionally, the at least one power grid parameter is obtained via another NW internal or external node, or an 0 AM function.

[0182] The determining 406 of the one or more KPI target values may comprise determining a modified KPI target value (e.g., a modified target NW KPI), e.g. by adding a power grid parameter-dependent bias term (e.g., an energy price-dependent bias term) to a baseline KPI target value (e.g., a KPI lower bound value). As a result, the corresponding NES configuration based on the modified KPI target value may be subject to the performance of the network node 100 not falling below the KPI lower bound value.

[0183] In any aspect and any embodiment thereof, the power grid parameter-dependent bias term may be positive or negative depending on the KPI type. I.e., higher KPI in terms of throughput is associated positive bias whereas higher KPI in terms of latency is associated with a negative bias (reduced latency).

[0184] In any aspect and any embodiment thereof, the power grid parameter-dependent bias term is the opposite of the above if the power grid parameter is indicative of a power shortage (e.g., if energy price is unfavorable) in relation to an absolute or relative parameter threshold (e.g., a price threshold) or to an intent-related parameter threshold (e.g., a price threshold).

[0185] In any aspect and any embodiment thereof, the determining 406 of the one or more (e.g., modified) KPI target value (e.g., a modified target NW KPI) by adopting a power grid parameter-dependent KPI target value (e.g., an energy price-dependent target KPI value). The power grid parameter-dependent KPI target value may be different from the baseline target KPI value, optionally subject to not falling below the KPI lower bound value.

[0186] Any one of the one or more KPI target values may function as a threshold for selectively activating or deactivating the at least one NES feature at the network node according to the configuring step 310.

[0187] Any aspect and any embodiment thereof may be implemented in a setup (e.g., a telecommunication system) comprising at least one of the following features or obeying to at least one of the following principles.

[0188] The examples (e.g., as detailed in the detailed embodiments below) may refer to a NES mode (e.g., the configured at least one NES feature). In one example, a NES feature can be defined as a feature which is saving more energy than a default mode of operation of a gNB 100. E.g., a gNB 100 is typically associated to a product, and a product may entail a specific number of receive (RX) and / or transmit (TX) chains, antennas, transmit power (TP), bandwidth (BW), number of carriers and so on. In one example, the maximum requirements possible by a gNB 100 may be defined as default. While in another example, a typical default mode of operation is either defined in the products or can be obtained through e.g., operator documentations, a database, etc. Such default mode of operation can also be learned, e.g., through collecting data over a reasonable time duration, e.g., 24h. Example, a default mode of operation can be applying 64 antennas in DL / UL, 55 dBm (i.e., decibel milliwatt) transmit power, 100 MHz BW, 8 carriers, etc. And this any feature leading to lower any of such requirements can be defined as a NES feature. Other definitions of a NES feature, e.g., as part of standards and regulations are not excluded.

[0189] There exist a number of NW configuration features that can strongly reduce gNB energy consumption - e.g. putting Primary Cell (Pcell) / Special Cell (Spcell) / Scells to sleep, applying UE and cell DTX / DRX, extending / adapting SSB periodicity and reducing monitored physical random access channel (PRACH) or scheduling request (SR) occasions, turning off parts of transceiver circuitry and MIMO modes, turning off support to multiple RATs or inactivating, Dynamic Spectrum Sharing (DSS), adapting common signals / channels, etc. - that will be jointly referred to as NES features.

[0190] In addition, most NES features can be made more aggressive or less aggressive by tuning configuration parameters. For example, the low-energy scheduling solution (LESS) NES feature can increase or decrease the allowed scheduling latency, e.g. wherein a larger maximum scheduling latency will result in increased energy saving but with a larger negative KPI impact. A MIMO sleep NES function in a 4-TX radio unit (RU) may change the number of antennas (e.g. 1, 2, 3, or 4) that remain active during MIMO sleep occasions. Also, switch-up and switchdown thresholds and timers may be adjusted.

[0191] Conventionally, the NW (e.g., the gNB 100) may apply them based on a trade-off to achieve the minimum possible total energy consumption while maintaining a minimum acceptable performance KPIs like service capacity and quality of experience (QoE), where the required service capacity and / or QoE may depend on the user load, current services or quality requirements of the UEs present, etc. The gNB 100 energy consumption rate with or without different NES features activated may vary many-fold (e.g. 2-10 times or more). The conventional NES feature adaptation approaches have focused primarily on maximizing energy savings while minimizing adverse impact to NW performance KPIs.

[0192] The present technique may be embodied to use a NES feature and / or other network (NW) configuration control responsive to the power grid parameter, e.g. the economic impact of average and instantaneous energy consumption in the gNB 100. The radio access network (RAN) nodes 100 adapt the activation or configuration of NES features not only based on meeting performance KPIs but including power grid-related (e.g., energy-related costs) on the medium-term time scale (e.g. a time scale on the order of 1 hour (h), e.g. as illustrated in Fig. 8A based on e.g. auctioned spot prices) and / or on the short-term scale (e.g. a time scale on the order of 10 s, e.g. as illustrated in Fig. 8B based on e.g. power grid voltage frequency monitor and related incentives). By suitably modifying KPI targets, the NES mode selection is biased towards more effective NES features when the current energy price is less favorable and towards reduced NES features when the energy price is more favorable.

[0193] In some embodiments the performance requirements (e.g. 5th-percentile user throughput, median user throughput, 5th-percentile service latency, etc.) are adjusted based on a current or future energy price information. A network energy performance optimizer algorithm may then try out different combinations of NES feature activation and parameters and select the configuration with the lowest energy consumption that fulfill all of the currently active electricity price dependent performance requirements.

[0194] Independent of the skilled person's choice of splitting the subject technique into two method aspects 300 and 400, any aspect of the technique may comprise at least one of the following steps in a wireless network

[0195] In a step 302 or 404, a current power grid parameter (e.g., an energy cost metric) may be obtained.

[0196] In a step 304 or 406, one or more current KPI target values (e.g. target NW KPIs) may be determined based on the current power grid parameter (e.g., an energy cost metric), optionally and a lower performance limit (e.g., a KPI lower bound information).

[0197] In a step 306 or 408, an available NES feature configuration (e.g., configuration set information) may be obtain, e.g. including an estimated KPI impact and / or a NES potential for each configuration.

[0198] In a step 310 or 410, a NES configuration that minimizes current energy consumption subject to meeting the current target KPIs is applied (e.g., directly at the network node 100 or indirectly by means of the control message from the network (NW) entity 200).

[0199] More specifically, the step 302 or 404 may use at least one of the following features or sub-steps for obtaining the power grid parameter (e.g., an energy cost metric info).

