Information sending method and apparatus, information receiving method and apparatus, device, and storage medium

Abstract

A signal sending method and apparatus, a signal receiving method and apparatus, a device, and a storage medium, relating to the technical field of communications. The signal sending method is performed by a first device, and comprises: using a measurement result for a first resource set as an input of an encoding model, and the encoding model outputting sequence information, the first resource set comprising at least one reference signal resource; and sending the sequence information to a second device, the sequence information being used as an input of a decoding model, and the decoding model outputting a prediction result related to a spatial transmission filter. By using the encoding model to encode the measurement result at a sending end, the measurement result to be sent is compressed, thereby reducing the overhead of reporting the measurement result. Then, by using the decoding model to decode the received sequence information at a receiving end and performing prediction to obtain the prediction result, a beam suitable for the current communication environment can be decided on the basis of the prediction result.

Description

Information sending method, information receiving method, device, equipment and storage medium TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of communication, in particular to an information sending method, an information receiving method, a device, an equipment and a storage medium. BACKGROUND

[0002] In view of the great success of AI (Artificial Intelligence) technology, especially deep learning in computer vision, natural language processing and other aspects, the communication field begins to try to use deep learning to solve technical problems that traditional communication methods cannot solve.

[0003] Taking beam prediction as an example, a user equipment measures a reference signal, and then can infer an AI / ML model to obtain a beam suitable for a current communication environment according to a measurement result. However, how to use AI technology to realize beam prediction still needs further discussion and research.

[0004] SUMMARY

[0005] Embodiments of the present application provide an information sending method, an information receiving method, a device, an equipment and a storage medium. The technical solutions provided by the embodiments of the present application are as follows:

[0006] According to an aspect of the embodiments of the present application, an information sending method is provided, the method is executed by a first device, and the method comprises:

[0007] inputting a measurement result for a first resource set into an encoding model, the first resource set comprising at least one reference signal resource, and outputting sequence information from the encoding model;

[0008] sending the sequence information to a second device, the sequence information being used as input of a decoding model, and outputting a prediction result related to a spatial domain transmission filter from the decoding model.

[0009] According to an aspect of the embodiments of the present application, an information receiving method is provided, the method is executed by a second device, and the method comprises:

[0010] receiving sequence information sent by a first device, the sequence information being obtained by an encoding model based on a measurement result for a first resource set, the first resource set comprising at least one reference signal resource;

[0011] inputting the sequence information into a decoding model, and outputting a prediction result related to a spatial domain transmission filter from the decoding model.

[0012] According to an aspect of some embodiments of the present application, an information sending apparatus is provided, the apparatus comprising:

[0013] a processing module configured to output, as an input of an encoding model, a measurement result for a first resource set, the first resource set comprising at least one reference signal resource, and output, by the encoding model, sequence information;

[0014] a sending module configured to send, to a second device, the sequence information, the sequence information being used as an input of a decoding model, and output, by the decoding model, a prediction result related to a spatial domain transmission filter.

[0015] According to an aspect of some embodiments of the present application, an information receiving apparatus is provided, the apparatus comprising:

[0016] a receiving module configured to receive, from a first device, sequence information, the sequence information being obtained by an encoding model based on a measurement result for a first resource set, the first resource set comprising at least one reference signal resource;

[0017] a processing module configured to output, as an input of a decoding model, the sequence information, and output, by the decoding model, a prediction result related to a spatial domain transmission filter.

[0018] According to an aspect of some embodiments of the present application, a communication device is provided, the communication device comprising a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the information sending method or the information receiving method. The communication device is a first device, or the communication device is a second device.

[0019] According to an aspect of some embodiments of the present application, a computer readable storage medium is provided, the storage medium storing a computer program, and the computer program being configured to be executed by a processor to implement the information sending method or the information receiving method.

[0020] According to an aspect of some embodiments of the present application, a chip is provided, the chip comprising a programmable logic circuit and / or program instructions, and when the chip is running, the chip is configured to implement the information sending method or the information receiving method.

