Antenna panel activation

Abstract

According to an example aspect of the present invention, there is provided an apparatus configured to obtain, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face, select, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and direct the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

Description

ANTENNA PANEL ACTIVATIONFIELD

[0001] The present disclosure relates to management of multiple antenna panels in a user equipment of a cellular telecommunication system.BACKGROUND

[0002] User equipments in cellular communication systems may be furnished with capability to receive and transmit in a directional manner, rather than omnidirectionally. This provides a benefit in terms of energy efficiency, since wireless transmissions may be sent to a direction of the intended receiver, avoiding transmitting interference into other directions. Likewise a wireless receiver may be provided with directionality, to enable reception of wireless signals from the intended transmitter, such that the receiver admits less noise from other directions.

[0003] In multipath radio channels, the relevant transmission may be provided into more than one direction, and the receiving of the transmission may be performed from more than one direction. Beamforming or multi-antenna panel transceivers may be used to accomplish directionality in the transmit or receive side.SUMMARY

[0004] According to some aspects, there is provided the subject-matter of the independent claims. Some embodiments are defined in the dependent claims. The scope of protection sought for various embodiments of the invention is set out by the independent claims. The embodiments, examples and features, if any, described in this specification that do not fall under the scope of the independent claims are to be interpreted as examples useful for understanding various embodiments of the invention.

[0005] According to a first aspect of the present disclosure, there is provided an apparatus comprising at least one processing core and at least one memory storing instructions that, when executed by the at least one processing core, cause the apparatus at least to obtain, from a network, antenna panel selection information which associates at least one geographic location with directions of arrival of wireless signals from the network, select, based on a current geographic location of the apparatus and on the antenna panel selection information, one or more antenna panels of a user equipment, and direct the user equipment to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0006] Various embodiments of the first aspect may comprise one or more features from the following bulleted list:• the antenna panel selection information comprises at least one of at least one measurement report, at least one set of angles of arrival of the wireless signals, a machine learning model, or antenna panel activation rate information for the one or more antenna panels of the user equipment.• the antenna panel selection information comprises the at least one set of angles of arrival of the wireless signals, wherein each of the at least one set of angles of arrival of the wireless signals comprises at least one of a mean angle of arrival or a standard deviation of angle of arrival.• each of the at least one set of angles of arrival of the wireless signals comprises plural angles of arrival of the wireless signals.• the apparatus is further configured to perform a measurement using plural of the antenna panels of the user equipment, to compile and to send a measurement report which comprises the current location of the apparatus, and angular information specifying directions from which the wireless signals from the network were determined to arrive in the measurement.

[0007] According to a second aspect of the present disclosure, there is provided an apparatus comprising at least one processing core and at least one memory storing instructions that, when executed by the at least one processing core, cause the apparatus at least to obtain, from plural user equipments, measurement reports which comprise current locations of the respective user equipments, and angular information specifying directions from which wireless signals from a network were determined to arrive in the respective user equipments, and provide, to a first user equipment, antenna panel selectioninformation which associates at least one geographic location with directions of arrival of the wireless signals from the network, wherein the apparatus is configured to obtain the antenna panel selection information based on the measurement reports.

[0008] According to a third aspect of the present disclosure, there is provided a method comprising obtaining, in an apparatus and from a network, antenna panel selection information which associates at least one geographic location with directions of arrival of wireless signals from the network, selecting, based on a current geographic location of the apparatus and on the antenna panel selection information, one or more antenna panels of a user equipment, and directing the user equipment to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0009] According to a fourth aspect of the present disclosure, there is provided a method comprising obtaining, from plural user equipments, measurement reports which comprise current locations of the respective user equipments, and angular information specifying directions from which wireless signals from a network were determined to arrive in the respective user equipments, and providing, to a first user equipment, antenna panel selection information which associates at least one geographic location with directions of arrival of the wireless signals from the network, wherein the method further comprises obtaining the antenna panel selection information based on the measurement reports.

[0010] According to a fifth aspect of the present disclosure, there is provided an apparatus comprising means for obtaining, in an apparatus and from a network, antenna panel selection information which associates at least one geographic location with directions of arrival of wireless signals from the network, selecting, based on a current geographic location of the apparatus and on the antenna panel selection information, one or more antenna panels of a user equipment, and directing the user equipment to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0011] According to a sixth aspect of the present disclosure, there is provided an apparatus comprising means for obtaining, from plural user equipments, measurement reports which comprise current locations of the respective user equipments, and angular information specifying directions from which wireless signals from a network were determined to arrive in the respective user equipments, and providing, to a first userequipment, antenna panel selection information which associates at least one geographic location with directions of arrival of the wireless signals from the network, and for obtaining the antenna panel selection information based on the measurement reports.

[0012] According to a seventh aspect of the present disclosure, there is provided a non-transitory computer readable medium having stored thereon a set of computer readable instructions that, when executed by at least one processor, cause an apparatus to at least obtain, from a network, antenna panel selection information which associates at least one geographic location with directions of arrival of wireless signals from the network, select, based on a current geographic location of the apparatus and on the antenna panel selection information, one or more antenna panels of a user equipment, and direct the user equipment to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0013] According to an eighth aspect of the present disclosure, there is provided a non-transitory computer readable medium having stored thereon a set of computer readable instructions that, when executed by at least one processor, cause an apparatus to at least obtain, from plural user equipments, measurement reports which comprise current locations of the respective user equipments, and angular information specifying directions from which wireless signals from a network were determined to arrive in the respective user equipments, and provide, to a first user equipment, antenna panel selection information which associates at least one geographic location with directions of arrival of the wireless signals from the network, and to obtain the antenna panel selection information based on the measurement reports.

