Determining validity of beam prediction model
By assessing beam prediction model validity using a first model to compare beam quality quantities, the approach addresses inefficiencies in conventional beam prediction systems, enabling timely and efficient model recovery and accurate predictions.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- HUAWEI TECH CO LTD
- Filing Date
- 2024-11-06
- Publication Date
- 2026-05-15
AI Technical Summary
Conventional beam prediction models in wireless communication systems lack efficient methods for determining their validity, leading to increased computational load and power consumption due to dedicated monitoring measurements for model failure detection and recovery.
A communication device measures reference signals using a first beam prediction model to assess the validity of a second beam prediction model by comparing beam quality quantities, allowing early detection of model failures and triggering timely recovery procedures without additional measurements.
This approach reduces computational complexity and power consumption by proactively detecting beam prediction errors, enabling rapid model failure recovery and ensuring accurate beam predictions, thus optimizing resource utilization.
Smart Images
Figure CN2024130144_15052026_PF_FP_ABST
Abstract
Description
DETERMINING VALIDITY OF BEAM PREDICTION MODELTechnical Field
[0001] Embodiments of the invention relate to first and second communication devices for determining the validity of a beam prediction model. Furthermore, embodiments of the invention also relate corresponding methods and a computer program.Background
[0002] Artificial intelligence (AI) , and more precisely machine learning (ML) , methods have proven their worth in a multitude of fields in wireless networks, covering different problems, including classification, regression and prediction, pattern detection, dimensionality reduction and interaction with dynamic environments. This potential can be particularly useful in the air interface of wireless communication systems.
[0003] ML models are capable of capturing non-trivial dependencies in data, that conventional signal processing techniques, typically used in air interface, are incapable of leveraging. Some operations in the air interface and in the wireless network can be enhanced and rendered more energy efficient by exploiting a ML approach.Summary
[0004] An objective of embodiments of the invention is to provide a solution which mitigates or solves the drawbacks and problems of conventional solutions.
[0005] Another objective of embodiments of the invention is to provide a solution which makes it possible to determine the validity of beam prediction models in communication systems and networks.
[0006] The above and further objectives are solved by the subject matter of the independent claims. Further embodiments of the invention can be found in the dependent claims.
[0007] According to a first aspect of the invention, the above mentioned and other objectives are achieved with a first communication device configured to:
[0008] measure a set of reference signals transmitted from a set of transmitter beams of a second communication device and received in a set of receiver beams of the first communication device;
[0009] determine a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals and a first beam prediction model;
[0010] determine a validity or an invalidity for a second beam prediction model based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals and the second beam prediction model; and
[0011] transmit a report to the second communication device, wherein the report indicates:
[0012] a beam prediction for a data transmission when the second beam prediction model is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model.
[0013] An advantage of the first communication device according to the first aspect is that predictability of beam quantities is assessed in a low complexity manner before performing the bulk of the computations for beam and / or beam pair predictions. The coupling of the two models can be assured by coupling the first and second prediction models during training or by assessing the correlation in their predictions. This will result in coupling the performance of the two models hence enabling the first communication device to infer the likely performance of the second model from that of the first model. Additionally, the first communication device can compare the outcome of the prediction from the first model with ground truth obtained from measurements on a second subset of measured reference signals. Consequently, beam prediction errors of the second beam prediction model, used as the main model, and model failures may be assessed, proactively. Additionally, beam failure or beam prediction model failure can be detected early hence adequate recovery procedures can be triggered in a timely manner, thereby reducing the timeline of model recovery.
[0014] In an implementation form of a first communication device according to the first aspect, the report indicates:
[0015] a model failure for the second beam prediction model when the second beam prediction model is determined invalid.
[0016] An advantage with this implementation form is that eventual model failure, i.e., inability to provide accurate or trustworthy beam / beam pair prediction, can be detected early without having to perform the bulk of computations for the second beam prediction model. Consequently, model failure recovery procedure can be initiated in a timely manner and beam failure can hence be avoided.
[0017] In an implementation form of a first communication device according to the first aspect, the model failure indicates a switching request for the second beam prediction model, or an alternative beam prediction model to the second beam prediction model, or a fallback beam prediction model to the second beam prediction model.
[0018] An advantage with this implementation form is that the first communication device is able to initiate the adaptation of the prediction model when prediction accuracy is not guaranteed. Consequently, model failure recovery can be initiated and performed rapidly.
[0019] In an implementation form of a first communication device according to the first aspect, the first subset of measured reference signals is received in at least one first time instance and the second subset of measured reference signals is received in at least one second time instances after the first time instance.
[0020] An advantage with this implementation form is that the measurements and predictions needed to determine the validity or invalidity of the second beam prediction model, are performed early enough in time so that the first set of beam quality quantities, obtained though prediction by the first beam prediction model, can be compared with the outcome of the measurements on the second subset of reference signals. In this manner, the accuracy or any other performance metric for the prediction of the first prediction model can be assessed in a precise manner and hence model monitoring can be performed based on the measurements used as input for the second prediction model. Consequently, model monitoring can be performed without requiring additional measurements dedicated for monitoring.
[0021] In an implementation form of a first communication device according to the first aspect, the first subset of measured reference signals comprises K number of time instances and the set of measured reference signals comprises M number of time instances, where K<M.
[0022] An advantage with this implementation form is that the first prediction model which is performed to determine the validity of the second prediction model, considered as the main model, can be determined on a subset of input measurements. More precisely, the first subset of measured reference signals is used given its earlier reception in time. This means that the first prediction can be obtained before all M number of time instances are measured, which results in timely assessment of the validity or invalidity and expected performance of the second beam prediction model.
[0023] In an implementation form of a first communication device according to the first aspect, the first communication device is configured to:
[0024] receive a first control message from the second communication device, the first control message indicating at least one of K and M.
[0025] An advantage with this implementation form is that the number of input measurements in the first subset of measured reference signals can be adapted dynamically or semi-statically in control signaling. Since beam switching and dwelling time depend on the mobility of the first communication device, it is beneficial to be able to adapt the number of measurements so the actual radio and propagation conditions are taken into consideration.
[0026] In an implementation form of a first communication device according to the first aspect, the first communication device is configured to:
[0027] determine the validity for the second beam prediction model based on a comparison of the first set of beam quality quantities with the second set of beam quality quantities and at least one threshold value.