[0021] According to an aspect of some embodiments of the present application, a computer program product is provided, the computer program product comprising computer instructions, the computer instructions being stored in a computer readable storage medium, and a processor reading and executing the computer instructions from the computer readable storage medium to implement the information sending method or the information receiving method.

[0022] The technical solutions provided by the embodiments of the present application can have the following beneficial effects:

[0023] Based on the encoding model of the first device and the decoding model of the second device, optimal beam prediction based on a double-end coding and decoding model is realized. The first device encodes the measurement results of the resource set based on the encoding model to obtain sequence information, and sends the sequence information output by the encoding model to the second device. The second device decodes and predicts the received sequence information based on the decoding model to obtain a prediction result related to the beam. By using the encoding model to encode the measurement results at the sending end, the measurement results to be sent are compressed, reducing the overhead of reporting the measurement results. Then, by using the decoding model to decode and predict the received sequence information at the receiving end, a prediction result is obtained, and based on the prediction result, it is helpful to decide the beam suitable for the current communication environment. BRIEF DESCRIPTION OF DRAWINGS

[0024] FIG. 1 is a schematic diagram of a network architecture according to an embodiment of the present application;

[0025] FIG. 2 is a schematic diagram of a fully connected neural network according to an embodiment of the present application;

[0026] FIG. 3 is a schematic diagram of a CSI (Channel State Information) autoencoder based on AI according to an embodiment of the present application;

[0027] FIG. 4 is a schematic diagram of spatial domain beam prediction according to an embodiment of the present application;

[0028] FIG. 5 is a schematic diagram of time domain beam prediction according to an embodiment of the present application;

[0029] FIG. 6 is a schematic diagram of frequency domain beam prediction according to an embodiment of the present application;

[0030] FIG. 7 is a flowchart of an information sending method and an information receiving method according to an embodiment of the present application;

[0031] FIG. 8 is a flowchart of an information sending method and an information receiving method according to another embodiment of the present application;

[0032] FIG. 9 is a flowchart of an information sending method and an information receiving method according to another embodiment of the present application;

[0033] FIG. 10 is a flowchart of an information sending method and an information receiving method according to another embodiment of the present application;

[0034] FIG. 11 is a block diagram of an information sending apparatus according to an embodiment of the present application;

[0035] FIG. 12 is a block diagram of an information receiving apparatus according to an embodiment of the present application;

[0036] FIG. 13 is a schematic diagram of a communication device according to an embodiment of the present application. DETAILED DESCRIPTION

[0037] For the purposes of the present application, the technical solutions and advantages will be more apparent from the following detailed description of the embodiments of the present application, taken in conjunction with the accompanying drawings.

[0038] The network architecture and service scenarios described in the embodiments of the present application are used to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that, as the network architecture evolves and new service scenarios appear, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

[0039] Please refer to FIG. 1, which shows a schematic diagram of a network architecture 100 provided by an embodiment of the present application. The network architecture 100 can include a terminal device 10, an access network device 20 and a core network element 30.

[0040] The terminal device 10 can refer to a UE (User Equipment), a STA (Station), an access terminal, a user unit, a user station, a mobile station, a mobile station, a remote station, a remote terminal, a mobile device, a wireless communication device, a user agent or a user apparatus. In some embodiments, the terminal device 10 can also be a cellular phone, a cordless phone, a SIP (Session Initiation Protocol) phone, a WLL (Wireless Local Loop) station, a PDA (Personal Digital Assistant), a handheld device with wireless communication function, a computing device or other processing device connected to a wireless modem, a vehicle-mounted device, a wearable device, a terminal device in a 5GS (5th Generation System) or a terminal device in a future evolved PLMN (Public Land Mobile Network), etc., and the embodiments of the present application are not limited thereto. For the convenience of description, the above-mentioned devices are collectively referred to as terminal devices. The number of terminal devices 10 is usually multiple, and one or more terminal devices 10 can be distributed in a cell managed by each access network device 20. The terminal device can also be referred to as a terminal or a UE, and those skilled in the art can understand its meaning.