[0014] According to a ninth aspect of the present disclosure, there is provided an apparatus comprising at least one processing core and at least one memory storing instructions that, when executed by the at least one processing core, cause the apparatus at least to obtain, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face, select, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, anddirect the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0015] According to a tenth aspect of the present disclosure, there is provided an apparatus comprising at least one processing core and at least one memory storing instructions that, when executed by the at least one processing core, cause the apparatus at least to obtain, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment, obtain a machine learning model, the model being obtained from the information specifying directions from which wireless signals from a network were determined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning model being configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and provide, to a first user equipment, the machine learning model.

[0016] According to an eleventh aspect of the present disclosure there is provided a method comprising obtaining, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face, selecting, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and directing the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0017] According to a twelfth aspect of the present disclosure, there is provided a method comprising obtaining, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment, obtaining a machinelearning model, the model being obtained from the information specifying directions from which wireless signals from a network were determined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning model being configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and providing, to a first user equipment, the machine learning model.

[0018] According to a thirteenth aspect of the present disclosure, there is provided a non-transitory computer readable medium having stored thereon a set of computer readable instructions that, when executed by at least one processor, cause an apparatus to at least obtain, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face, select, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and direct the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0019] According to a fourteenth aspect of the present disclosure, there is provided a non-transitory computer readable medium having stored thereon a set of computer readable instructions that, when executed by at least one processor, cause an apparatus to at least obtain, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment, obtain a machine learning model, the model being obtained from the information specifying directions from which wireless signals from a network were determined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning model being configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and provide, to a first user equipment, the machine learning model.

[0020] According to a fifteenth aspect of the present disclosure, there is provided an apparatus comprising means for obtaining, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face, selecting, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and directing the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0021] According to a sixteenth aspect of the present disclosure, there is provided an apparatus comprising means for obtaining, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment, obtaining a machine learning model, the model being obtained from the information specifying directions from which wireless signals from a network were determined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning model being configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and providing, to a first user equipment, the machine learning model.BRIEF DESCRIPTION OF THE DRAWINGS

[0022] FIGURE 1 illustrates an example system in accordance with at least some embodiments of the present invention;

[0023] FIGURE 2A illustrates radiative patterns of antenna panels of a UE;

[0024] FIGURE 2B illustrates an example multipath situation;

[0025] FIGURE 3 illustrates an example apparatus capable of supporting at least some embodiments of the present invention;

[0026] FIGURE 4 illustrates signalling in accordance with at least some embodiments of the present invention, and

[0027] FIGURE 5 is a flow graph of a method in accordance with at least some embodiments of the present invention.EMBODIMENTS

[0028] Processes are disclosed herein which enable the network to advise UEs concerning directions from which signals may arrive, in the current location of the UE. The UE may take this advice into account when selecting which antenna panels to activate for a measurement, such as a mobility measurement. The UE measurement may be a radio resource management, RRM, measurement which assesses signal strengths from a serving cell and one or more neighbouring cells. The UE, on the other hand, enabled the network to determine the advice by transmitting uplink data to the network, informing the network of directions from which the UE received wireless signals.

[0029] FIGURE 1 illustrates an example system in accordance with at least some embodiments of the present disclosure. This system includes a base station 130 in communication with UEs 110, 120. Radio link 131 connects base station 130 with UE 110, and radio link 132 connects base station 130 with UE 120. The radio links 131, 132 may be bidirectional, comprising an uplink, UL, to convey information from the respective UE 110, 120 toward the base station 130 and a downlink, DL, to convey information from the base station 130 toward the respective UE 110, 120.

[0030] Base station 130 is further coupled communicatively with core network node 140, which may comprise, for example, a mobility management entity, MME, or access and mobility management function, AMF. The core network node 140 may be coupled with further core network nodes, and with a network 150, which may comprise the Internet or a corporate network, for example. The system may communicate with further networks via network 150. Examples of the further core network nodes, which are not illustrated in FIGURE 1 for the sake of clarity, include gateways and subscriber information repositories.

[0031] Radio links 131, 132 may be in accordance with a long term evolution, LTE, technology, or a new radio, NR, technology, which is also known as fifth generation, 5G, for example. In particular, radio links 131, 132 comprise a set of subbands, each subband being a frequency resource in use between the base station and UE. There may be, for example, up to 12 or up to 18 subbands. There may be more than 18 subbands. The subbands in the set of subbands may form a contiguous block of subbands in frequency space, for example.

[0032] In FIGURE 1, radio link 131 is a so-called line-of-sight- LOS, link where UE 110 receives the signal from a single direction, which is the direction of base station 130 from the point of view of UE 110. On the other hand, in FIGURE 1 radio link 132 is a multipath link which comprises a LOS element as well as an element reflected from building 160. UE 120 thus receives radio link 132 from two directions, firstly the LOS direction and secondly from the reflected element, from the direction of building 160 from the point of view of UE 120. UEs 110, 120 may be configured to determine the directions from where they receive wireless signals from base station 130, or indeed from plural base stations. This may be performed as part of a receive beamforming process, or otherwise.

[0033] UEs 110, 120, or UEs in general, may be provided with more than one antenna panel. Individual antenna panels may be used for both transmission and reception, and individual antenna panels transmit and receive to / from a direction which the antenna panel faces, broadly speaking. Thus placing the antenna panels on different sides of a UE, a solution may be obtained which is capable of selectively transmitting to, and receiving from, different directions using different antenna panels. This will be explored in more detail herein below.