[0028] An advantage with this implementation form is that the verification of the validity for the second beam prediction model can be performed in a low complexity manner and leveraging already supported beam quality quantities such as RSRP and L1-SINR. As an example, the first communication device can use the first beam prediction to obtain RSRP for the measured reference signal resources in the first measurements subset, and compare the outcome with the RSRP obtained for the same reference signal resources, in the second measurement subset. If the difference between the predicted and measured RSRPs are within an interval e.g., defined by a configured threshold, the prediction by the second beam prediction model is determined valid. Otherwise, the prediction by the second beam prediction model is considered as invalid.
[0029] In an implementation form of a first communication device according to the first aspect, the first communication device is configured to:
[0030] determine the validity for the second beam prediction model upon reception of a second control message from the second communication device, the second control message indicating an activation of determining the validity of the second beam prediction model.
[0031] An advantage with this implementation form is that the proposed solution can be used as opportunistic, aperiodic behavior by the first communication device that is triggered when needed. In this case, the second communication device can trigger the determination of the validity of the second beam prediction model when it assesses that the current active model is dropping in accuracy and / or that new radio conditions have been detected and hence validity assessment of the second beam prediction model become necessary.
[0032] In an implementation form of a first communication device according to the first aspect, the first communication device is configured to:
[0033] determine the beam prediction for the data transmission when the second beam prediction model is determined valid.
[0034] An advantage with this implementation form is that beam prediction is used to determine the transmission and / or receiver beams for the data transmission, only when the second beam prediction model is determined as valid. Hence, probability for beam failure is reduced and probability for successful data transmission is increased.
[0035] In an implementation form of a first communication device according to the first aspect, the first set of beam quality quantities and the second set of beam quality quantities comprises:
[0036] RSRPs, SINRs or RSSIs for the set of transmitter beams, the set of receiver beams, or a set of transmitter beam and receiver beam pairs; or
[0037] an ascending or a descending order of the set of transmitter beams, the set of receiver beams, or a set of transmitter beam and receiver beam pairs in terms of a RSRP, a SINR or a RSSI.
[0038] An advantage with this implementation form is that already supported beam quality quantities can be used thus reducing the complexity of the proposed scheme. While validity conditions may be based on the comparison of the difference between measured and predicted beam quality quantities, the order of beams, in terms of RSRP, SINR or RSSI can be used. This can be particularly useful when the goal is to determine the strongest beams with certain flexibility in terms of prediction and measurements accuracy.
[0039] In an implementation form of a first communication device according to the first aspect, the beam prediction comprises:
[0040] beam identities or reference signal resource indicators for the set of transmitter beams, the set of receiver beams, or a set of transmitter beam and receiver beam pairs; or
[0041] RSRPs, SINRs or RSSIs for the set of transmitter beams, the set of receiver beams, or a set of transmitter beam and receiver beam pairs determined for one or multiple time instances following a prediction time instance for the beam prediction.
[0042] An advantage with this implementation form is that multiple scenarios can be supported by the present solution, including transmit, receive and joint-transmit-receive beamforming. Additionally, the proposed solution can be used at different stages of the beam management procedure.
[0043] According to a second aspect of the invention, the above mentioned and other objectives are achieved with a second communication device configured to:
[0044] transmit a set of reference signals to a first communication device in a set of transmitter beams of the second communication device and received in a set of receiver beams of the first communication device;
[0045] receive a report from the first communication device, wherein the report indicates:
[0046] a beam prediction for a data transmission when a second beam prediction model is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model.
[0047] An advantage of the second communication device according to the second aspect is that predictability of beam quantities is assessed in a low complexity manner before performing the bulk of the computations for beam and / or beam pair predictions of the second beam prediction model. Consequently, beam prediction errors of the second beam prediction model, used as the main model, and model failures may be assessed, proactively. Additionally, beam failure or beam prediction model failure can be detected early hence adequate recovery procedures can be triggered in a timely manner, thereby reducing the timeline of model recovery.
[0048] In an implementation form of a second communication device according to the second aspect, the report indicates:
[0049] a model failure for the second beam prediction model when the second beam prediction model is determined invalid.
[0050] An advantage with this implementation form is that eventual model failure, i.e., inability to provide accurate or trustworthy beam / beam pair prediction, can be detected early without having to perform the bulk of computations for the second beam prediction model. Consequently, model failure recovery procedure can be initiated in a timely manner and beam failure can hence be avoided.
[0051] In an implementation form of a second communication device according to the second aspect, the model failure indicates a switching request for the second beam prediction model, or an alternative beam prediction model to the second beam prediction model, or a fallback beam prediction model to the second beam prediction model.
[0052] An advantage with this implementation form is that the first communication device is able to initiate the adaptation of the prediction model when prediction accuracy is not guaranteed. Consequently, model failure recovery can be initiated and performed rapidly.
[0053] In an implementation form of a second communication device according to the second aspect, the first subset of measured reference signals comprises K number of time instances and the set of measured reference signals comprises M number of time instances, where K<M.
[0054] An advantage with this implementation form is that the first prediction model which is performed to determine the validity of the second prediction model, considered as the main model, can be determined on a subset of input measurements. More precisely, the first subset of measured reference signals is used given its earlier reception in time. This means that the first prediction can be obtained before all M number of time instances are measured, which results in timely assessment of the validity or invalidity and expected performance of the second beam prediction model.
[0055] In an implementation form of a second communication device according to the second aspect, the second communication device is configured to:
[0056] transmit a first control message to the first communication device, the first control message indicating at least one of K and M.
[0057] An advantage with this implementation form is that the number of input measurements in the first subset of measured reference signals can be adapted dynamically or semi-statically in control signaling. Since beam switching and dwelling time depend on the mobility of the first communication device, it is beneficial to be able to adapt the number of measurements so the actual radio and propagation conditions are taken into consideration.
[0058] In an implementation form of a second communication device according to the second aspect, the second communication device is configured to:
[0059] transmit a second control message to the first communication device, the second control message indicating an activation of determining the validity of the second beam prediction model.
[0060] An advantage with this implementation form is that the proposed solution can be used as opportunistic, aperiodic behavior by the first communication device that is triggered when needed. In this case, the second communication device can trigger the determination of the validity of the second beam prediction model when it assesses that the current active model is dropping in accuracy and / or that new radio conditions have been detected and hence validity assessment of the second beam prediction model become necessary.