[0041] The access network device 20 is a device deployed in an access network to provide wireless communication functions for the terminal device 10. The access network device 20 can include various forms of macro base stations, micro base stations, relay stations, APs (Access Points), and the like. In systems using different wireless access technologies, the names of devices with access network device functions can be different, for example, in a 5G NR (New Radio) system, it is called gNodeB or gNB (Next Generation Node B). With the evolution of communication technology, the name of the "access network device" can change. For ease of description, in the embodiments of the present application, the above-mentioned devices that provide wireless communication functions for the terminal device 10 are collectively referred to as access network devices. In some embodiments, through the access network device 20, a communication relationship can be established between the terminal device 10 and the core network element 30. Illustratively, in the LTE (Long Term Evolution) system, the access network device 20 can be an EUTRAN (Evolved Universal Terrestrial Radio Access Network) or one or more eNodeBs in the EUTRAN; in the 5G NR system, the access network device 20 can be a RAN (Radio Access Network) or one or more gNBs in the RAN. In the embodiments of the present application, the "network device" refers to the access network device 20, such as a base station, unless otherwise specified.

[0042] The core network element 30 is a network element deployed in the core network, and the main functions of the core network element 30 are to provide user connection, manage users, and complete bearer for services, and provide an interface to external networks as a bearer network. For example, the core network element in the 5G NR system can include AMF (Access and Mobility Management Function) entities, UPF (User Plane Function) entities, and SMF (Session Management Function) entities.

[0043] In some embodiments, the access network device 20 and the core network element 30 communicate with each other through some air interface technology, such as the NG interface in the 5G NR system. The access network device 20 and the terminal device 10 communicate with each other through some air interface technology, such as the Uu interface.

[0044] The "5G NR system" in the embodiments of the present application can also be referred to as a 5G system or an NR system, but those skilled in the art can understand its meaning. The technical solutions described in the embodiments of the present application can be applicable to the LTE system, and can also be applicable to the 5G NR system, and can also be applicable to the subsequent evolution system (for example, the B5G (Beyound 5G) system, the 6G system (6th Generation System, the sixth generation mobile communication system)) of the 5G NR system, and can also be applicable to other communication systems such as the NB-IoT (Narrow Band Internet of Things, Narrow Band Internet of Things) system, and the present application does not limit this.

[0045] In the embodiments of the present application, the network device can provide services for a cell, and the terminal device communicates with the network device through the transmission resource (for example, the frequency domain resource, or the spectrum resource) on the carrier used by the cell. The cell can be a cell corresponding to the network device (for example, a base station), and the cell can belong to a macro base station or a base station corresponding to a small cell (Small cell). The small cell here can include a metro cell, a micro cell, a pico cell, a femto cell, etc. These small cells have the characteristics of small coverage and low transmit power, and are suitable for providing high-speed data transmission services.

[0046] Before introducing the technical solutions of the present application, some related technical knowledge involved in the present application will be introduced and explained. The following related technologies can be combined with the technical solutions of the embodiments of the present application as optional solutions, and all belong to the protection scope of the embodiments of the present application. The embodiments of the present application include at least part of the following contents.

[0047] I. Measurement and reporting of NR beam management

[0048] In related technologies, the beam management mechanism is standardized.

[0049] Beam management mainly includes beam related measurement and reporting, and beam indication by NW (Network) according to UE's reporting. Beam measurement is based on periodic, semi-persistent and aperiodic reference signals. The reference signals are CSI-RS (CSI-Reference Signal) or SSB (Synchronization Signal and PBCH Block). UE measures one or more SSB resource sets and / or CSI-RS resource sets. The best N (1, 2 or 4) reference signal resource indexes (i.e. SSB RI (SSB Resource Indicator) or CRI (CSI-RS Resource Indicator)) and corresponding link quality (with L1-RSRP (L1-Reference Signal Received Power) or L1-SINR (L1-Signal to Inference and Noise Ratio) as indicators) are reported to NW.