[0200] A NW entity (referred to above as the second NW entity 200) may receive 404 a first information from a source about current or projected energy prices. The energy price may include the current spot price, short-term penalty when instantaneous frequency below 50 Hz, short-term reward when instantaneous frequency below 50 Hz, taxes, distribution fees, etc. The energy price may be associated with one or more energy providers, where the operator / NW may have the option to switch to another provider if a more favorable price is obtained.

[0201] The NW entity 200 receiving said first and / or second information (and optionally later providing it to one or more RAN nodes 100 in the step 410) may be a 3 GPP node such as a core network node. For example, the node may be a newly introduced node or an existing node such as the Policy Control Function (PCF) that directly or indirectly via any of the other core NW nodes such as Access and Mobility Management Function (AMF), Session Management Function (SMF), User-plane Function (UPF) or alike provide the first / second information to the gNB. Alternately, any of the nodes AMF / SMF / UPF are the entities that receive the first energy price and provide the first / second information to the gNB. It can even be so that the gNB itself, or another dedicated RAN node, retrieves the first price information. Alternately, the entity providing the first / second information to the gNB may be an Open RAN (O-RAN) entity. Another option is that a third party application either provides the information or controls the 3GPP NW behavior via the Network Exposure Function (NEF) existing or new application programming interfaces (APIs). The source of the first (price) info can be an external application / server or an internal / external operational, administration and maintenance (0AM) entity from which the information can be retrieved. The gNB may thus be able to control its operations directly or indirectly based on info about the past average, current, and pending energy prices, penalties, and rewards from other NW nodes, via 0 AM services, RAN interfaces, etc.

[0202] In one non-exclusive example, the entity may be an energy optimization unit working within a gNB, or managing multiple gNBs. It can be part of RAN, or NAS, openRAN, or other non-RAN related components of the NW, e.g., Core network node. In the simplest form, the said first / second information is to be used by the gNB for its operations in all of cells and applicable to all UEs. But in another aspect, any potential KPI compromise (reward / penalty) for the sake of energy saving is only applicable to certain UEs or certain cells of the gNB. Alternately, certain UEs or cells are exempt from any KPI impact inflicted by the methods mentioned herein.

[0203] The favorability of the current energy price may be assessed in absolute terms, e.g. in relation to absolute price thresholds, or in relative terms, e.g. fractional thresholds in relation to average energy prices during a predetermined longer time interval.

[0204] An order to reduce energy consumption from the national grid operator can be treated as an extremely high price signal or penalty value.

[0205] Alternatively or in addition, the step 304 or 406 may use at least one of the following features or sub-steps for target KPI determination.

[0206] The NW entity next determines current KPI targets based on the current the energy cost metric. The target KPI values may be determined by applying a bias or an offset biasing to a baseline performance KPI value. The biasing may be in a form of modifying the minimum acceptable KPI value or the currently acceptable KPI impact / reduction, e.g. allowing a larger additional latency or reduced system or UE throughput as a result of applying a NES feature if the current energy cost is less favorable. Lower bounds for acceptable KPI or QoE constraints may be additionally applied; those bounds may be UE-, traffic type-, or service / application- specific.

[0207] In one formulation, the target KPI determination may be captured by the principle that the gNB can compromise on the resulting QoE, where the higher the price, the more compromise is permitted. Compromise is applied on functions that would have been turned on or used in legacy operation, i.e. the KPI modification is distinguishable from the energy-price-agnostic cases where energy saving schemes would have been used anyway. For example, regardless of the energy price, if SNR for a UE is high or only single-layer operation is possible, we can mute antennas in legacy operation. When applying the current invention, we may sacrifice QoE and turn off antennas or not offer MIMO scheduling to a UE in good conditions despite higher rank reported if the energy price is high. The energy price-dependent KPI / QoE thresholds may be illustrated with the following examples. One way to control this could be that during low-to-normal energy prices the network aim at achieving e.g. a 5th percentile UE throughput of X Mbit / s, while during high electricity cost hours this target is reduced to alpha x X, where 0 < alpha < 1. Viable example numbers could be e.g. X = 20 Mbit / s, alpha = 0.25, high price threshold = 0.5 Euro / kWh.

[0208] There could also be more than one threshold, e.g. (alpha l = 0.9, alpha_2 = 0.5, alpha_3 = 0.25) with corresponding (fixed or relative to mean price) electricity cost thresholds (threshold l = 0.2, threshold_2 = 0.4, threshold_3 = 0.6 Euro / kWh).

[0209] Alternatively, the electricity threshold may be adaptive (e.g. relative to the current average price). This mode may be the preferable mode when the average price is relatively high and there are real cost savings opportunities, e.g. based on the variance or another variability metric of the instantaneous prices. Since electricity prices fluctuate much over time, it may not be preferable to apply this strategy during consistently low price periods — the NW could end up reducing performance during summer months when electricity prices have been low for 3 consecutive months and electricity price goes above 0.1 Euro / kWh.

[0210] In addition, the NW could also apply electricity price thresholds per service or per UE with corresponding performance reduction factors (alpha values). E.g. the target Voice over Internet Protocol (VoIP) latency is allowed to increase with a factor of alpha voip = 2 when electricity cost is above threshold voip. Accuracy requirements of positioning services or for certain UEs could e.g. be relaxed (from a normal requirement of 0.1 meter to a high electricity cost requirement of 0.5 meter). Critical machine type communication (C-MTC) requirements could be allowed to be relaxed with slightly higher latency (alpha cmtc latency = 1.5) and slightly higher error probability (alpha cmtc errorrate = 2).

[0211] The NW could also allow for higher interference or distortion by allowing its power amplifier (PA) to operate in a less linear mode when electricity cost is high. The error vector magnitude (EVM) requirement could be increased with alpha evm when electricity cost more than threshold evm. Same applies to adjacent channel leakage power ratio (ACER), and other radio performance related requirements. In one embodiment, this (e.g. allowing for higher PA non-linearity) is achieved by e.g. limiting the peak modulation order. The highest modulation (256 QAM) could have an electricity price threshold (threshold_256qam) prohibiting the use of 256 QAM when the electricity price is above that threshold.

[0212] Other high-energy usage functions like high rank transmissions, high bandwidth operation, activating some carriers, deactivating transceivers, operating with reduced output power, etc. could also have prohibiting electricity price thresholds.

[0213] In one embodiment, traffic load and is additionally considered as a KPI adjustment tradeoff. In some cases, typical high energy price hours are correlated with high traffic / loads where spending additional energy may be acceptable for getting data transmitted / received. Therefore, a variable minimum QoE or KPI level may be considered, where the level may depend on the NW load. Intent-based automation techniques including e.g., Al methods, closed loop control algorithms, etc. can help in reduced cost related to energy consumption while still fulfilling a level of QoE. For example, the intent could be "I want my NW energy cost to not exceed a certain limit while a minimum QoE which is still 5G worthy is satisfied". The gNB would then compromises on QoE based on the price, the higher the price, the more compromise. In an intent-based automation setup, the intent value may be a target energy cost and the favorability may be assessed in relation to the target level. Some intents (such as average / maximum / . . . energy cost during X / Y / Z hours / scenarios) may be used to adapt the NES configurations based on the current energy price. The IB A framework may be used to reduce the cost of manual configurations after observing specific events.