[0034] With the use of mmWave in 3rdGeneration Partnership Project, 3GPP, 5G NR, the need to compensate for additional path loss at higher frequencies leads to proposals to use antenna arrays at the base stations and UEs. Patch arrays for mmWave at UE level are clearly directional with up to a 30 dB front-to-back sensitivity ratio. Thus using multiple array antenna panels covering multiple spatial directions becomes an attractive choice.

[0035] FIGURE 2A illustrates radiative patterns of antenna panels of a UE. Like numbering denotes like structure as in FIGURE 1. Here UE 110 has three antenna panels 201, 203 and 205 on distinct sides of UE 110. Panel 201 has a radiative pattern 201R,panel 203 has a radiative pattern 203R, and panel 205 has a radiative pattern 205R. The radiative patterns may overlap in places, however overall the radiative patterns are directional in that each antenna panel 201, 203, 205 is most sensitive, and transmits most strongly, in the forward direction in front of the respective panel. It is of note, that the radiative patterns of antenna panels 201, 203, 205 co-rotate with UE 110 as UE 110 is rotated.

[0036] FIGURE 2B illustrates an example multipath situation. In the situation of FIGURE 2B, UE 110 has four antenna panels 201, 203, 205 and 207, attached to four different sides of UE 110, as illustrated. Two base stations 231, 233 transmit signals to UE 110. In particular, base station 231 transmits using wireless link 211, and base station 233 transmits using wireless link 213, 215. This latter wireless link is a multi-path wireless link, and the distinct paths comprised in the wireless link are here given distinct reference numbers, unlike the case in FIGURE 1.

[0037] UE 110 is capable of receiving wireless link 211 from base station 231 via antenna panel 205, and UE 110 is capable of receiving wireless link 213, 215 from base station 233 via antenna panels 205 and 203. In other words, UE 110 can receive the signals from base stations 231, 233 in this multipath propagation situation via antenna panels 203 and 205. Antenna panels 201 and 207 may be inactive since they cannot in this situation receive signals. It should be noted, that in real-life use cases multipath propagation may be more complex than that illustrated in FIGURE 2B, in particular there may be more reflections and more distinct wireless links, from more base stations or cells. Further, in case uplink and downlink use different frequencies, the multipath propagation channels may differ markedly between the uplink and downlink, wherefore a different set of antenna panels may be usable in the uplink and the downlink. The multipath propagation situation is dependent on the location of the UE within a coverage area of a serving cell, since as the UE moves, reflections and LOS wireless links become available and unavailable in dependence of the geometry of the surroundings of the UE. Although the rotation of the UE may be same, that is, the panels are orientated in a same way with regard to compass directions, the same UE may need to activate different panels in different mobility occasions for the measurements of the same cell, in different locations in the cell’s coverage area. This is due to reflectors and blocking objects that shape the propagation of the signals.

[0038] UEs need to perform measurements concerning signals transmitted in the downlink from nearby cells, to facilitate decisions on handovers and cell reselections, for example. Not all UEs are capable of using all their panels at the same time. For example, in a multipanel UE assumption 1, multiple panels are implemented in a UE and only one panel can be activated at a time, for UL and DL. In this case the UE has multiple antenna panels but only a single baseband processing chain to measure synchronization signal blocks, SSBs, on each of them. In a multipanel UE assumption 2, multiple panels are implemented in a UE and multiple panels can be activated at a time and one or more panels can be used for transmission and reception at the same time. In a multipanel UE assumption 2, multiple panels are implemented in a UE and multiple panels can be activated at a time for reception and measurements, since each of them has its own baseband processing chain, but only one panel can be used for transmission at any one time. These assumptions are defined to simplify the discussions and multi-panel UE, MPUE, capabilities in terms of measurements. However, each panel of an MPUE can be activated independently with different activation frequency. The activation periodicity of each panel determines the sampling rate of each panel, that is, how often a panel performs sampling of cell measurements over time. Thus to measure for signals arriving from all directions, the UE may need to perform plural measurements using its different antenna panels, one after another.

[0039] In battery-powered devices, such as many UEs, and also in non-battery- powered devices, conservation of energy is of interest. UE measurements are a key part of mobility management in cellular networks. UEs measure the quality of serving cell and neighbour cell-originated signals, and report to the network where those measurements are used by the network to decide on handover of the UE from one cell to another, for example. The reporting may be accomplished by transmitting a measurement report to the network. Inaccurate cell quality measurements lead to faulty handover decisions in the network and cause the UEs to experience service interruption, such as radio link failure, RLF, handover failure, HOF or ping-pong behaviour. On the other hand, frequent activations of antenna panels for measurements leads to a lot of energy consumption and causes battery drainage. Therefore, it is of interest for UEs to achieve accurate cell quality measurements and good mobility performance in appropriate balance with the power consumption.

[0040] If the UE is aware of the multipath characteristics of the wireless link(s), it can activate only the antenna panels necessary to perform cell quality measurements. However, if the UE doesn’t know the multipath characteristics of signals at its present location, the UE needs to activate all panels to monitor the received signal power of the neighbouring cells, as well as that of the serving cell and decide which panels are to be used for cell quality measurements. Therefore, in the absence of multipath information, the UE needs to activate all of its panels first to detect which ones of its panels receive signal, which leads more power consumption since merely activating a panel already consumes some power.

[0041] The challenge mentioned above is more pronounced in high-speed train and factory scenarios. In those scenarios, UE rotation, and thus panel rotation, is less dynamic relative to a direction of UE movement. In a high-speed train scenario, the UE traverses cell borders very often due to its fast geolocation change, and the observed UE rotation is almost stationary with respect to the cell border crossings. In a factory automation scenario automated guided vehicles which are network-connected move along predefined paths in a deterministic manner, without changing UE panel rotational state. In those cases, the UEs need to activate all panels to detect the received signal power from the serving and neighbouring cells even if the environment and UE panels are rotationally stationary and the UE receives the signal only from certain directions.