[0050] The specific reporting format is shown in Table 1, where L1-RSRP is reported in a differential manner.

[0051] Beam reporting is associated with measurement resources, which include periodic beam reporting, such as using PUCCH (Physical Uplink Control Channel) resources for reporting; beam reporting can also be semi-persistent reporting, such as using PUCCH or PUSCH resources for reporting, and semi-persistent reporting can use MAC CE (Media Access Control-Control Element) to activate or deactivate; beam reporting can also be aperiodic reporting, such as using PUSCH (Physical Uplink Shared Channel) resources, and aperiodic measurement reporting is triggered based on NW sending DCI (Downlink Control Information).

[0052] Table 1: UCI (Uplink Control Information) format for traditional beam measurement reporting

[0053] When the beam measurement report is encoded, the UE will insert CRC (Cyclic Redundancy Check) check bits to facilitate the NW to determine whether the decoding of the report is correct.

[0054] II. Measurement and reporting of LTM (L1 / L2 Triggered Mobility)

[0055] In the related art, the NR system only supports layer 3 (i.e., RRC (Radio Resource Control) layer) mobility. The NR standardizes the enhancement of layer 1 / layer 2 triggered mobility (LTM). From the perspective of the physical layer (i.e., L1), the UE pre-measures multiple candidate cells (including the current serving cell) through the downlink measurement resources configured by the NW, and then reports the measurement results to the NW. The NW indicates the selected candidate cell (i.e., target cell) and / or indicates the candidate beam through the cell switching command.

[0056] For measurement and reporting, the UE can measure in multiple candidate cells (including the current serving cell) configured by the NW, and give at most M x L (depending on the UE capability, L is the number of candidate cells, and M is the number of beams per cell) measurement results in one report. Obviously, for the measurement of a large number of candidate cells, a large amount of reference signal overhead is required, and the time delay required for the UE to measure these reference signals, after all, the UE cannot measure multiple beams of multiple candidate cells at the same time.

[0057] When the UE completes the measurement and reports the measurement results to the NW, the NW can indicate the UE to switch to the target cell through the MAC CE of the CSC (Cell Switch Command), which at least includes the target cell information (such as candidate cell index) and / or the selected beam information (i.e., indicated by a unified TCI state ID).

[0058] III. Measurement and reporting of data collection

[0059] In the AI / ML-based beam management in the related art, data collection for training the model often requires the UE to perform a large amount of measurement and reporting (assuming the model on the NW side).

[0060] For example, for the training of the model, the measurement results of Set B are needed as the input of the model, and Set A is needed as the label of the output of the model. The input of the model and the corresponding label can be considered as a sample of data. The training of the model requires a large number of such samples.

[0061] For BM-Case 1 (use case of spatial domain beam prediction), Set B is a smaller set of measurements relative to Set A, such as a set of reference signals corresponding to only 8 measurement beams. Set A can be considered as a full set of beam sets, such as a set of 64 predicted beams.

[0062] IV. Introduction of AI / ML (Maching Learning) Model

[0063] Deep Neural Network

[0064] A simple neural network is shown in FIG. 2, which includes an input layer, a hidden layer, and an output layer. Different connections, weights, and activation functions of multiple neurons can produce different outputs, thereby fitting the mapping relationship from input to output. Each upper node is connected to all lower nodes. This full connection model can also be called a DNN (Deep Neural Networks) in the present case, i.e., a deep neural network. The DNN model can perform spatial domain beam prediction.