[0214] In one example, the KPI, its threshold, biasing, etc., as exemplified here can be determined in a gNB, e.g., in the RAN, where it can also become UE specific, area specific, or gNB specific depending on the need or application. E.g., a common signal KPI such as SSB, can be defined to be area or gNB specific, but a DL data channel related one can be UE specific. In another example or complementary, the KPI can be defined over multiple gNBs 100, e.g., the NW capacity over multiple gNBs. As such, either the gNB or a unit managing multiple gNB such as core or energy optimization unit can keep track of the underlying KPI.

[0215] The current KPI target information is then provided to one or more RAN nodes 100. The NW entity 200 related to step 404 may provide the first information itself, or a second KPI- or NES-related policy information based on the first information, to a RAN node 100, where the NES features are to be configured. Examples of second information can be usage of certain NES feature, or setting energy level targets (e.g., average / instantaneous / . . . energy levels), or setting indirect KPIs / policies for other features (such as changing the KPI for UE / Cell throughput, or latency, broadcast transmission rate, or alike) which in turn affect energy levels used by the system.

[0216] In one embodiment, the step 304 or 406 may be performed in above-mentioned second NW entity 200, e.g. in a core NW or NEF or 0AM functions or other specialized nodes for performing KPI steering based on power grid parameters (e.g., energy cost metrics), and the gNB 100 may receive the determined KPI values or an index to a KPI set in a predetermined KPI list, to perform NES adaptation in the next step.

[0217] In another embodiment, the target KPI determination step 304 or 406 may be performed in the RAN node (gNB) 100 that will be applying the NES adaptation, based on receiving an energy cost metric-dependent gNB operation mode / regime index or indication and applying predefined KPI target values corresponding to that index.

[0218] By controlling the current KPI target based on the current energy cost metric, and providing only the KPI target as input to NES feature adaptation, there will subsequently be extensive freedom for any algorithm that configures the network nodes to do what is necessary to reach the target with as little energy usage as possible.

[0219] Alternatively or in addition, the step 306 or 408 may use at least one of the following features or sub-steps for NES configuration selection. The target KPIs determined in step 304 or 406 are next used in one or more RAN nodes (gNBs) 100 to select 306 or 408 NW configurations to operate at an appropriate energy consumption level. The features to configure may include explicit NES features, e.g. antenna / channel state information reference signal (CSI-RS) port adaptation, UE discontinuous reception (DRX) and cell discontinuous transmission (DTX) and DRX configurations, Secondary cell (Scell) activation latency configuration, common signals / channels output power adaptation such as SSB, PRACH, SR occasion / periodicity adaptation, spatial, power scheduling request occasion adaptation, SSB and / or frequency domain energy saving adaptations, System Information provision rate adaptation, etc.

[0220] Instead of, or in addition to, explicit NES features, the configuration selection may relate to, more generally, gNB features that have a noticeable energy consumption impact. This may also include general gNB algorithm and hardware (HW) configurations that affect the gNB power consumption, e.g. DL transmission and UL monitoring rates, transmitter (TX) power allocation, clock gating and power domain activation / deactivation in HW units.

[0221] Lowering TX power, e.g. user-specifically or modulation-and-coding-scheme (MCS)- specifically, can also be a way of reducing a KPI value. Limiting or reducing the bandwidth both in frequency and time domain can also be used to reduce the KPI and the power consumption. For example, the bandwidth is reduced from 100 MHz to 50 MHz or the number of available time slots are reduced with 50 % which allow more sleep. Both changes above reduce the available capacity with 50 % but also enabling a flexible way of reducing the power consumption when needed due to energy cost fluctuations.

[0222] The possible NES features that may be applied include medium-term measures that can be reconfigured via radio resource control (RRC) or system information (SI) transmissions that are adapted at the longer time scales (e.g. full cell deactivation, SSB / SI configuration, etc.), e.g. according to spot price fluctuation.

[0223] It also includes short-term measures that can be reconfigured via Ll / 2 signaling or L1 / L3 signaling (e.g. cell-DRX / DTX, port adaptation, like ignoring PRACH, putting Scells to sleep, turning off MIMO, etc.) may be adapted according at shorter time scales, e.g. according to instantaneous frequency fluctuations.

[0224] The NES configuration selection (e.g., in the step 408) may be performed in two sub-steps. First, a list of possible configuration aspects or their combinations may be determined, and for each such aspect or combination, reachable KPI values and associated energy consumption metrics are determined, e.g. based on previously measured or otherwise estimated information. Second, the NES configuration with minimum energy consumption is chosen that meets all (or a predetermined subset of) target KPIs.

[0225] In one embodiment, the NES optimization may include RAT optimization, e.g., if most UEs support 5G, 4G support may be turned off, or if most support 4G then 5G could be deactivated, or dynamic spectrum sharing (DSS) (where LTE and NR is operated in parallel on the same carrier) may be avoided, and so on. In case the net price (which may be the aggregate of the spot price, incentive, distribution fee, taxes, and other price components) for short-term consumption is negative, e.g. the spot price is negative or the stabilization reward exceeds the spot price, the gNB may inactivate all NES features and deliberately increase the instantaneous energy consumption or steer the traffic time to these occasions even when the KPIs would be fulfilled with the features activated. In one embodiment, the target KPI values are ignored (or set to maximum target values) and the gNB is configured in its maximum (or high) energy consumption mode.

[0226] Alternatively or in addition, the step 310 or 410 may use at least one of the following features or sub-steps for an operation with selected NES configuration.

[0227] The gNB 100 subsequently applies 310 the NES configuration in one or more operating features and operates using the selected configuration. During the operation, the NES feature adaptation algorithm, i.e., the method 300 or 400 may continue analyzing (e.g., in the step 302 or 304) the current and upcoming at least one power grid parameter (e.g., energy price), optionally in combination with information on a use case or scenario to update the NES configuration (e.g., in the step 302 or 312 or the steps 406 and 410) or according to the any of the above-mentioned steps.

[0228] The technique may be embodied based on any power grid parameter indicative of environmentally harmful emissions (which may be highly correlated with energy costs) per energy unit or power plan types involved in the power grid network for some RAN deployments using other energy carriers (e.g. gas, diesel etc.).

[0229] Fig. 5 shows a first exemplary communication system 700 comprising a device 100 (e.g. network node) according to Fig. 1 for performing the method 300 according to the Fig. 3. The network node 100 may be connected to the power grid network 500. The power grid network 500 may provide (e.g., supply) energy to the one or more network node 100 (e.g., as a power load of the power grid network 500). The power grid network may be supplied by one or more types of power plants, e.g., a solar power plant, a wind power plant, a water power plant, a fossil power plant, etc.