[0042] Thus, a UE with multiple antenna panels needs to activate, in the absence of multipath information, all of its panels to determine if any of the reference signal received power, RSRP, of any cell, either neighbour or serving cell, is received on each one of its panels. This leads drainage of the UE battery due to unnecessary panel activation.

[0043] Solutions are herein described, where a multi-panel UE is enabled to activate its antenna panels only when it is necessary, which improves UE power efficiency as well as maintains measurement accuracy. In detail, the UE observes the parameters that are related to spatio-temporal characteristics of the propagation channel and reports them to the network. The network collects these reported parameters and derives the spatiotemporal characteristics of the propagation channel of its coverage area, as a function of UE location, from reports from plural UEs in different locations. The network then provides UEs antenna panel selection information that is needed to activate only the necessary panels of the UEs for mobility measurements, for example. The disclosedmethod uses, in some embodiments, a machine learning, ML, approach to derive the spatio-temporal characteristics of the propagation channel. Both UE and network side ML approaches may be used in these embodiments of this method. Herein a ML model may include a statistical, computational, adaptable or trainable inference model.

[0044] In this method, the network configures the UE to collect measurements that are needed for deriving the spatio-temporal characteristics of the propagation channel, as a function of UE location. Those collected measurements are then reported by the UE to the network. The network may use a ML model, or a deterministic mathematical model, to model the spatio-temporal characteristics of the propagation channel, using the reported collected measurements. The spatio-temporal characteristics may be used to derive antenna panel selection information. Once the modeling is accomplished, the network provides the antenna panel selection information that the UEs need for achieving power efficient panel scheduling without sacrificing measurement accuracy. Hence the UE uses the network- provided antenna panel selection information and schedules its panel activation accordingly. The antenna panel selection information may comprise sampling rate or activation rate for each of the UE’s antenna panels, or for a subset of the UE’s antenna panels. The antenna panel selection information may alternatively be referred to as antenna panel information.

[0045] The network may determine a list of parameters that the UE should measure concerning the multipath environment along with the reference signal received power, RSRP, values to report the network for modeling the spatio-temporal characteristics of the multipath propagation environment. Alternatively, the UE may know the parameters in advance, and the network may merely trigger the collection and reporting of this parameter data.

[0046] Examples of parameters the UE may collect to facilitate determining the multipath propagation environment include an azimuth angle of arrival of the serving and neighbour cell measurements, a zenith angle of arrival of the serving and neighbor cell measurements, and coordinates of the UE when performing the measurement. The azimuth angle may be an angle about a normal of the surface, the normal pointing directly away from the Earth’s centre, wherein zero in azimuth angle may point north, for example. The zenith angle may correspond to an elevation from ground, with zero zenith correspondingto a signal arriving parallel to the ground and 90 degrees corresponding to a signal arriving directly downward, for example.

[0047] For example, the UE may determine the afore-mentioned parameters for each antenna panel, or for each multipath signal component arriving at the UE. Further, the network may be provided supplementary information, such as angular placement of the antenna panels of the UE, orthogonality information describing how independent the panels are from each other, and a timing advance parameter describing a distance from the UE to a base station controlling the serving cell. Angular placement and orthogonality of the antenna panels are UE-specific, or UE-type specific, parameters that affect the panel scheduling algorithm for different propagation characteristics. Reporting these along with the angle of arrival and location information enables the network to model the spatiotemporal characteristics in accordance with the panel scheduling algorithm that UE will later use, to further improve the power efficiency. Alternatively, the UE may report the angle of arrival data (azimuth and, optionally, zenith) for the incoming multipath components, without including information on panel configuration of the UE.

[0048] In a first group of embodiments, the network provides antenna panel selection information to UEs which is collected over the time from the reports that are sent by other UEs in the past. In other words, earlier reports received in the uplink direction form the antenna panel selection information are sent as such to other UEs. Then, it is up to these UEs how to make use of this information to achieve a power efficient panel activation scheduler algorithm. The network may provide the antenna panel selection information obtained by the network based on reports received from the other UEs.

[0049] For example, the network may share with UEs information from reports received from the other UEs as such, without processing the information. A potential drawback of this is, that the information in unprocessed form may be too voluminous to send to UEs if a large amount of collected measurement data is required to derive meaningful panel activation schedules.

[0050] Alternatively, the network may provide to UEs processed information as antenna panel selection information, obtained from the collected and reported measurements from the other UEs. This may be from a certain time frame, such as, from reports received within the past week, or the past month, to enable adapting to changes in reflecting and blocking objects. Statistical analysis of the reported information, such asmean and standard deviation of angle of arrival, for both azimuth and zenith. An angular spread of arrival for both azimuth and zenith may also be included in the processed information. Providing an angular spread may comprise providing the extreme values, such as maximum and minimum azimuth and zenith angles received in the uplink reports from that UE location. This processed information may be clustered over quantized coordinates and timing advance, then to be shared with UEs, for example. For example, the network, for example a base station, may use a clustering process to obtain, for each quantized location in the coverage area, a set of incoming signal azimuth and, optionally, zenith angles, and provide these to UEs as antenna panel selection information.

[0051] Thus, for example, the network may provide to a UE average azimuth angles of wireless signals to be expected at a quantized location within a cell coverage area. Average in this context does not refer to averaged values of all azimuth angles of all incoming multipath components, but averaged angels of distinct multipath components separately. The quantized location may be a square area, for example, depending on the coordinate system used, in general the quantized location defined an area within the cell coverage area. In some embodiments, also the zenith angles are given, and in yet further embodiments, the standard deviations of the angles are given, and possibly also the spreads of the reported angles. The UE is thus enabled to select, which antenna panels to activate for a mobility measurement, such as an RSRP measurement of serving and neighboring cells.