[0061] In an implementation form of a second communication device according to the second aspect, the beam prediction comprises:
[0062] beam identities or reference signal resource indicators for the set of transmitter beams, the set of receiver beams, or a set of transmitter beam and receiver beam pairs; or
[0063] RSRPs, SINRs or RSSIs for the set of transmitter beams, the set of receiver beams, or a set of transmitter beam and receiver beam pairs determined for one or multiple time instances following a prediction time instance for the beam prediction.
[0064] An advantage with this implementation form is that multiple scenarios can be supported by the present solution, including transmit, receive and joint-transmit-receive beamforming. Additionally, the proposed solution can be used at different stages of the beam management procedure.
[0065] According to a third aspect of the invention, the above mentioned and other objectives are achieved with a method for a first communication device, the method comprises:
[0066] measuring a set of reference signals transmitted from a set of transmitter beams of a second communication device and received in a set of receiver beams of the first communication device;
[0067] determining a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals and a first beam prediction model;
[0068] determining a validity or an invalidity for a second beam prediction model based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals and the second beam prediction model; and
[0069] transmitting a report to the second communication device, wherein the report indicates:
[0070] a beam prediction for a data transmission when the second beam prediction model is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model.
[0071] The method according to the third aspect can be extended into implementation forms corresponding to the implementation forms of the first communication device according to the first aspect. Hence, an implementation form of the method comprises the feature (s) of the corresponding implementation form of the first communication device.
[0072] The advantages of the methods according to the third aspect are the same as those for the corresponding implementation forms of the first communication device according to the first aspect.
[0073] According to a fourth aspect of the invention, the above mentioned and other objectives are achieved with a method for a second communication device, the method comprises:
[0074] transmitting a set of reference signals to a first communication device in a set of transmitter beams of the second communication device and received in a set of receiver beams of the first communication device; and
[0075] receiving a report from the first communication device, wherein the report indicates:
[0076] a beam prediction for a data transmission when a second beam prediction model is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model.
[0077] The method according to the fourth aspect can be extended into implementation forms corresponding to the implementation forms of the second communication device according to the second aspect. Hence, an implementation form of the method comprises the feature (s) of the corresponding implementation form of the second communication device.
[0078] The advantages of the methods according to the fourth aspect are the same as those for the corresponding implementation forms of the second communication device according to the second aspect.
[0079] Embodiments of the invention also relate to a computer program, characterized in program code, which when run by at least one processor causes the at least one processor to execute any method according to embodiments of the invention. Further, embodiments of the invention also relate to a computer program product comprising a computer readable medium and the mentioned computer program, wherein the computer program is included in the computer readable medium, and may comprises one or more from the group of: read-only memory (ROM) , programmable ROM (PROM) , erasable PROM (EPROM) , flash memory, electrically erasable PROM (EEPROM) , hard disk drive, etc.
[0080] Further applications and advantages of embodiments of the invention will be apparent from the following detailed description.Brief Description of the Drawings
[0081] The appended drawings are intended to clarify and explain different embodiments of the invention, in which:
[0082] -Fig. 1 shows a first communication device according to embodiments of the invention;
[0083] -Fig. 2 shows a flow chart of a method for a first communication device according to embodiments of the invention;
[0084] -Fig. 3 shows a second communication device according to embodiments of the invention;
[0085] -Fig. 4 shows a flow chart of a method for a second communication device according to embodiments of the invention;
[0086] -Fig. 5 shows a communication system according to embodiments of the invention;
[0087] -Fig. 6 illustrates further aspects of embodiments of the invention; and
[0088] -Fig. 7 shows a signaling diagram illustrating further embodiments of the invention.Detailed Description
[0089] While the aforementioned use cases are quite challenging and important, given that 3GPP new radio (NR) is a beamforming-based air interface, it is worth highlighting that the potential of AI / ML can be leveraged in other important use cases such as, link adaptation enhancements, hardware impairment and mobile terminal (MT) / gNB-side implementation enhancement, L1 / L2-mobility support enhancement, adaptation to different traffic types, and cross link interference measurement and management.
[0090] Beam management is one of the most critical features of NR, especially in the high frequency ranges. NR beam management can be broken down into beam selection during initial access and three main procedures, namely: P-1 which is used to enable user equipment (UE) measurement on different transmission and reception point (TRP) transmit (Tx) beams to support selection of TRP Tx beams / UE receiver (Rx) beam (s) ; P-2 which is used to enable UE measurement on different TRP Tx beams to possibly change inter / intra-TRP Tx beam (s) from a typically smaller set of beams for beam refinement than in P-1; and P-3 which is used to enable UE measurement on the same TRP Tx beam to change UE Rx beam in case the UE uses beamforming.
[0091] In Rel-18 study item on ML operations for beam management, two variants are considered, namely, beam prediction in the spatial domain and the time domain, respectively. Beam prediction in the time domain is considered in order to better handle UE mobility. Depending on UE mobility pattern and velocity, the optimal beam (s) would vary over time. The rate of this change may depend on beam design, UE mobility pattern and velocity, among other things. In order to keep track of the optimal beam (s) , reference signal (RS) measurements are performed in the downlink (DL) or the uplink (UL) , or both in DL and UL.
[0092] Model life cycle procedures include a dedicated procedure for beam model failure detection and recovery. However, these procedures typically involve dedicated monitoring measurements. Hence, the measurement load will increase, from the UE perspective. Several of considered conventional approaches for beam model failure detection and recovery involve additional monitoring measurements which increase the computations and power consumption in the UE.
[0093] From the above discussion, embodiments of the invention aim, among other things, at enabling beam prediction model failure detection and recovery in a low-cost and low-complex manner. Thus, early detection of model failure is provided so that resources are not spent on performing a main beam prediction model if failure is expected. Further, no additional reference signal measurements for the sake of model failure detection are needed.
[0094] Fig. 1 shows a first communication device 100 according to an embodiment of the invention. In the embodiment shown in Fig. 1, the first communication device 100 comprises a processor 102, a transceiver 104 and a memory 106. The processor 102 is coupled to the transceiver 104 and the memory 106 by communication means 108 known in the art. The first communication device 100 may be configured for wireless and / or wired communications in a communication system. The wireless communication capability may be provided with an antenna or antenna array 110 coupled to the transceiver 104, while the wired communication capability may be provided with a wired communication interface 112 e.g., coupled to the transceiver 104. The antenna or antenna array 110 comprises a set of receiver beams 120.