[0230] Herein, the words "power" and "energy" are often used interchangeably in this description for historical reasons in this field, but it is clear to the person skilled in the art that power (as a unit of energy per unit of time) is physically different from the energy and related, e.g. by considering appropriate time steps as the unit of time.

[0231] In fact, it is not possible for the power grid network to be constantly supplied with the same power and the same parameters, e.g., the same frequency. Therefore, an feedback-loop control of the power grid network 500 may use some policies to encourage or discouraging power consumers (e.g., power loads) to use more or less power, respectively, to better balance the input (e.g., production) and output (e.g., consumption) of the power of the grid.

[0232] The power consumption of the network nodes 100 may depend on, for example, on the number of radio devices 600 (e.g. UEs) to be served by the network node 100. For example, the right-hand network node 100 may consume more power compared to the left-hand network node 100 of the exemplary communication system 700. The network entity 200 or the network node 100 may sense (e.g., obtain and / or monitor) the parameters of the grid 500 according to the step 404 or 302, respectively. For example, if the sensed electrical frequency 806 of the grid 500, as illustrated in Fig. 8B, is lower than the nominated electrical frequency, the network node 100 may interpret this as a sign of the imbalance in the grid, herein the output of the grid 500 is higher than the input of the grid 500. Therefore, according to the method 300, the network node 100 may activate the NES and / or configure (e.g., reconfigure) at least one of the NES features of the network node 100.

[0233] The method 300 or 400 may be node-specific. For example, the left-hand network node 100 may not necessarily reconfigure and / or activate the at least one NES feature.

[0234] Fig. 6 shows a second exemplary communication system 700 comprising an embodiment of the device 100 (e.g. network node) according to Fig. 1 for performing method 300 according to the Fig. 3. The network node 100 may be connected to the power grid network 500. In this example, the grid 500 may be supplied by three different types of power supply (a, b and c). The left-hand network node 100 may be supplied by two grids 500-a and 500-b, optionally via different power interfaces. In an exemplary situation, when the network node 100 detects (e.g., by sense or obtain the power grid parameter in the step 301 or 302) that the grid 500-b is unstable (e.g., obtained unstable grid parameters), the network node 100 may, according to the method 300, switch to the stable grid 500-a.

[0235] In another exemplary situation, if the grid 500-b is perceived by the network node 100 to be unstable (e.g., a higher electrical frequency is obtained) according to the method 300, the network node 100 may switch to the grid 500-b, as this may be a sign of low consumption of the grid 500-b. In this example, low consumption of the grid 500-b may often be correlated with a lower price of the energy, therefore the method 300 may have the advantage of lowering the price of the energy consumed by the network node 100 while consuming power and providing quality of experience to the radio device 600.

[0236] Fig. 7 shows a third exemplary communication system 700 comprising an embodiment of the device 100 (e.g. network node) according to Fig. 1 and a device 200 according to Fig. 2 for performing method 300 and 400 according to the Fig. 3 and 4, respectively.

[0237] The network node 100 may transmit 314 a feedback message to the network entity 200 and the network entity 200 may receive 402 the feedback message from the network node 100. Therefore, the network entity 200 may have sufficient or comprehensive information of the RAN 700, e.g. as to local power consumption of the individual network node 100 and / or the at least one power grid parameter, which may be advantageously measured locally at the individual network node 100 to detect or predict a local fluctuation or a local outage. In this example the network entity 200 may provide an effective configuration of the at least one NES feature of the one or more network nodes of the RAN 700.

[0238] Responsive to the temporal changes of the power grid parameter, the network entity 200 may transmit 410 a control message to the one or more network nodes 100 of the RAN and the network node 100 may receive 301 the control a message. The network node 100 may receive 301 the control message from the network entity 200. The communication system 700 may comprise one or more network entities 200, e.g., one network entity may be embodied in a cloud system (e.g., a distributed server network) and / or in the core node of the RAN 700.

[0239] Multiple network entities 200 may be in communication, for example, exchanging NES configurations and / or the historical data to improve the machine leaning.

[0240] Fig. 8A shows an exemplary schematic diagram of a change in one exemplary parameter of the power grid network 500 over time, e.g., according to the second rate. The exemplary parameter of the grid 500 is shown to be correlated with the energy price (e.g., SEK / MWh) provided by the grid 500 as an indicator of emissions, wherein environmental costs are factored in the price 804. Therefore, embodiments of the methods 300 and 400 can reduce, e.g. CO2, emissions.

[0241] Fig. 8B shows an exemplary schematic diagram of a change in an exemplary parameter of the power grid network 500 over time, e.g., according to a first rate. Here, the exemplary parameter of the grid 500 is the electrical frequency (in Hz) provided by the grid 500. As has been shown, the provided electrical frequency may vary over time from the nominated frequency (e.g., 50 Hz in this example). The change in the parameter of the grid 500 may trigger the configuration (e.g., reconfiguration) of the at least one NES features of the network node 100.

[0242] The first rate of obtaining 302 or 404 (e.g., sampling) the short-term scale may be, e.g. 10 s. The power grid parameter may be the instantaneous power grid voltage frequency 806. When the instantaneous consumption exceeds instantaneous production, the grid frequency 806 drops as energy is being drained from the flywheels on the distribution system. Conversely, the frequency 806 may increase if production exceeds the consumption. The grid operator may provide instantaneous incentives to large energy users for rapid adaptation of instantaneous consumption, e.g. negative prices during intervals where the frequency 806 increases above the reference value or penalties during frequency 806 drops due to aggregate over-consumption.

[0243] The Fig. 8B depicts an example instantaneous frequency fluctuation around the 50 Hz reference during a one hour, where the deviation approaches a tolerance limit of 0.2 %. Around the time mark 15:02 (3:02 PM), the at least one NES feature is activated responsive to the decrease in the power grid parameter 806.

[0244] Minimizing the power and energy consumption of a radio BS (e.g., gNB 100), or more generally, improving its energy performance, has been a long-standing objective. Several key features of the 5G NR system solution, e.g. the lean signaling architecture with 20 ms Synchronization Signal Block (SSB) period or dynamic Secondary Cell (Scell) activation / deactivation, were introduced to minimize unnecessary gNB energy consumption. In a 3GPP Release 18, Network Energy Saving (NES) Work Item (WI), various technical improvements are being introduced to further improve the NW energy performance using techniques in e.g. spatial, power, and frequency domains. The energy consumption of a gNB may vary widely, 2-1 Ox, depending on its selected NES feature configuration. Any of the afore- mentioned examples may be used for implementing the at least one NES feature in the present technique.