[0052] In a second group of embodiments, the network provides to UEs a trained ML model, the ML model being trained over time using the reports that are sent by other UEs in the past as training data. In this approach, the ML model that is used on the UE side for power efficient panel activation scheduler is trained on the network side and given to the UE. The network may also configure the UE on how to facilitate a scheduling algorithm since it is known by the network which inputs to be used for the trained ML model.

[0053] In the following, details of an example ML model design are explained. A target of the ML model is an ability of new UEs to select the antenna panels for activation, that is, which antenna panel to switch on in order to achieve power saving while maintaining measurement accuracy from different cells. Challenges include that UEs may have different numbers of antenna panels and different mounting geometries for the panels,that UEs may rotate and so the panel coordinates will change frequently, and spatial temporal information is by nature fairly large datasets to be transferred over the air.

[0054] Due to the different UE vendor design for number of antenna panels and mounting directions, in addition to the frequent movement of UEs, the external shape of the UEs may be represented in a unified manner in order to allow the ML model to provide useful panel selection advice. This may be performed as follows: The UE is defined to be within a cube with six different faces, where each face has a unique identifier (that is; 1, 2, 3, 4, 5, 6). The UE (panels) can rotate freely, but the defined cube is assumed to be fixed with: face 1 always toward true south and perpendicular to earth surface, face 2 always toward true east and perpendicular to earth surface, face 3 always toward true north and perpendicular to earth surface, face 4 always toward true west and perpendicular to earth surface, face 5 is UP and horizontal to earth surface, and finally face 6 is down and horizontal to earth surface. Determining which panel is currently pointing toward which face of the cube may be accomplished by the UE.

[0055] An ML model is used that is able to give statistically independent output probabilities for each face of the six faces of the cube which represents the UE, as a function of UE location. The UE location may be quantized, as was the case for the first group of embodiments. Each output may show the probability that a signal will be received through that corresponding cube’s face, or each output identifies which faces will receive signals.

[0056] The classification model could be designed, for example, using a neural network comprising an input layer followed with plural, for example three, hidden fully connected dense layers, and finally an output layer. The output layer which may have an output dimension equal to the number of targeted faces to be predicted (6 for exmaple) and sigmoid as the activation function, and binary cross entropy as the loss function.

[0057] The ML model may be trained as follows. Firstly, the input is selected so that it will only contribute to the learning of the model. In the method disclosed herein the input will consist of input samples, where each sample comprises a vector of N azimuth angles of arrival (horizontal plane), where N is the maximum number of measured cells or multipath components. The input sample further comprises a vector of N zenith angles of arrival (vertical plane), where N is again the maximum number of measured cells. The input samples further comprise measurement coordinates, which are the UE location whenthe measurement was performed, and RSRP values for each measured cell. Secondly, a ground truth (label) is set per each selected input sample. The UE may report the cube face(s) that corresponds to the current panel(s) that is, or are, receiving the signals. The reported faces may form a vector of size 6, for example, with “1” indicating a signal is received, and “0” indicating no signal is received in the corresponding panel. The node in the network responsible for the training may be configured to use the reported input samples and ground truth for training the multi -label classification model.

[0058] To exemplify the multi-label classification mechanism, the following use case example is provided:

[0059] There are 6 faces of the cube in total and our ground truth, the labels, may be either 0 (Off) or 1 (ON) for each face independently. If a signal is received through an antenna panel corresponding to that face, then for that face all panel(s) are switched ON. In another words, for an input sample (angles, UE location) we assume the received signal is through faces 1, 5, and 6, then we set the true label (ground truth) = [0 1 0 0 0 1 1]

[0060] So the model’s Ytrue for input [angles, UE location] is = [0 1 0 0 1 1] (counting starts from zero: 0,1, 2, 3, 4, 5). From this the model may start predicting which faces are associated with our input sample. This vector may indicate the antenna panels based on the cube face, such that all panels on a cube face labelled “1” indicated as to be activated.

[0061] The number of neurons in the output layer may be equal to the number of total faces, 6 in our use case. The neural network will start outputting a vector with total faces size which contains 0's and l's. 0's face not predicted, the panel is correspondingly set OFF , 1 face is predicted the panel is correspondingly set ON.

[0062] Using Sigmoid in the output layer with binary cross-entropy as a loss function, each neuron in the output layer may yield a probability for each face independently in the output vector. So, for example and with a well-trained model, if Ytrue [0 0 0 1 1 1 ] for an input sample X, then the prediction output of our model Ypredicted may be something like [0.1 0.2 0.15 0.8 0.7 0.98], which means faces 1,2,3 have high probability OFF (low probability are ON), while faces 4,5,6 with high probability ON. Moreover, face 6 almost sure that will be ON as it got very high probability of 0.98.

[0063] For example, the network-trained ML model may be used in the UE as follows. The UE first activates a subset, but not all, of its antenna panels, and determines at least one angle of arrival of an incoming signal. In some cases, the UE may determine angles of arrival of at least two incoming signals, using the subset of activated panels. These determined angles of arrival are then provided to the ML model as input, optionally together with the UE location, and the ML model outputs a prediction concerning which further antenna panels of the UE, outside of the subset, need to be activated to receive more incoming signals than those received via the subset of antenna panels. In other words, the ML model aids the UE in selecting which further antenna panels, which are not comprised in the subset of antenna panels, need to be activated in order to receive more incoming signals. A benefit is obtained in that panels unlikely to be of use in receiving signals are not needlessly activated, for example to perform a measurement in the UE. In some cases, the subset comprises only a single antenna panel. In some cases, the subset comprises exactly two, and not more than two, antenna panels. The subset of antenna panels may be activated to communicate with a serving cell, for example.