[0095] The processor 102 may be referred to as one or more general-purpose central processing units (CPUs) , one or more digital signal processors (DSPs) , one or more application-specific integrated circuits (ASICs) , one or more field programmable gate arrays (FPGAs) , one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, or one or more chipsets. The memory 106 may be a read-only memory, a random access memory (RAM) , or a non-volatile RAM (NVRAM) . The transceiver 104 may be a transceiver circuit, a power controller, or an interface providing capability to communicate with other communication modules or communication devices, such as network nodes and network servers. The transceiver 104, memory 106 and / or processor 102 may be implemented in separate chipsets or may be implemented in a common chipset.
[0096] That the first communication device 100 is configured to perform certain actions can in this disclosure be understood to mean that the first communication device 100 comprises suitable means and devices, such as e.g., the processor 102 and the transceiver 104, configured to perform the actions.
[0097] According to embodiments of the invention the first communication device 100 is configured to:measure a set of reference signals 530 transmitted from a set of transmitter beams 320 of a second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; determine a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals 530 and a first beam prediction model M1; determine a validity or an invalidity for a second beam prediction model M2 based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals 530 and the second beam prediction model M2; and transmit a report 540 to the second communication device 300, wherein the report 540 indicates: a beam prediction for a data transmission 550 when the second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model M2.
[0098] Furthermore, in an embodiment of the invention, the first communication device 100 comprises a processor configured to: measure a set of reference signals 530 transmitted from a set of transmitter beams 320 of a second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; determine a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals 530 and a first beam prediction model M1; determine a validity or an invalidity for a second beam prediction model M2 based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals 530 and the second beam prediction model M2. The first communication device 100 comprises a transceiver configured to:transmit a report 540 to the second communication device 300, wherein the report 540 indicates: a beam prediction for a data transmission 550 when the second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model M2.
[0099] Moreover, in yet another embodiment of the invention, the first communication device 100 for a communication system 500 comprises a processor and a memory having computer readable instructions stored thereon which, when executed by the processor, cause the processor to: measure a set of reference signals 530 transmitted from a set of transmitter beams 320 of a second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; determine a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals 530 and a first beam prediction model M1; determine a validity or an invalidity for a second beam prediction model M2 based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals 530 and the second beam prediction model M2; and transmit a report 540 to the second communication device 300, wherein the report 540 indicates: a beam prediction for a data transmission 550 when the second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model M2.
[0100] Fig. 2 shows a flow chart of a corresponding method 200 which may be executed in a first communication device 100, such as the one shown in Fig. 1. The method 200 comprises: measuring 202 a set of reference signals 530 transmitted from a set of transmitter beams 320 of a second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; determining 204 a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals 530 and a first beam prediction model M1; determining 206 a validity or an invalidity for a second beam prediction model M2 based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals 530 and the second beam prediction model M2; and transmitting 208 a report 540 to the second communication device 300, wherein the report 540 indicates: a beam prediction for a data transmission 550 when the second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model M2.
[0101] Fig. 3 shows a second communication device 300 according to an embodiment of the invention. In the embodiment shown in Fig. 3, the second communication device 300 comprises a processor 302, a transceiver 304 and a memory 306. The processor 302 is coupled to the transceiver 304 and the memory 306 by communication means 308 known in the art. The second communication device 300 may be configured for wireless and / or wired communications in a communication system. The wireless communication capability may be provided with an antenna or antenna array 310 coupled to the transceiver 304, while the wired communication capability may be provided with a wired communication interface 312 e.g., coupled to the transceiver 304. The antenna or antenna array 310 comprises a set of transmitter beams 320.
[0102] The processor 302 may be referred to as one or more general-purpose CPUs, one or more DSPs, one or more ASICs, one or more FPGAs, one or more programmable logic devices, one or more discrete gates, one or more transistor logic devices, one or more discrete hardware components, one or more chipsets. The memory 306 may be a read-only memory, a RAM, or a NVRAM. The transceiver 304 may be a transceiver circuit, a power controller, or an interface providing capability to communicate with other communication modules or communication devices. The transceiver 304, the memory 306 and / or the processor 302 may be implemented in separate chipsets or may be implemented in a common chipset.
[0103] That the second communication device 300 is configured to perform certain actions can in this disclosure be understood to mean that the second communication device 300 comprises suitable means and devices, such as e.g., the processor 302 and the transceiver 304, configured to perform the actions.
[0104] According to embodiments of the invention the second communication device 300 is configured to: transmit a set of reference signals 530 to a first communication device 100 in a set of transmitter beams 320 of the second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; receive a report 540 from the first communication device 100, wherein the report 540 indicates: a beam prediction for a data transmission 550 when a second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model M2.
[0105] Furthermore, in an embodiment of the invention, the second communication device 300 comprises a transceiver configured to: transmit a set of reference signals 530 to a first communication device 100 in a set of transmitter beams 320 of the second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; and receive a report 540 from the first communication device 100, wherein the report 540 indicates: a beam prediction for a data transmission 550 when a second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model M2.
[0106] Moreover, in yet another embodiment of the invention, the second communication device 300 for a communication system 500 comprises a processor and a memory having computer readable instructions stored thereon which, when executed by the processor, cause the processor to: transmit a set of reference signals 530 to a first communication device 100 in a set of transmitter beams 320 of the second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; receive a report 540 from the first communication device 100, wherein the report 540 indicates: a beam prediction for a data transmission 550 when a second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model M2.
[0107] Fig. 4 shows a flow chart of a corresponding method 400 which may be executed in a second communication device 300, such as the one shown in Fig. 3. The method 400 comprises: transmitting 402 a set of reference signals 530 to a first communication device 100 in a set of transmitter beams 320 of the second communication device 300 and received in a set of receiver beams 120 of the first communication device 100; receiving 404 a report 540 from the first communication device 100, wherein the report 540 indicates: a beam prediction for a data transmission 550 when a second beam prediction model M2 is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model M2.
[0108] Fig. 5 shows a communication system 500 according to an embodiment of the invention. The communication system or mobile network 500 in the disclosed embodiment comprises a first communication device 100 and a second communication device 300 configured to communicate and operate in the communication system 500. For simplicity, the shown communication system 500 only comprises one first communication device 100 and one second communication device 300. However, the communication system 500 may comprise any number of first communication devices 100 and any number of second communication devices 300 without deviating from the scope of the invention.