[0065] V. AI / ML-based CSI feedback method

[0066] In view of the great success of AI technology, especially deep learning in computer vision, natural language processing, etc., the communication field has begun to try to use deep learning to solve technical problems that are difficult to solve by traditional communication methods. The neural network architecture commonly used in deep learning is nonlinear and data-driven, which can extract features from actual channel matrix data and restore the channel matrix information compressed and fed back by the UE side at the base station side as much as possible. This ensures the restoration of channel information and also provides the possibility for the UE side to reduce the CSI feedback overhead.

[0067] For example, the CSI feedback based on deep learning regards channel information as a "image" to be compressed, uses a deep learning auto-encoding model to compress and feed back the input channel information, and reconstructs the compressed channel "image" at the receiving end, which can preserve channel information to a greater extent.

[0068] Using the AI-based CSI auto-encoding model method, the entire feedback system is divided into an encoding model and a decoding model, which are deployed at the UE (as the sending end of the CSI feedback) and the NW (as the receiving end of the CSI feedback), respectively. Therefore, it can also be called a two-end model based on AI / ML.

[0069] Specifically, the UE obtains the channel information through channel estimation, and inputs the channel information into the encoding model. The neural network of the encoding model compresses and encodes the channel information matrix, and feeds back the compressed bit sequence to the NW through the air interface feedback link. Correspondingly, the NW recovers the channel information according to the feedback bit sequence through the decoding model, and outputs the complete feedback channel information.

[0070] The AI / ML model of the encoding model and the decoding model shown in FIG. 3 can adopt a DNN composed of multiple full connection layers, or a CNN (Convolutional Neural Networks) composed of multiple convolution layers, or an RNN (Recurrent Neural Network) with structures such as LSTM (Long Short-Term Memory) and GRU (Gate Recurrent Unit), or various neural network architectures such as residual and self-attention mechanisms to improve the performance of the encoding model and the decoding model.

[0071] The above-mentioned CSI input and output can be full channel information, or feature vector information obtained based on the full channel information. For the feedback method based on the feature vector, it is the feedback architecture supported in the current NR system. It can achieve higher CSI feedback accuracy with the same feedback bit overhead, or significantly reduce the feedback overhead under the condition of achieving the same CSI feedback accuracy.

[0072] Six, AI / ML-based beam management use cases

[0073] In the ongoing 3GPP R18 discussion, AI / ML-based beam management is one of the main use cases of R18 AI project. RAN1 is preparing to standardize spatial and / or temporal beam prediction mechanisms in R19. The beam prediction in the spatial domain (BM-Case1) and the beam prediction in the temporal domain (BM-Case2) are defined as follows.

[0074] Agreement protocol

[0075] For AI / ML-based beam management, support BM-Case1 and BM-Case2 for characterization and baseline performance evaluations

[0076] •BM-Case 1: Spatial-domain DL beam prediction for Set A of beams based on measurement results of Set B of beams (as shown in Figure 4, the prediction result of spatial-domain beam prediction based on measurement results of Set B includes at least one beam in different directions in Set A)

[0077] •BM-Case 2: Temporal DL beam prediction for Set A of beams based on the historic measurement results of Set B of beams (as shown in Figure 5, the prediction result based on the temporal beam prediction of Set B includes beams at at least one moment in Set A)

[0078] • FFS: Details of BM-Case 1 and BM-Case 2

[0079] •FFS: other sub-use cases

[0080] Note: For BM-Case 1 and BM-Case 2, beams in Set A and Set B can be in the same frequency range.

[0081] Here, Set B is the UE's measurement set, which serves as the model's input; Set A is the model's prediction set, which is related to the model's output. For some BM-Case 1 / BM-Case 2 model training, it is necessary to collect the measurement results of Set B (as the model's input) and the measurement results of Set A (as the model's labels).

[0082] Finally, for the beam prediction in frequency domain, there is no related work in the current 3GPP standardization process, but it is possible to enter the standardization in the future with the prediction type of AI / ML use case. Specifically, as shown in FIG. 6, the measurement set Set B is on one carrier or band, but the prediction set of the optimal beam is on another carrier or band. The benefit of this is that no measurement resource can be configured on some measurement resource limited CC / band (carrier / frequency band), so that the optimal beam information and / or link quality on the CC / band can be predicted by the AI / ML model (with the measurement of other CC / band as input).