[0064] Concerning the model at the network side, as the spatial temporal information from different UEs already has been collected and reported to the network side then it makes sense to perform the training at the network side. A trained model may be transferred as a whole to the UE side as antenna panel selection information to be used for local prediction: in this case the UE will use either 1) its own newly collected samples as input to the model for predicting and no further information is needed from the network, or 2) the network will share statistical analysis of the reported information from different UEs, for example, mean and / or standard deviation of angles of arrival, for azimuth and, optionally, zenith). The network may also share angular spread of arrival for azimuth and, optionally, zenith). This information from the network may be clustered over quantized US locations and / or timing advance values. The UE may be configured to use this information to perform the prediction. Alternatively, the UE may use its location as input to the ML model, to obtain the prediction on which panels might be activated.

[0065] As an alternative, the network-trained ML model may be configured to accept the UE location as input and to output a set of angles of arrival of multipath components. The angles of arrival may be azimuth angles or azimuth-zenith angle pairs for individual multipath components. In some embodiments, instead of the set of angles of arrival, the ML model outputs, from an input comprising the UE location, a vectorindicating which antenna panels should be activated. In these embodiments, the input may comprise solely the UE location.

[0066] Optionally, the network may train the model as an encoder-decoder at the network side and share the decoder only with the UE. Then, the network may performs initial predictions on the network side using the encoder then send the compressed output of the encoder to the UE, which will use as input to its decoder to get the final output vector of probabilities of which face to consider as ON and which as OFF.

[0067] FIGURE 3 illustrates an example apparatus capable of supporting at least some embodiments of the present invention. Illustrated is device 300, which may comprise, for example, a UE or, in applicable parts, a base station of FIGURE 1, for example. Comprised in device 300 is processor 310, which may comprise, for example, a single- or multi-core processor wherein a single-core processor comprises one processing core and a multi-core processor comprises more than one processing core. Processor 310 may comprise, in general, a control device. Processor 310 may comprise more than one processor. When processor 310 comprises more than one processor, device 300 may be a distributed device wherein processing of tasks takes place in more than one physical unit. Processor 310 may be a control device. A processing core may comprise, for example, a Cortex-A8 processing core manufactured by ARM Holdings or a Zen processing core designed by Advanced Micro Devices Corporation. A processing core or processor may be, or may comprise, at least one qubit. Processor 310 may comprise at least one Qualcomm Snapdragon and / or Intel Atom processor. Processor 310 may comprise at least one application-specific integrated circuit, ASIC. Processor 310 may comprise at least one field-programmable gate array, FPGA. Processor 310, optionally together with memory and computer instructions, may be means for performing method steps in device 300, such as obtaining, selecting, directing, performing and providing. Processor 310 may be configured, at least in part by computer instructions, to perform actions.

[0068] A processor may comprise circuitry, or be constituted as circuitry or circuitries, the circuitry or circuitries being configured to perform phases of methods in accordance with embodiments described herein. As used in this application, the term “circuitry” may refer to one or more or all of the following: (a) hardware-only circuit implementations, such as implementations in only analogue and / or digital circuitry, and (b) combinations of hardware circuits and software, such as, as applicable: (i) a combinationof analogue and / or digital hardware circuit(s) with software / firmware and (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a UE or base station, to perform various functions) and (c) hardware circuit(s) and or processor(s), such as a microprocessor s) or a portion of a microprocessor s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.

[0069] This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.

[0070] Device 300 may comprise memory 320. Memory 320 may comprise randomaccess memory and / or permanent memory. Memory 320 may comprise at least one RAM chip. Memory 320 may be a computer readable medium. Memory 320 may comprise solid- state, magnetic, optical and / or holographic memory, for example. Memory 320 may be at least in part accessible to processor 310. Memory 320 may be at least in part comprised in processor 310. Memory 320 may be means for storing information. Memory 320 may comprise computer instructions that processor 310 is configured to execute. When computer instructions configured to cause processor 310 to perform certain actions are stored in memory 320, and device 300 overall is configured to run under the direction of processor 310 using computer instructions from memory 320, processor 310 and / or its at least one processing core may be considered to be configured to perform said certain actions. Memory 320 may be at least in part comprised in processor 310. Memory 320 may be at least in part external to device 300 but accessible to device 300. Memory 320 may be transitory or non-transitory. The term “non-transitory”, as used herein, is a limitation of the medium itself (that is, tangible, not a signal) as opposed to a limitation on data storage persistency (for example, RAM vs. ROM).

[0071] Device 300 may comprise a transmitter 330. Device 300 may comprise a receiver 340. Transmitter 330 and receiver 340 may be configured to transmit and receive, respectively, information in accordance with at least one cellular or non-cellular standard. Transmitter 330 may comprise more than one transmitter. Receiver 340 may comprise more than one receiver. Transmitter 330 and / or receiver 340 may be configured to operate in accordance with global system for mobile communication, GSM, wideband code division multiple access, WCDMA, 5G, long term evolution, LTE, IS-95, wireless local area network, WLAN, Ethernet and / or worldwide interoperability for microwave access, WiMAX, standards, for example.

[0072] Device 300 may comprise a near-field communication, NFC, transceiver 350. NFC transceiver 350 may support at least one NFC technology, such as NFC, Bluetooth, Wibree or similar technologies.