[0109] In this particular example, the first communication device 100 is configured as a client device, such as a UE, while the second communication device 300 is configured as a network access node, such as a base station. The network access node may be part of a radio access network (RAN) and can include a communication interface for communication with a network (NW) such as a core network (CN) . The communication between the first communication device 100 and the second communication device 300 may be performed using channels in the DL and UL.
[0110] In Fig. 5, the first communication device 100 transmits a report 540 to the second communication device 300. The report 540, which also may be denoted beam report, indicates a beam prediction for a coming / subsequent data transmission 550 when a second beam prediction model M2 is determined valid according to embodiments of the invention. Previous to transmitting the report 540, the first communication device 100 has received reference signals 530 transmitted from the second communication device 300 and made measurements on the reference signals 530 so as to derive the validity or invalidity of the second beam prediction model M2. This is however not shown in Fig. 5.
[0111] Further details related to embodiments of the invention may fully or partially be described in a 3GPP context so as to provide deeper understanding of embodiments of the invention. Thus, 3GPP terminology, definitions, expressions and system architecture may be used. Especially, the first communication device 100 according to the invention may in these embodiments be configured to perform any of the described functions of a 3GPP UE. Correspondingly, the second communication device 300 according to the invention may in these embodiments be configured to perform any of the described functions of a 3GPP gNB. It may however be noted that embodiments of the invention are not limited thereto and e.g., the reverse case is possible. Furthermore, sidelink communications between first and second communication devices are also withing the scope of the present disclosure.
[0112] Embodiments of the invention may be implemented in a 3GPP standard such as 5G NR or future wireless networks air interfaces. If adapted, embodiments of the invention may impact how beam prediction is performed at the UE side, the content of channel state information (CSI) reporting following beam prediction, how model life cycle management is performed, and so on.
[0113] Fig. 6 illustrates further aspects of embodiments of the invention involving among other thing timing aspects and beam quality quantities used for evaluating beam prediction models.
[0114] In Fig. 6, a first beam prediction model M1 and a second beam prediction model M2 are illustrated on a timeline. It is shown how a first subset of measured reference signals is used as input to a first beam prediction model M1. The first beam prediction model M1 based on the first subset of measured reference signals hence outputs a first set of beam quality quantities. It is also shown how a first and a second subset of measured reference signals are used as input to a second beam prediction model M2. Based on the first and second subset of measured reference signals, the second beam prediction model M2 outputs a second set of beam quality quantities. It may be noted that the first subset of measured reference signals is received in a set of first time instances TI1 and the second subset of measured reference signals is received in a set of second time instances TI2 which is after the set of first time instance TI1 in time. The set of time instances may comprise a plurality of time instances.
[0115] By comparing the first set of beam quality quantities with the second set of beam quality quantities it may be determined whether the second beam prediction model M2 is considered valid or not. This may involve the use of one or more threshold values or ranges to which the beam quality quantities are compared. Hence, in embodiments of the invention, the validity for the second beam prediction model M2 is determined based on a comparison of the first set of beam quality quantities with the second set of beam quality quantities and at least one threshold value. For example, the validity can be determined by evaluating the difference between predicted reference signal received power (RSRP) , obtained from the first prediction model, and measured RSRP obtained from measurements based on the second subset of reference signals. If the difference is lower than a certain threshold value then the second beam prediction model M2 is determined valid. Indeed, the first beam prediction model M1 is deemed accurate enough then the second beam prediction model M2 is more likely to justify the same accuracy requirements. This can be assured by coupling the first and second prediction models during training or by assessing the correlation in their predictions. The first beam prediction model M1 may therefore be considered as a model for checking the validity of the second beam prediction model M2. Thus, the first beam prediction model M1 may also be denoted an evaluation model or a sanity check model while the second beam prediction model M2 may be denoted a main beam prediction model.
[0116] The beam quality quantities herein used may indicate how well different beams are suited for transmission or reception of a coming data transmission. Therefore, in embodiments of the invention, the first set of beam quality quantities and the second set of beam quality quantities comprises RSRPs, signal to noise and interference ratios (SINRs) or received signal strength indicators (RSSIs) for the set of transmitter beams 320, the set of receiver beams 120, or a set of transmitter beam 320 and receiver beam 120 pairs. The proposed scheme can be used with already supported beam quality quantities. The RSRP, SINR and RSSI can be used as both input and output of the first M1 and second M2 beam prediction models. The input may be measured beam quality quantities and the output may be predicted beam quality quantities.
[0117] In further embodiments of the invention, the first set of beam quality quantities and the second set of beam quality quantities comprises an ascending or a descending order of the set of transmitter beams 320, the set of receiver beams 120, and / or a set of transmitter beam 320 and receiver beam 120 pairs in terms of RSRP, SINR or RSSI. Indeed, in some cases, the goal is to determine the strongest beams or beam pairs. Hence, the accuracy of the order of beams and / or beam pairs, whether ascending or descending, is convenient and important, and thus certain error margin is acceptable for the predicted beam quality quantities.
[0118] Moreover, as also illustrated in Fig. 6, the first subset of measured reference signals comprises K number of time instances and the set of measured reference signals 530 comprises M number of time instances, where K<M. This means that the measurements used as input for the first prediction model M1 are a subset of the measurements used as input for the second beam prediction model M2. Additionally, the first subset of measurements is performed first in the set of measured reference signals 530. Consequently, the outcome of the first bream prediction model M1 can be compared, in a timely manner with the quantities obtained from the second subset of measurements.
[0119] Regarding the beam prediction sent to the second communication device 300 many alternatives are possible. In non-limiting examples of the invention, the beam prediction may comprise beam identities (IDs) or reference signal resource indicators for the set of transmitter beams 320, the set of receiver beams 120, and / or a set of transmitter beam 320 and receiver beam 120 pairs. In case of a valid second beam prediction model M2, the first communication device 100 would report predicted beams and / or beam pairs. The predicted beams and / or beam pairs can be part of a measured beam set. Identification of the beam and / or beam pair may be done using reference signal resource indicators or beam or beam pair identifiers.