[0083] For the beam prediction mechanism standardized in R19, whether it is BM-case 1 or BM-case 2, it is based on a single-side AI / ML model, which can be deployed on the UE side or on the NW side.

[0084] When the model is deployed on the UE side, the UE completes the measurement of Set B, and uses the UE-side model to complete the inference of the optimal beam information and / or link quality. Finally, the result of predicting the optimal beam is reported to the NW. Once an error occurs, according to the existing protocol, the UE has the opportunity to complete the optimal beam reporting only when it reports next time.

[0085] When the model is deployed on the NW side, the UE completes the measurement of Set B and reports the measurement result to the NW, so that the NW completes the inference of the optimal beam information and / or link quality. Similarly, the reporting of Set B can also have errors. Once the reporting of Set B has an error, the model on the NW side cannot use Set B to perform inference.

[0086] However, whether it is a UE-side model or a NW-side model, the reporting content is encoded according to the traditional way of source encoding, channel encoding, modulation, etc. This has the following disadvantages.

[0087] (1) The correlation between the reporting contents can have a lot of redundancy, which can be reduced by the compression of the two-end model, and the uplink overhead can be reduced;

[0088] (2) The source encoding and channel encoding of the reporting content are independent, and there is no dynamic balance between reporting redundancy and error protection.

[0089] Please refer to FIG. 7, which shows the flowchart of the information sending method and the information receiving method provided by an embodiment of the present application. The method is executed by the interaction of the first device and the second device, and the method includes at least one of the steps 710-730.

[0090] At step 710, the first device outputs sequence information from the encoding model, taking the measurement results of the first resource set as input of the encoding model, the first resource set including at least one reference signal resource.

[0091] A reference signal resource is a resource for transmitting a reference signal, such as a time-frequency resource for transmitting a reference signal. A time-frequency resource refers to a resource allocated in the time domain and the frequency domain. Each reference signal resource can occupy a certain time domain resource and a certain frequency domain resource for transmitting a reference signal. The first resource set can include one or more reference signal resources.

[0092] A reference signal is a known signal provided by a transmitting end to a receiving end for channel estimation or channel sounding. Illustratively, an uplink reference signal is used for uplink channel estimation or uplink channel quality measurement. For example, the uplink reference signal can be a DM-RS (Demodulation Reference Signal), a SRS (Sounding Reference Signal), etc. The DM-RS is associated with the transmission of PUSCH and PUCCH, and is used to obtain a channel estimation matrix to help demodulation of the two channels. The SRS is transmitted independently, and is used for estimation of uplink channel quality and channel selection, and calculation of the SINR of the uplink channel. Illustratively, a downlink reference signal is used for downlink channel estimation, downlink channel quality measurement, or cell search. For example, a downlink reference signal can be a CRS (Cell Reference Signal), a PRS (Positioning Reference Signal), a CSI-RS, or an SSB. The CRS (cell-specific reference signal, also called a common reference signal) is used for channel estimation and related demodulation of all downlink transmission technologies except for beamforming technologies based on a codebook. Cell-specific means that the reference signal corresponds to an antenna port (antenna ports 0-3) of a network device. The PRS is a positioning reference signal, and is used to assist the receiving end in positioning. The CSI-RS is a channel state information reference signal, and is used to obtain channel state information. The SSB is used for initial synchronization and cell search.

[0093] In some embodiments, the first device is a terminal device, and the second device is a network device. The reference signal is a downlink reference signal transmitted by the network device to the terminal device, such as a CSI-RS.

[0094] In some embodiments, the reference signal is used for measurement related to a spatial domain transmission filter.