[0073] Device 300 may comprise user interface, UI, 360. UI 360 may comprise at least one of a display, a keyboard, a touchscreen, a vibrator arranged to signal to a user by causing device 300 to vibrate, a speaker or a microphone. A user may be able to operate device 300 via UI 360, for example to accept incoming telephone calls, to originate telephone calls or video calls, to browse the Internet, to manage digital files stored in memory 320 or on a cloud accessible via transmitter 330 and receiver 340, or via NFC transceiver 350, and / or to play games.

[0074] Device 300 may comprise or be arranged to accept a user identity module 370. User identity module 370 may comprise, for example, a subscriber identity module, SIM, card installable in device 300. A user identity module 370 may comprise information identifying a subscription of a user of device 300. A user identity module 370 may comprise cryptographic information usable to verify the identity of a user of device 300 and / or to facilitate encryption of communicated information and billing of the user of device 300 for communication effected via device 300.

[0075] Processor 310 may be furnished with a transmitter arranged to output information from processor 310, via electrical leads internal to device 300, to other devices comprised in device 300. Such a transmitter may comprise a serial bus transmitter arranged to, for example, output information via at least one electrical lead to memory 320 for storage therein. Alternatively to a serial bus, the transmitter may comprise a parallel bus transmitter. Likewise processor 310 may comprise a receiver arranged to receiveinformation in processor 310, via electrical leads internal to device 300, from other devices comprised in device 300. Such a receiver may comprise a serial bus receiver arranged to, for example, receive information via at least one electrical lead from receiver 340 for processing in processor 310. Alternatively to a serial bus, the receiver may comprise a parallel bus receiver.

[0076] Device 300 may comprise further devices not illustrated in FIGURE 3. For example, where device 300 comprises a smartphone, it may comprise at least one digital camera. Some devices 300 may comprise a back-facing camera and a front-facing camera, wherein the back-facing camera may be intended for digital photography and the frontfacing camera for video telephony. Device 300 may comprise a fingerprint sensor arranged to authenticate, at least in part, a user of device 300. In some embodiments, device 300 lacks at least one device described above. For example, some devices 300 may lack a NFC transceiver 350 and / or user identity module 370.

[0077] Processor 310, memory 320, transmitter 330, receiver 340, NFC transceiver 350, UI 360 and / or user identity module 370 may be interconnected by electrical leads internal to device 300 in a multitude of different ways. For example, each of the aforementioned devices may be separately connected to a master bus internal to device 300, to allow for the devices to exchange information. However, as the skilled person will appreciate, this is only one example and depending on the embodiment various ways of interconnecting at least two of the aforementioned devices may be selected without departing from the scope of the present invention.

[0078] FIGURE 4 illustrates signalling in accordance with at least some embodiments of the present invention. On the vertical axes are disposed, on the left, three UEs UE1, UE2 and UE3, in the middle, base station 130, and on the right, UE 110. Time advances from the top toward the bottom.

[0079] In phases 410, 420 and 430, UEs UE1, UE2 and UE3, respectively, report to base station 130 their measurements. These reports include the respective UE location when the measurement was done, and angles of arrival of wireless signals detected in the measurements. The angles of arrival include azimuth angles and, optionally, also zenith angles for individual multipath components. The number of reports may be more than three, for example, there may be one hundred reports from UEs at different locationswithin the cell coverage area. In phase 440, base station 130 determines characteristics of the reports of phases 410 - 430.

[0080] In phase 450, base station 130 provides to UE 110 antenna panel selection information determined based on the characteristics determined in phase 440. For example, base station 130 may select which information to provide based on a location of UE 110. Phase 450 may take place as a response to UE 110 requesting the antenna panel selection information, wherein such requesting may be specifically for a current location of UE 110. In phase 460, UE 110 selects which antenna panels to activate, based at least in part on the antenna panel selection information received in phase 450. In phase 470, UE 110 performs an RRM measurement using the antenna panels selected in phase 460, leaving other antenna panels in an inactive state.

[0081] Finally, in phase 480 UE 110 reports to base station 130, or indeed to another base station, results of the measurement of phase 470.

[0082] FIGURE 5 is a flow graph of a method in accordance with at least some embodiments of the present invention. The phases of the illustrated method may be performed in UE 110, for example, or in a control device configured to control the functioning thereof, when installed therein.

[0083] Phase 510 comprises obtaining, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face. Phase 520 comprises selecting, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus. Finally, phase 530 comprises directing the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

[0084] It is to be understood that the embodiments of the invention disclosed are not limited to the particular structures, process steps, or materials disclosed herein, but are extended to equivalents thereof as would be recognized by those ordinarily skilled in the relevant arts. It should also be understood that terminology employed herein is used for the purpose of describing particular embodiments only and is not intended to be limiting.

[0085] Reference throughout this specification to one embodiment or an embodiment means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, appearances of the phrases “in one embodiment” or “in an embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment. Where reference is made to a numerical value using a term such as, for example, about or substantially, the exact numerical value is also disclosed.

[0086] As used herein, a plurality of items, structural elements, compositional elements, and / or materials may be presented in a common list for convenience. However, these lists should be construed as though each member of the list is individually identified as a separate and unique member. Thus, no individual member of such list should be construed as a de facto equivalent of any other member of the same list solely based on their presentation in a common group without indications to the contrary. In addition, various embodiments and example of the present invention may be referred to herein along with alternatives for the various components thereof. It is understood that such embodiments, examples, and alternatives are not to be construed as de facto equivalents of one another, but are to be considered as separate and autonomous representations of the present invention.