[0120] However, the beam prediction may also comprise RSRPs, SINRs or RSSIs for the set of transmitter beams 320, the set of receiver beams 120, and / or a set of transmitter beam 320 and receiver beam 120 pairs determined for one or multiple time instances following a prediction time instance for the beam prediction. Indeed, together with beam identities or reference signal resource indicators for the set of transmitter beams 320, the set of receiver beams 120, and / or a set of transmitter beam 320 and receiver beam 120 pairs, the first communication device 100 may further report associated beam quality quantities which can be used by the second communication device 300 for later beam management operations or to perform further model monitoring from the perspective of the second communication device 300.
[0121] Fig. 7 shows a signaling diagram involving configuration of the UE 100 and control signaling according to embodiments of the invention.
[0122] In step I in Fig. 7, the UE 100 is configured by the gNB 300. The configuration may involve legacy RRC configuration such as CSI measurements and reporting configurations, reference signals configurations, transmission configuration information (TCI) state configuration, radio resource management (RRM) configuration, among others.
[0123] Furthermore, the configuration in step I may also involve specific configuration related to embodiments of the invention. Thus, the UE 100 may report its capabilities for early model failure and recovery, i.e., being able to perform a sanity / validity check of the second beam prediction model M2.
[0124] If training of the beam prediction models is performed at the UE 100, the UE 100 may train the first M1 and second M2 beam prediction models. The initial training parameters of the beam prediction models, e.g., number of input measurements and prediction horizon may be configured by the network or be selected by the UE 100. If training is performed at the gNB 300, the UE 100 may be configured with the first M1 and second M2 beam prediction models.
[0125] The network may via the gNB 300 further configure the UE 100 with measurement resources, e.g., reference signals transmitted for beam prediction input collection. The network may also configure the UE 100 with specific measurement resources for model performance monitoring in the disclosed solution. Additionally, the network may pre-configure the UE 100 with alternatives of model failure when the second beam prediction model M2 is considered invalid as described below.
[0126] In step II in Fig. 7, the gNB 300 configures the UE 100 further by also transmitting first 510 and second control 520 messages to the UE 100 according to embodiments of the invention.
[0127] The first control message 510 carries information about parameters K and M. Thus, the first control message 510 may be part of RRC signaling and hence part of the previous configuration step I. However, the signaling according to the first control message 510 may in an alternative be part of a separate control signaling procedure such as medium access control (MAC) control element (CE) or downlink control information (DCI) , or radio resource control (RRC) reconfiguration.
[0128] The second control message 520 is on the other hand a trigger mechanism for the UE 100 to start evaluating the validity of the second beam prediction model M2. Hence, the UE 100 determines the validity for the second beam prediction model M2 upon reception of the second control message 520 since an activation of determining the validity of the second beam prediction model M2 is indicated by the second control message 520. Thus, the second control message 520 may be part of DCI signaling in a MAC CE for dynamic signaling. However, the activation may in an alternative be performed during the configuration step I as part of RRC signaling.
[0129] In step III in Fig. 7, the gNB 300 transmits DL reference signals to the UE 100 in its set of transmitter beams 320. The DL reference signals may be synchronization signal block (SSB) or channel state information reference signal (CSI-RS) .
[0130] In step IV in Fig. 7, the UE 100 receives the DL reference signals in its set of receiver beams 120. Since the UE 100 has received the first 510 and second control 520 messages in previous step II, the UE 100 is informed about parameters K and M and also that an evolution of the second beam prediction model M2 should be initiated by the UE 100.
[0131] The UE 100 may therefore measure prediction model input, i.e., beam / RS measurements over the multiple measurement time instances M of the second beam prediction model M2, in order to collect input beam quality quantities, e.g., RSRP, for a set of beams or beam pairs or beams directly. Multiple measurements can be used in order to collect input for time-domain prediction model and / or space-domain prediction model, with a large number of beams which may be spread input measurements over multiple time instances. Parameter K delimits the time instances or time duration for collecting the input to the first beam prediction model M2. The UE 100 hence delimits the measurements in time according to parameters K and M.
[0132] In step V in Fig. 7, the UE 100 therefore determines whether the second beam prediction model M2 is valid or not by evaluating first and second set of beam quality quantities derived as described above.
[0133] The outcome of the beam prediction model evaluation determines the information content of the report 540 which is to be transmitted to the gNB 300.
[0134] If the second beam prediction model M2 is considered valid, the report 540 indicates a beam prediction for a data transmission 550 as previously mentioned. For example, in case, predicted beams comparison with subsequent measurements show acceptable accuracy or prediction beam quality quantity is above a given threshold value, the UE 100 may decide that the second beam prediction model M2 passed the sanity check and transmits the report 540 of e.g., predicted beam quantities and / or beams in the report 540 for N number of multiple future time instances TI3, as shown in Fig. 6.
[0135] However, if the second beam prediction model M2 is considered invalid, the report 540 may instead indicate a model failure for the second beam prediction model M2. The model failure may indicate any of a switching request for the second beam prediction model M2, or an alternative beam prediction model to the second beam prediction model M2, or a fallback beam prediction model to the second beam prediction model M2. For example, in case predicted beams comparison with subsequent measurements or prediction quality quantity, e.g. RSRPs or beams, is below a given threshold value, the UE 100 may decide model failure of the second beam prediction model M2.
[0136] The switching request may mean that the UE 100 requests the switching of the second beam prediction model M2, either to a new main model that will be signaled by the network or to another main model already available at the UE 100 e.g., through pre-configuration.
[0137] The alternative beam prediction model may be selected by the UE 100 and signaled to the gNB 300. Different main models may have different parameters, e.g., number of coefficients, prediction window, input size, etc. Consequently, subsequent adaptation of reference signals measurements and transmissions may be required hence the need to indicate the alternative beam prediction main model.
[0138] The fallback beam prediction model may be a main model that has been pre-configured as a fallback model, in case of active model invalidity or failure. In this case, when the UE 100 determines that the second beam prediction model M2 is invalid, the UE 100 triggers switching to the fallback main model and a simple invalidity indication in the report 540 is sufficient for the gNB 300 to know which beam prediction model is being used.