[0245] Optionally, the technique may be coupled to an Intent Based Automation (IBA). For example, it has become highly complex to operate the RAN 700 efficiently, and automated operational excellence is becoming ever more important. Rather than using human operation personnel for tuning the RAN 700, especially for complex dynamic scenarios, Intent- driven / based techniques is a level of automation used with possibility to automate with one or more KPIs used by the subject technique. IAB can balance on requirements while fulfilling a certain KPI compared to traditional methods. Intent-based operation relies on an autonomous and self-organizing platform that provides the capability needed to operate an intent.

[0246] Embodiments of the technique can address at least one of the following prior art problems. Existing gNB energy performance improvement solutions typically pursue absolute energy consumption minimization, subject to maintaining a predetermined minimum necessary level of NW key performance indicators (KPIs), like user or system throughput or NW access latency. However, since energy or electricity parameters (which may correlate with prices) may vary over time, such strategy may not provide overall system optimization in terms of jointly considering power grid stability (and, given correlation, RAN operation costs) and RAN performance.

[0247] The conventional techniques according to the documents CN113873627B and CN111242326A teach an intelligent turn-off energy-saving operation and maintenance system for 5G communication base station propose energy-saving approaches using different criteria to determine whether certain gNBs or certain RATs (e.g., 5G) may be turned off or should remain active. These criteria may include the need to maintain NW functionality (e.g. NB-IoT service), energy cost, KPIs or active service types. However, turning on and off gNBs is a blunt tool that can dramatically affect the user experience and cannot be used dynamically. It is therefore not a suitable tool for rapid and gradual energy consumption adaptation.

[0248] The conventional technique energy-saving method for 5G base station taught in the document CN112566226B is proposed to maximize energy performance by balancing energy consumption and performance while considering network KPIs. According to the network type, gNB configuration, load type and coverage characteristics the energy saving strategy is determined and energy-saving parameters adjusted based on real-time KPI monitoring ensuring that the network performance is basically not affected. However, following predetermined KPIs as criteria may lead to suboptimal operation of the power grid network 500 (e.g., correlated with suboptimal cost / performance trade-offs) or can even add to destabilizing the power grid network 500.

[0249] To leverage rapidly varying power grid parameters (which may correlate with electricity prices), including instantaneous situations of excess of electricity (which may correlate with instantaneously negative prices instants), backup battery charging / discharging pattern can be controlled at RAN nodes to contribute to grid frequency stabilization according to the document WO2021 / 148127A1 and the nodes may decide when to charge and discharge the batteries based on electricity price, with the objective to minimize the operational cost. However, the battery capacity, and thus the extent and the duration of the achievable impact, is limited and using a considerable fraction of the capacity for total cost optimization also reduces the usability of the batteries for their actual intended purpose - providing maximal extended lifetime in case of an actual power failure.

[0250] In contrast, embodiments of the methods 300 and 400 enable the network node 200 (e.g., RAN nodes such gNBs) to scalably and gradually contribute to power grid stability (which may correlate with reduced energy-related expenses) over a time period with varying power grid parameters (which may or may not be related to electricity prices) while maintaining reasonable user- and NW-related KPIs, and without compromising backup energy availability.

[0251] Fig. 9A shows a first example of a schematic diagram of a performance indicator 901 (PI), e.g., any one of the 3GPP-defined key performance indicators (KPIs), of at least one network node 200 of the RAN 700. More specifically, the PI target value 910 of the PI is a function of time responsive to changes in at least one parameter of the power grid network, e.g. any one or a combination of the parameters 804 and 806.

[0252] The network node 100 performs in a best effort mode close to the PI capability value 902 of the network node, when the power grid parameter is indicative of excess power in the power grid network 500. Only in temporal phases when the power grid parameter is indicative of a shortage of power, the at least one NES feature is configured (e.g. activated).

[0253] In one variant of any embodiment, which is schematically illustrated in Fig. 9A, the PI target value 910 sets a lower performance limit 908 in the PI space of the network node 100, e.g. a 95%-quantile for the PI current value 906 not falling below PI target value 910.

[0254] In another variant of any embodiment, which is schematically illustrated in Fig. 9B, the PI target value 910 sets an upper performance limit 904 in the PI space of the network node 100, e.g. a 95%-quantile for the PI current value 906 not exceeding below PI target value 910.

[0255] In any variant, the lower performance limit 908 may correspond to a QoS requirement of the radio device served by the network node 100.

[0256] As indicated in each of the Figs. 9A and 9B, the transition (in either direction) between the activated NES feature and the deactivated NES feature may occur driven by the event 912, such as switching between different types power plants in the power grid network 500 and / or when a threshold as to the CO2 emissions per energy unit is exceeded.

[0257] Fig. 10 shows a schematic block diagram for an embodiment of the device 100. The device 100 comprises processing circuitry, e.g., one or more processors 1004 for performing the method 300 and memory 1006 coupled to the processors 1004. For example, the memory 1006 may be encoded with instructions that implement at least one of the modules 104 and 110 or any other module functionally disclosed herein.

[0258] The one or more processors 1004 may be a combination of one or more of a microprocessor, controller, microcontroller, central processing unit, digital signal processor, application specific integrated circuit, field programmable gate array, or any other suitable computing device, resource, or combination of hardware, microcode and / or encoded logic operable to provide, either alone or in conjunction with other components of the device 100, such as the memory 1006, network node functionality. For example, the one or more processors 1004 may execute instructions stored in the memory 1006. Such functionality may include providing various features and steps discussed herein, including any of the benefits disclosed herein. The expression "the device being operative to perform an action" may denote the device 100 being configured to perform the action.

[0259] As schematically illustrated in Fig. 10, the device 100 may be embodied by a network node 1000, e.g., functioning as a gNB. The network node 1000 comprises a radio interface 1002 coupled to the device 100 for radio communication with one or more radio devices, e.g., functioning as a UE.

[0260] In view of the embodiments disclosed herein, the skilled person appreciates that the (e.g., key) PI target values may be based on the at least one power grid parameter in various ways. In the context of adapting the configuration (e.g., activation) of the Network Energy Saving (NES) features in a gNB 100 (or other Radio Access Network nodes 100) based on the status of the power grid network 500, the Key Performance Indicators (KPIs) are affected by power grid parameters indirectly, as operational objectives of the RAN 700 are aligned with the current energy system status.

[0261] The KPI target value 910 may be influenced by the at least one power grid parameter in at least one of the following ways.

[0262] 1. Grid Frequency Deviations: When the power grid frequency deviates from its nominal value (e.g., 50 Hz or 60 Hz, depending on the region), it signals an imbalance between electricity production and consumption. The method 300 and 400 can respond to such incentives from the grid 500 to either consume more energy when there's an excess (frequency above nominal) or reduce consumption when there's a shortage (frequency below nominal). This can lead to the adaptation of gNB operations, which may affect network KPIs like throughput, latency, and connectivity, yet to a limited extent at the benefit of long term grid stability.