[0087] Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the preceding description, numerous specific details are provided, such as examples of lengths, widths, shapes, etc., to provide a thorough understanding of embodiments of the invention. One skilled in the relevant art will recognize, however, that the invention can be practiced without one or more of the specific details, or with other methods, components, materials, etc. In other instances, well-known structures, materials, or operations are not shown or described in detail to avoid obscuring aspects of the invention.

[0088] While the forgoing examples are illustrative of the principles of the present invention in one or more particular applications, it will be apparent to those of ordinary skill in the art that numerous modifications in form, usage and details of implementation can be made without the exercise of inventive faculty, and without departing from the principles and concepts of the invention. Accordingly, it is not intended that the invention be limited, except as by the claims set forth below.

[0089] The verbs “to comprise” and “to include” are used in this document as open limitations that neither exclude nor require the existence of also un-recited features. The features recited in depending claims are mutually freely combinable unless otherwise explicitly stated. Furthermore, it is to be understood that the use of "a" or "an", that is, a singular form, throughout this document does not exclude a plurality.

[0090] As used herein, “at least one of the following: ” and “at least one of ” and similar wording, where the list of two or more elements are joined by “and” or “or”, mean at least any one of the elements, or at least any two or more of the elements, or at least all the elements. INDUSTRIAL APPLICABILITY

[0091] At least some embodiments of the present invention find industrial application in operation of user equipments in a cellular network.ACRONYMS LISTRRM radio resource management

Claims

CLAIMS:

1. An apparatus comprising at least one processing core and at least one memory storing instructions that, when executed by the at least one processing core, cause the apparatus at least to:- obtain, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face;- select, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and- direct the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

2. The apparatus according to claim 1, wherein the apparatus is configured to send, to the network, a measurement report comprising information specifying directions from which wireless signals from the network are determined to arrive at the apparatus and information specifying at least one panel of the apparatus that received the wireless signals, wherein the information specifying the at least one panel of the apparatus is expressed as at least one face of a cube enclosing the apparatus.

3. The apparatus according to any of claims 1 - 2, wherein a face of the cube enclosing the apparatus faces north and is perpendicular to the Earth’s surface.

4. The apparatus according to claim 2, wherein the measurement report further comprises information specifying a location of the apparatus at a time when the directions from which the wireless signals from the network were determined to arrive, wherein the machine learning model is further configured to output probabilities for at least the subset of the six first faces based on information specifying a current location of the apparatus and wherein the apparatus is configured to select one or more antenna panels further, based on the information specifying the current location of the apparatus.

5. The apparatus according to any of claims 1 - 4, wherein the apparatus is configured to determine, based on the output of the machine learning model, antenna panel activation rate information for the one or more antenna panels of the user equipment.

6. The apparatus according to any of claims 1 - 5, wherein the apparatus is configured to update the machine learning model using as training data measurement data obtained in the apparatus by performing a measurement by the apparatus.

7. The apparatus according to any of claims 1 - 6, wherein the apparatus is comprised in a user equipment as a control chip of the user equipment, or where the apparatus is the user equipment.

8. An apparatus comprising at least one processing core and at least one memory storing instructions that, when executed by the at least one processing core, cause the apparatus at least to:- obtain, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment;- obtain a machine learning model, the model being obtained from the information specifying directions from which wireless signals from a network were determined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning model being configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and- provide, to a first user equipment, the machine learning model.

9. The apparatus according to claim 7, wherein a first face of the cube enclosing the user equipment faces north and is perpendicular to the Earth’s surface.

10. The apparatus according to claim 8, wherein the measurement reports further comprise information specifying a location of the user equipment and wherein the machine learning model is further determined based on the information specifying the location of the user equipment.

11. The apparatus according to any of claims 8 - 10, wherein the machine learning model associates at least one geographic location with directions of arrival of wireless signals from a network in which the apparatus is comprised.

12. The apparatus according to any of claims 7 - 9, wherein the apparatus is configured to train the machine learning model using the measurement reports as training data.

13. A method comprising:- obtaining, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face;- selecting, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and- directing the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

14. A method comprising:- obtaining, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment;- obtaining a machine learning model, the model being obtained from the information specifying directions from which wireless signals from a network weredetermined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning model being configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and- providing, to a first user equipment, the machine learning model.

15. A non-transitory computer readable medium having stored thereon a set of computer readable instructions that, when executed by at least one processor, cause an apparatus to at least:- obtain, from a network, a machine learning model representing a first apparatus as a first cube with six first faces, the machine learning model being configured to output probabilities for at least a subset of the six first faces, that a signal will be received through a respective first face;- select, based at least in part on the machine learning model and at least information specifying directions from which wireless signals from the network are determined to arrive at the apparatus, one or more antenna panels of the apparatus, and- direct the apparatus to perform a radio measurement of the wireless signals from the network using the selected one or more antenna panels.

16. A non-transitory computer readable medium having stored thereon a set of computer readable instructions that, when executed by at least one processor, cause an apparatus to at least:- obtain, from plural user equipments, measurement reports wherein each of the measurement reports comprises information specifying directions from which wireless signals from a network were determined to arrive at a user equipment and information specifying at least one panel of the user equipment that received the wireless signals, wherein the information specifying at least one panel of the user equipment is expressed as at least one face of a cube enclosing the user equipment;- obtain a machine learning model, the model being obtained from the information specifying directions from which wireless signals from a network were determined to arrive and the information specifying at least one panel receiving the wireless signals expressed as the at least one face of the cube, the machine learning modelbeing configured to output probabilities for at least a subset of six first faces of a first cube, the probabilities indicating whether a signal will be received through the at least a subset of the six first faces of the first cube, and- provide, to a first user equipment, the machine learning model.

Images