[0139] Moreover, the UE 100 may include a recovery reference signal request or an early reference signal transmission or measurement termination request in the report 540. When the UE 100 determines invalidity of the second beam prediction model M2, it can interrupt further measurements of input of the second beam prediction model M2 and request recovery reference signals, e.g., reference signals transmitted with wider beams. Additionally, the UE 100 may skip the rest of the measurements and request reference signal termination in order to reduce power consumption for both the UE 100 and the gNB 300. The requested reference signal termination may be indicated in the report 540.
[0140] In step VI in Fig. 7, the UE 100 transmits the report 540 to the gNB 300 e.g., in PUCCH.
[0141] In step VII in Fig. 7, the gNB 300 receives the report 540 from the UE 100. The gNB 300 demodulates and decodes the report 540 thereby deriving the information disclosed therein, i.e., information about the validity / invalidity of the second beam prediction model M2 and possible associated information.
[0142] If the second beam prediction model M2 was considered valid by the UE 100 in the report 540, the associated information indicates a beam prediction for a data transmission 550. Thus, the gNB 300 can select its beams and TCI states indications for subsequent data transmissions based on the indicated beams in the received report 540.
[0143] However, if the second beam prediction model M2 was considered invalid by the UE 100, the associated information indicates a model failure for the second beam prediction model M2. Depending on the type of model failure, the gNB 30 may use different procedures.
[0144] If the model failure indicates a switching request, the gNB 300 may indicate the identifier of an alternative main model to be used as a second beam prediction model at the UE 100. This alternative main model can be selected from a set of second beam prediction models already available at the UE 100 for this purpose or being transmitted to the UE 100 in control signaling, e.g., MAC CE or RRC.
[0145] If the model failure indicates an alternative beam prediction model, the gNB 300 may perform subsequent adaptation of reference signals measurements and transmissions in order to match the requirements of the new main model as indicated by the UE 100.
[0146] If the model failure indicates the fallback beam prediction model, the gNB 300 may perform subsequent adaptation of reference signals measurements and transmissions in order to match the requirements of the main model configured as the fall back model, e.g., reduced prediction window and number of reference signal measurements used as input for the main models.
[0147] [Rectified under Rule 91, 04.12.2024]In step VIII in Fig. 7, a data transmission 550 is performed from the gNB 300 to the UE 100. The beams used for the data transmission 550 and reception of the data transmission 550 are selected by the gNB 300, based on the report 540 from the UE 100, and are indicated using control signaling, e.g., TCI in DCI. Depending on whether the second beam prediction model M2 was determined as valid or invalid, the beams used for the data transmission 550 may differ. The data transmission 550 may be any suitable data transmission such as physical uplink control channel (PUCCH) and physical uplink shared channel (PUSCH) , scheduled through dynamic grant or in a semi-persistent manner. It may be noted that the same can be done for UL reference signals transmissions, such as sounding reference signals (SRS) .
[0148] A network access node herein may also be denoted as a radio network access node, an access network access node, an access point (AP) , or a base station (BS) , e.g., a radio base station (RBS) , which in some networks may be referred to as transmitter, “gNB” , “gNodeB” , “eNB” , “eNodeB” , “NodeB” or “B node” , depending on the standard, technology and terminology used. The radio network access node may be of different classes or types such as e.g., macro eNodeB, home eNodeB or pico base station, based on transmission power and thereby the cell size. The radio network access node may further be a station, which is any device that contains an IEEE 802.11-conformant media access control (MAC) and physical layer (PHY) interface to the wireless medium (WM) . The radio network access node may be configured for communication in 3GPP related long term evolution (LTE) , LTE-advanced, fifth generation (5G) wireless systems, such as new radio (NR) and their evolutions, as well as in IEEE related Wi-Fi, worldwide interoperability for microwave access (WiMAX) and their evolutions.
[0149] A client device herein may be denoted as a user device, a user equipment (UE) , a mobile station, an internet of things (IoT) device, a sensor device, a wireless terminal and / or a mobile terminal, and is enabled to communicate wirelessly in a wireless communication system, sometimes also referred to as a cellular radio system. The UEs may further be referred to as mobile telephones, cellular telephones, computer tablets or laptops with wireless capability. The UEs in this context may be, for example, portable, pocket-storable, hand-held, computer-comprised, or vehicle-mounted mobile devices, enabled to communicate voice and / or data, via a RAN, with another communication entity, such as another receiver or a server. The UE may further be a station, which is any device that contains an IEEE 802.11-conformant MAC and PHY interface to the WM. The UE may be configured for communication in 3GPP related LTE, LTE-advanced, 5G wireless systems, such as NR, and their evolutions, as well as in IEEE related Wi-Fi, WiMAX and their evolutions.
[0150] Furthermore, any method according to embodiments of the invention may be implemented in a computer program, having code means, which when run by processing means causes the processing means to execute the steps of the method. The computer program is included in a computer readable medium of a computer program product. The computer readable medium may comprise essentially any memory, such as previously mentioned a ROM, a PROM, an EPROM, a flash memory, an EEPROM, or a hard disk drive.
[0151] Moreover, it should be realized that the first communication device 100 and the second communication device 300 comprise the necessary communication capabilities in the form of e.g., functions, means, units, elements, etc., for performing or implementing embodiments of the invention. Examples of other such means, units, elements and functions are: processors, memory, buffers, control logic, encoders, decoders, rate matchers, de-rate matchers, mapping units, multipliers, decision units, selecting units, switches, interleavers, de-interleavers, modulators, demodulators, inputs, outputs, antennas, amplifiers, receiver units, transmitter units, DSPs, TCM encoder, TCM decoder, power supply units, power feeders, communication interfaces, communication protocols, etc. which are suitably arranged together for performing the solution.
[0152] Therefore, the processor (s) of the first communication device 100 and the second communication device 300 may comprise, e.g., one or more instances of a CPU, a processing unit, a processing circuit, a processor, an ASIC, a microprocessor, or other processing logic that may interpret and execute instructions. The expression “processor” may thus represent a processing circuitry comprising a plurality of processing circuits, such as e.g., any, some or all of the ones mentioned above. The processing circuitry may further perform data processing functions for inputting, outputting, and processing of data comprising data buffering and device control functions, such as call processing control, user interface control, or the like.
[0153] Finally, it should be understood that the invention is not limited to the embodiments described above, but also relates to and incorporates all embodiments within the scope of the appended independent claims.