[0263] 2. Spot Prices and Demand Response: Electricity spot prices vary with supply and demand. For instance, prices could be higher during peak usage times. The gNBs 100 can adjust their activity during these periods by leveraging NES features. For example, during higher price periods, a gNB 100 may be configured to be more aggressive with energy savings at the expense of some KPIs.

[0264] 3. Load Management and Grid Stability: The grid 500 may request the RAN 700 to reduce their load at critical times to maintain grid stability in the steps 302 or 404. Such requests can prompt the methods 300 and 400 to activate NES features more aggressively to lower energy consumption, which might result in temporarily lower KPI performance at the benefit of long term grid stability.

[0265] 4. Power Grid Cooperation Mechanisms: Some power grids offer mechanisms for power loads that offer flexibility in their electricity consumption. For example, the RAN 700 may curtail energy usage even without drawing upon backup power sources (like batteries) during periods of peak demand. Embodiments can comply with these cooperation mechanism, e.g., even without the aging caused to batteries. 5. Renewable Energy Availability: If a gNB 100 is connected to a power source heavily dependent on renewable energy, such as solar or wind, its NES configuration may be synchronized with the renewable energy's intermittent nature. During high renewable energy production periods, a gNB 100 might operate at higher power usage levels without aggressive NES features, thus maintaining or improving KPIs. Conversely, when renewable energy is scarce, more aggressive NES features may be configured according to the methods 300 and 400.

[0266] Fig. 11 shows a schematic block diagram for an embodiment of the device 200. The device 200 comprises processing circuitry, e.g., one or more processors 1104 for performing the method 400 and memory 1106 coupled to the processors 1104. For example, the memory 1106 may be encoded with instructions that implement at least one of the modules 204 and 210.

[0267] The one or more processors 1104 may be a combination of one or more of a microprocessor, controller, microcontroller, central processing unit, digital signal processor, application specific integrated circuit, field programmable gate array, or any other suitable computing device, resource, or combination of hardware, microcode and / or encoded logic operable to provide, either alone or in conjunction with other components of the device 200, such as the memory 1106, network entity functionality. For example, the one or more processors 1104 may execute instructions stored in the memory 1106. Such functionality may include providing various features and steps discussed herein, including any of the benefits disclosed herein. The expression "the device being operative to perform an action" may denote the device 200 being configured to perform the action.

[0268] As schematically illustrated in Fig. 11, the device 200 may be embodied by a network entity 1100, e.g., functioning as a core network node. The network entity 1100 comprises a backhaul or midhaul interface 1102 coupled to the device 200 for (e.g., radio, e.g. microwave) communication with one or more network nodes 1000, e.g., functioning as gNB or DU thereof.

[0269] With reference to Fig. 12, in accordance with an embodiment, a communication system 1200 includes a telecommunication network 1210, such as a 3 GPP -type cellular network, which comprises an access network 1211, such as a radio access network, and a core network 1214. The access network 1211 comprises a plurality of base stations 1212a, 1212b, 1212c, such as NBs, eNBs, gNBs or other types of wireless access points, each defining a corresponding coverage area 1213a, 1213b, 1213c. Each base station 1212a, 1212b, 1212c is connectable to the core network 1214 over a wired or wireless connection 1215. A first user equipment (UE) 1291 located in coverage area 1213c is configured to wirelessly connect to, or be paged by, the corresponding base station 1212c. A second UE 1292 in coverage area 1213a is wirelessly connectable to the corresponding base station 1212a. While a plurality of UEs 1291, 1292 are illustrated in this example, the disclosed embodiments are equally applicable to a situation where a sole UE is in the coverage area or where a sole UE is connecting to the corresponding base station 1212.

[0270] Any of the base stations 1212 may embody the device 100. Alternatively or in addition, the core network 1214 may embody the device 200. The telecommunication network 1210 is itself connected to a host computer 1230, which may be embodied in the hardware and / or software of a standalone server, a cloud-implemented server, a distributed server or as processing resources in a server farm. The host computer 1230 may be under the ownership or control of a service provider, or may be operated by the service provider or on behalf of the service provider. The connections 1221, 1222 between the telecommunication network 1210 and the host computer 1230 may extend directly from the core network 1214 to the host computer 1230 or may go via an optional intermediate network 1220. The intermediate network 1220 may be one of, or a combination of more than one of, a public, private or hosted network; the intermediate network 1220, if any, may be a backbone network or the Internet; in particular, the intermediate network 1220 may comprise two or more subnetworks (not shown).

[0271] The communication system 1200 of Fig. 12 as a whole enables connectivity between one of the connected UEs 1291, 1292 and the host computer 1230. The connectivity may be described as an over-the-top (OTT) connection 1250. The host computer 1230 and the connected UEs 1291, 1292 are configured to communicate data and / or signaling via the OTT connection 1250, using the access network 1211, the core network 1214, any intermediate network 1220 and possible further infrastructure (not shown) as intermediaries. The OTT connection 1250 may be transparent in the sense that the participating communication devices through which the OTT connection 1250 passes are unaware of routing of uplink and downlink communications. For example, a base station 1212 need not be informed about the past routing of an incoming downlink communication with data originating from a host computer 1230 to be forwarded (e.g., handed over) to a connected UE 1291. Similarly, the base station 1212 need not be aware of the future routing of an outgoing uplink communication originating from the UE 1291 towards the host computer 1230.

[0272] By virtue of the method 200 being performed by any one of the base stations 1212, the performance or stability of the OTT connection 1250 can be improved, e.g., by avoiding power outages of individual base station or the entire power grid network. More specifically, the host computer 1230 may indicate to the RAN 700 or the base station 1212 (e.g., on an application layer) a QoS of traffic, which may set the lower performance limit when the at least one NES features is activated.

[0273] As has become apparent from above description, at least some embodiments of the technique enable NW nodes (e.g., gNBs) to adapt their NES mode selection to reduce the risk of power grid instability (which may relate with overall energy cost reduction) for the RAN nodes (e.g., gNBs), reducing OPEX, and / or assisting the power grid network to remain stable, reducing the probability of black-outs or brown-outs during times of energy shortages. Same or further embodiments achieve this while still ensuring acceptable NW performance in terms of the one or more Pls or KPIs, over extended periods, and by adjusting actual operating power without compromising backup power availability.

[0274] Embodiments of the technique can be inherently robust in that the energy consumption optimization is never stretched beyond a point where the KPI current values are no longer meet KPI requirement values. Moreover, the KPI current values need not be fixed. Rather, the KPI targets values themselves may depend on the power grid parameter (which may correlate with price signals). E.g. for situations where the instantaneous CO2 emissions or cost is very high, the methods may apply KPI target values (e.g., predefined already ahead of such an event) that define lower KPI targets. Same or further embodiments can implement the technique as a control mechanism that is fully under the operator’s control and there will be no surprising KPI failures.

[0275] Many advantages of the present invention will be fully understood from the foregoing description, and it will be apparent that various changes may be made in the form, construction and arrangement of the units and devices without departing from the scope of the invention and / or without sacrificing all of its advantages. Since the invention can be varied in many ways, it will be recognized that the invention should be limited only by the scope of the following claims.

Claims

Claims1. A method (300) of configuring a network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, the method (300) being performed by the network node (100; 1000; 1212) connected to a power grid network (500), the method (300) comprising: obtaining (304) one or more target values (910) of one or more performance indicators, Pls (901), of the network node (100; 1000; 1212), wherein the one or more PI target values (910) are dependent on at least one parameter (802) of the power grid network (500); and configuring (310) at least one NES feature of the network node (100; 1000; 1212) based on the one or more obtained (304) PI target values (910).

2. The method (300) of claim 1, wherein the one or more obtained (304) PI target values (910) set a lower performance limit (908) on the one or more Pls (901) of the network node (100; 1000; 1212), optionally wherein the at least one NES feature of the network node (100; 1000; 1212) is configured (310) to operate closer to each of the one or more PI target values (910) than a corresponding PI capability value of the network node (100; 1000; 1212).

3. The method (300) of claim 1 or 2, wherein the one or more obtained (304) PI target values (910) set an upper performance limit (904) on the one or more Pls (901) of the network node (100; 1000; 1212), optionally wherein the one or more PI target values (910) are greater than one or more corresponding PI requirement values of a Quality of Service, QoS, requirement.

4. The method (300) of any one of claims 1 to 3, further comprising: obtaining (302), optionally measuring, the at least one parameter (802) of the power grid network (500).

5. The method (300) of any one or claims 1 to 4, further comprising: determining (306) the at least one NES feature of the network node (100; 1000; 1212) based on the one or more obtained (304) PI target values (910).

6. The method (300) of any one of claims 1 to 5, wherein the configuring (310) of the at least one NES feature of the network node (100; 1000; 1212) comprises at least one of: activation or deactivation of the NES, optionally activation or deactivation of the at least one NES feature; activation or deactivation of one or more cells served by the network node (100; 1000; 1212); changing a modulation scheme of the network node (100; 1000; 1212); changing a transmission power of the network node (100; 1000; 1212);changing a bandwidth of the network node (100; 1000; 1212); changing a number of antennas used by the network node (100; 1000; 1212); activating or deactivating a multiple input-multiple output, MEMO, operation of the network node (100; 1000; 1212); changing a rank of a MIMO transmission or MIMO reception of the network node (100; 1000; 1212); switching between a single-user MEMO, SU-MIMO, operation and multi-user MIMO, MU-MEMO, operation of the network node (100; 1000; 1212); and increasing or reducing a clock rate of one or more processors of the network node (100; 1000; 1212).

7. The method (300) of any one of claims 1 to 6, wherein the at least one parameter of the power grid network (500) comprises at least one of: a power demand in the power grid network (500); a power supplied to the power grid network (500); a type of power plant (912) supplying the power grid network (500); an alternating current frequency, AC frequency (806), of the power grid network (500); a voltage of the power grid network (500); an amount of carbon dioxide (914) emitted into the atmosphere per energy unit provided by the power grid network (500); and a price (804) per energy unit of the power grid network (500).

8. The method (300) of any one of claims 1 to 7, wherein the configuring (310) of the at least one NES feature comprises: activating the at least one NES feature or configuring the at least one NES feature with stricter limits on power consumption; or deactivating the at least one NES feature or configuring the at least one NES feature with higher limits on power consumption.

9. The method (300) of any one of claims 1 to 8, wherein the power grid network (500) comprises two or more power grids (500-a; 500-b; 500-c).

10. The method (300) of any one of claims 1 to 9, further comprising measuring (308) at least one of: a current power consumption of the network node (100; 1000; 1212); and one or more current values (906) of the one or more Pls (906) of the network node (100; 1000; 1212).

11. The method (300) of claim 10, wherein the configuration (310) for the at least one NES feature of the network node (100; 1000; 1212) controls or influences at least one of:the measured (308) power consumption of the network node (100; 1000; 1212); and the one or more measured (308) current values (906) of the one or more Pls of the network node (100; 1000; 1212).

12. The method (300) of any one of claims 1 to 11, wherein the obtaining (304) of the one or more PI target values and / or the configuring (310) of the at least one NES feature of the network node (100; 1000; 1212) is performed using an artificial intelligence, Al, model, optionally according to intent-based networking, IBN.

13. The method (300) of any one of claims 1 to 12, further comprising: adjusting (312) the configured (310) at least one NES feature of the network node (100; 1000; 1212) based on the measured (308) current power consumption of the network node (100; 1000; 1212) and / or the measured (308) current value (906) of the one or more Pls (901) of the network node (200; 1100; 1214).

14. The method (300) of any one of claims 1 to 13, further comprising transmitting (314) a feedback message to a network entity (200; 1100; 1214), wherein the feedback message is indicative of at least one of: a measured (308) current power consumption of the network node (100; 1000; 1212); one or more measured (308) PI current values (906) of the one or more Pls (901) of the RAN (700); the obtained (302) at least one parameter (802) of the power grid network (500); the determined (306) at least one NES feature of the network node (100; 1000; 1212); the configuration (310) of at least one NES features of the network node (100; 1000;1212) based on the obtained (304) one or more PI target values (910); and the obtained (304) one or more PI target values (910).

15. The method (300) of any one of claims 1 to 14, further comprising receiving (301) a control message from the network entity (200; 1100; 1214), wherein the control message is indicative of at least one of: the at least one parameter (802) of the power grid network (500); the one or more PI target values (910) of the network node (100; 1000; 1212); and the configuration (310) of the at least one NES feature of the network node (100; 1000; 1212) depending on the at least one parameter (802) of the power grid network (500).

16. The method (300) of any one of claims 1 to 15, wherein the one or more Pls of the network node (100; 1000; 1212) comprises at least one of: an access delay measured in one or more radio devices (600); a data serving latency measured in one or more radio devices (600); a system transmission power, TP, of the network node (100; 1000; 1212);a data rate of a radio connection to one or more radio devices (600); and a data throughput of one or more radio devices (600).

17. The method (300) of any one of claims 1 to 16, wherein the obtained (304) one or more PI target values (910) are further based on at least one of: a number or load of one or more radio device (600) served by the network node (100; 1000; 1212); a service currently provided by the network node (100; 1000; 1212); a quality requirement of radio devices (600) served by the RAN (700); a future service to be provided by the network node (100; 1000; 1212); and a capability of one or more UEs served by the network node (100; 1000; 1212).

18. The method (300) of any one of claims 1 to 17, wherein the configuring (310) of the at least one NES feature of the network node (100; 1000; 1212) is performed at a first rate and at a second rate that is slower than the first rate.