Claims
1.A first communication device (100) configured to:measure a set of reference signals (530) transmitted from a set of transmitter beams (320) of a second communication device (300) and received in a set of receiver beams (120) of the first communication device (100) ;determine a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals (530) and a first beam prediction model (M1) ;determine a validity or an invalidity for a second beam prediction model (M2) based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals (530) and the second beam prediction model (M2) ; andtransmit a report (540) to the second communication device (300) , wherein the report (540) indicates:a beam prediction for a data transmission (550) when the second beam prediction model (M2) is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model (M2) .2.The first communication device (100) according to claim 1, wherein the report (540) indicates:a model failure for the second beam prediction model (M2) when the second beam prediction model (M2) is determined invalid.3.The first communication device (100) according to claim 2, wherein the model failure indicates a switching request for the second beam prediction model (M2) , or an alternative beam prediction model to the second beam prediction model (M2) , or a fallback beam prediction model to the second beam prediction model (M2) .4.The first communication device (100) according to any one of the preceding claims, wherein the first subset of measured reference signals is received in at least one first time instance and the second subset of measured reference signals is received in at least one second time instances after the first time instance.5.The first communication device (100) according to any one of the preceding claims, wherein the first subset of measured reference signals comprises K number of time instances and the set of measured reference signals (530) comprises M number of time instances, where K<M.6.The first communication device (100) according to claim 5, configured to:receive a first control message (510) from the second communication device (300) , the first control message (510) indicating at least one of K and M.7.The first communication device (100) according to any one of the preceding claims, configured to:determine the validity for the second beam prediction model (M2) based on a comparison of the first set of beam quality quantities with the second set of beam quality quantities and at least one threshold value.8.The first communication device (100) according to any one of the preceding claims, configured to:determine the validity for the second beam prediction model (M2) upon reception of a second control message (520) from the second communication device (300) , the second control message (520) indicating an activation of determining the validity of the second beam prediction model (M2) .9.The first communication device (100) according to any one of the preceding claims, configured to:determine the beam prediction for the data transmission (550) when the second beam prediction model (M2) is determined valid.10.The first communication device (100) according to any one of the preceding claims, wherein the first set of beam quality quantities and the second set of beam quality quantities comprises:RSRPs, SINRs or RSSIs for the set of transmitter beams (320) , the set of receiver beams (120) , or a set of transmitter beam (320) and receiver beam (120) pairs; oran ascending or a descending order of the set of transmitter beams (320) , the set of receiver beams (120) , or a set of transmitter beam (320) and receiver beam (120) pairs in terms of a RSRP, a SINR or a RSSI.11.The first communication device (100) according to any one of the preceding claims, wherein the beam prediction comprises:beam identities or reference signal resource indicators for the set of transmitter beams (320) , the set of receiver beams (120) , or a set of transmitter beam (320) and receiver beam (120) pairs; orRSRPs, SINRs or RSSIs for the set of transmitter beams (320) , the set of receiver beams (120) , or a set of transmitter beam (320) and receiver beam (120) pairs determined for one or multiple time instances following a prediction time instance for the beam prediction.12.A second communication device (300) configured to:transmit a set of reference signals (530) to a first communication device (100) in a set of transmitter beams (320) of the second communication device (300) and received in a set of receiver beams (120) of the first communication device (100) ; andreceive a report (540) from the first communication device (100) , wherein the report (540) indicates:a beam prediction for a data transmission (550) when a second beam prediction model (M2) is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model (M2) .13.The second communication device (300) according to claim 12, wherein the report (540) indicates:a model failure for the second beam prediction model (M2) when the second beam prediction model (M2) is determined invalid.14.The second communication device (300) according to claim 13, wherein the model failure indicates a switching request for the second beam prediction model (M2) , or an alternative beam prediction model to the second beam prediction model (M2) , or a fallback beam prediction model to the second beam prediction model (M2) .15.The second communication device (300) according to any one of claims 12 to 14, wherein the first subset of measured reference signals comprises K number of time instances and the set of measured reference signals (530) comprises M number of time instances, where K<M.16.The second communication device (300) according to claim 15, configured to:transmit a first control message (510) to the first communication device (100) , the first control message (510) indicating at least one of K and M.17.The second communication device (300) according to any one of claims 12 to 16, configured to:transmit a second control message (520) to the first communication device (100) , the second control message (520) indicating an activation of determining the validity of the second beam prediction model (M2) .18.The second communication device (300) according to any one of claims 12 to 17, wherein the beam prediction comprises:beam identities or reference signal resource indicators for the set of transmitter beams (320) , the set of receiver beams (120) , or a set of transmitter beam (320) and receiver beam (120) pairs; orRSRPs, SINRs or RSSIs for the set of transmitter beams (320) , the set of receiver beams (120) , or a set of transmitter beam (320) and receiver beam (120) pairs determined for one or multiple time instances following a prediction time instance for the beam prediction.19.A method (200) for a first communication device (100) , the method (200) comprising:measuring (202) a set of reference signals (530) transmitted from a set of transmitter beams (320) of a second communication device (300) and received in a set of receiver beams (120) of the first communication device (100) ;determining (204) a first set of beam quality quantities based on a first subset of measured reference signals among the set of measured reference signals (530) and a first beam prediction model (M1) ;determining (206) a validity or an invalidity for a second beam prediction model (M2) based on a comparison of the first set of beam quality quantities with a second set of beam quality quantities, wherein the second set of beam quality quantities is determined based on a second subset of measured reference signals among the set of measured reference signals (530) and the second beam prediction model (M2) ; andtransmitting (208) a report (540) to the second communication device (300) , wherein the report (540) indicates:a beam prediction for a data transmission (550) when the second beam prediction model (M2) is determined valid, wherein the beam prediction is determined based on the first subset of measured reference signals, the second subset of measured reference signals and the second beam prediction model (M2) .20.A method (400) for a second communication device (300) , the method (400) comprising:transmitting (402) a set of reference signals (530) to a first communication device (100) in a set of transmitter beams (320) of the second communication device (300) and received in a set of receiver beams (120) of the first communication device (100) ; andreceiving (404) a report (540) from the first communication device (100) , wherein the report (540) indicates:a beam prediction for a data transmission (550) when a second beam prediction model (M2) is determined valid, wherein the beam prediction is determined based on a first subset of measured reference signals, a second subset of measured reference signals and the second beam prediction model (M2) .21.A computer program with a program code for performing a method according to claim 19 or 20 when the computer program runs on a computer.