19. The method (300) of any one of claims 1 to 18, wherein the obtaining (304) of the one or more PI target values (910) is performed periodically and / or triggered based on at least one of: the configuration (310) of the at least one NES feature of the network node (100; 1000; 1212); a change in the at least one parameter (802) of the power grid network (500); a change in the measured current power consumption of the network node (100; 1000;1212); and a control message received (301) from the network entity (200; 1100; 1214).

20. The method (300) of any one of claims 1 to 19, wherein the one or more PI target values fulfil a performance criterion of a QoS Class Identifier, QCI, or 5G QoS Identifier, 5QI, associated with the network node (100; 1000; 1212).

21. A method (400) of configuring at least one network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, the method (400) being performed by a network entity (200; 1100; 1214) of the RAN (700) connected to a power grid network (500), the method (400) comprising: obtaining (404) at least one parameter of the power grid network (500); and transmitting (410) a control message to the least one network node (100; 1000; 1212), wherein the control message is indicative of at least one of: one or more target values (910) of one or more performance indicators, Pls (901), of the RAN (700) for configuring at least one NES feature of the respective network node (100; 1000; 1212), the one or more PI target values being dependent on the obtained (404) at least one parameter (802) of the power grid network (500), anda configuration of at least one NES feature of the respective network node (100; 1000; 1212), the configuration being dependent on the obtained (404) at least one parameter (802) of the power grid network (500).

22. The method (400) of claim 21, further comprising at least one of determining (406) the one or more target values (910) of the one or more Pls (901)of the RAN (500) for the at least one network node (100; 1000; 1212), wherein the one or more PI target values (910) are based on the obtained (404) at least one parameter (802) of the power grid network (500), optionally wherein the configuration of the at least one NES feature of the respective network node (100; 1000; 1212) is dependent on the obtained (404) at least one parameter (802) of the power grid network (500) by determining (406) the configuration based on the determined (406) one or more PI target values (910); and determining (408) the configuration based on the obtained (404) at least one parameter of the power grid network (500).

23. The method (400) of claim 22, wherein the determining (406) of the one or more PI target values (910) for the at least one network node (100; 1000; 1212) is further based on the at least one of a received (402) feedback message from the at least one network node (100; 1000; 1212); and operational information of the RAN (700).

24. The method (400) of any one of claims 22 or 23, wherein the one or more PI target values (910) are dynamically determined (406) in real-time based on a predictive model that is trained by a machine learning technique to anticipates fluctuations in the at least one power grid parameter (802) using historical data, and current sensor data.

25. The method (400) of any one of claims 21 to 24, further comprising receiving (402) a feedback message from the at least one network node (100; 1000; 1212), wherein the feedback message is indicative of at least one of a measured (308) current power consumption of the network node (100; 1000; 1212); one or more measured (308) current values (906) of the one or more Pls (901); the obtained (404) at least one parameter of the power grid network (500), optionally measured (302) at the network node (100; 1000; 1212); the at least one NES feature of the network node (100; 1000; 1212) determined (306) by the network node (100; 1000; 1212) based on the one or more transmitted (410) PI target values (910); the configuration (310) of the at least one NES features of the network node (100; 1000; 1212) based on the one or more transmitted ( 16) PI target values.

26. The method (400) of any claims 21 to 25, wherein at least one of the transmitted (410) control message and the received (402) feedback message are communicated between the respective network node (100; 1000; 1212) and the network entity (200; 1100; 1214) through a midhaul interface, optionally an Fl interface, and / or wherein the network node (100; 1000; 1212) and the network entity (200; 1100; 1214) are functionally split into a distributed unit, DU, and a central unit, CU, respectively.

27. A computer program product comprising program code portions for performing the steps of any one of the claims 1 to 20 or 21 to 26 when the computer program product is executed on one or more computing devices (1004; 1104), optionally stored on a computer-readable recording medium (1006; 1106).

28. A network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, the network node (100; 1000; 1212) comprising memory (1006) operable to store instructions and processing circuitry (1004) operable to execute the instructions, such that the network node (100; 1000; 1212) is operable to: obtain (304) one or more target values (910) of one or more performance indicators, Pls (901), of the network node (100; 1000; 1212), wherein the one or more PI target values (910) are dependent on at least one parameter (802) of the power grid network (500); and configure (310) at least one NES feature of the network node (100; 1000; 1212) based on the one or more obtained (304) PI target values (910).

29. The network node (100; 1000; 1212) of claim 28, further comprising the features or operable to perform the steps of any one of claims 2 to 20.

30. A network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, the network node (100; 1000; 1212) being configured to: obtain (304) one or more target values (910) of one or more performance indicators, Pls (901), of the network node (100; 1000; 1212), wherein the one or more PI target values (910) are dependent on at least one parameter (802) of the power grid network (500); and configure (310) at least one NES feature of the network node (100; 1000; 1212) based on the one or more obtained (304) PI target values (910).

31. The network node (100; 1000; 1212) of claim 30, further comprising the features or configured to perform the steps of any one of claims 2 to 20.

32. A network entity (200; 1100; 1214) for configuring at least one network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, the network entity (200; 1100; 1214) comprising memory (1106) operable to store instructions and processing circuitry (1104) operable to execute the instructions, such that the network node (200; 1100; 1214) is operable to: obtain (404) at least one parameter of the power grid network (500); and transmit (410) a control message to the least one network node (100; 1000; 1212), wherein the control message is indicative of at least one of: one or more target values (910) of one or more performance indicators, Pls (901), of the RAN (700) for configuring at least one NES feature of the respective network node (100; 1000; 1212), the one or more PI target values being dependent on the obtained (404) at least one parameter (802) of the power grid network (500), and a configuration of at least one NES feature of the respective network node (100; 1000; 1212), the configuration being dependent on the obtained (404) at least one parameter (802) of the power grid network (500).

33. The network entity (200; 1100; 1214) of claim 32, further comprising the features or operable to perform any one of the steps of any one of claims 22 to 26.

34. A network entity (200; 1100; 1214) for configuring at least one network node (100; 1000; 1212) of a radio access network, RAN (700), for network energy savings, NESs, the network entity (200; 1100; 1214) being configured to: obtain (404) at least one parameter of the power grid network (500); and transmit (410) a control message to the least one network node (100; 1000; 1212), wherein the control message is indicative of at least one of: one or more target values (910) of one or more performance indicators, Pls (901), of the RAN (700) for configuring at least one NES feature of the respective network node (100; 1000; 1212), the one or more PI target values being dependent on the obtained (404) at least one parameter (802) of the power grid network (500), and a configuration of at least one NES feature of the respective network node (100; 1000; 1212), the configuration being dependent on the obtained (404) at least one parameter (802) of the power grid network (500).

35. The network entity (200; 1100; 1214) of claim 34, further comprising the features or configured to perform the steps of any one of claim 22 to 26.

Images