[0095] Spatial domain filtering refers to a technology that when several signals stacked together in time domain occupy the same frequency band, beam forming utilizes the spatial domain separation of signals from different directions to achieve spatial domain processing of signals. The concept of spatial domain filtering is proposed in the field of digital beam forming technology (DBF) in radar technology. It refers to forming a main lobe beam in a specific direction to receive useful signals and suppress interference signals from other directions. Because when several signals stacked together or arrived at the same time occupy the same frequency band, the general time domain filtering and frequency domain filtering cannot separate them, but these signals are generally from different directions, and beam forming is to utilize the spatial domain separation to achieve spatial domain processing of signals. Its essence is a multi-channel array signal processing system, which is a concept of spatial domain filtering. In the embodiments of the present application, the spatial domain transmission filter can also be referred to as a beam. In measurement and reporting, the downlink reference signal (such as CSI-RS or SSB) is used to associate the transmit beam. In the beam indication process, the TCI state (which contains the downlink reference signal) is used to refer to the transmit beam.

[0096] In some embodiments, the link corresponding to the reference signal resource is a beam. It can be understood that the reference signal is transmitted by using the reference signal resource, that is, the reference signal resource is used and the reference signal is transmitted by using the corresponding beam. It should be noted that the links corresponding to different reference signal resources can be the same beam or different beams. For example, the link corresponding to the reference signal resource 1 is the beam 1, the link corresponding to the reference signal resource 2 is the beam 2, and the link corresponding to the reference signal resource 3 is also the beam 1. In the digital beam forming technology, there are two different types of beams, namely wide beam and narrow beam. Among them, the wide beam is used to receive signals from a larger angle range, and has the characteristics of large coverage range, low gain and weak anti-interference ability; while the narrow beam is used to receive signals from a specified direction, and has the characteristics of small coverage range, high gain and strong anti-interference ability.

[0097] In some embodiments, the encoding model is an AI or ML model. An AI model refers to a mathematical representation obtained by running a deep learning algorithm based on an existing data set. It is a mathematical model that turns data into intelligence. AI models can be used to analyze, process, predict, and optimize data with certain regularity and predictability. AI models need to be professionally trained and trained before being applied in different applications. Through continuous data feeding to the machine and optimization of framework algorithms, the corresponding computing power is matched to complete the task. AI models have the characteristics of being powerful, efficient, and flexible. A ML model is an intelligent file that predefines an algorithm, can learn a specific pattern of data set, and extract insights from it to make predictions. When creating a ML model, the answer to be obtained needs to be defined, and the working and learning parameters of the model need to be set. When the ML model starts processing new data, actionable insights can be obtained. The training data must contain the correct answers, which are called targets or target attributes. The learning algorithm finds patterns in the training data that map input data attributes to targets, and then outputs a ML model that captures these patterns. AI models are classified based on learning methods, which can be divided into supervised learning, unsupervised learning, and reinforcement learning.

[0098] In some embodiments, the first device measures the N reference signal resources included in the first resource set to obtain measurement result data of each of the N reference signal resources, where N is a positive integer greater than 1. The measurement result data is obtained by receiving the reference signal in a specified beam to obtain performance parameters and quality parameters related to the specified beam. In some embodiments, the measurement result of the first resource set can include measurement result data of all or part of the N reference signal resources. In some embodiments, the measurement result data of the reference signal resource includes: an index of the reference signal resource, and / or link quality information corresponding to the reference signal resource.

[0099] The index of the reference signal resource is used to distinguish different reference signal resources. In some embodiments, the link quality information corresponding to the reference signal resource includes at least one of the following: L1-RSRP, L1-SINR, L1-RSRQ (L1-Reference Signal Received Quality), and L1-RSSI (L1-Received Signal Strength Indicator). L1-RSRP refers to the average received power of the reference signal in the first resource set. L1-SINR is used to represent the ratio of received signal strength to interference and noise. L1-RSRQ is used to identify the ratio of signal quality to all received signals. L1-RSSI is used to represent the total power of all signals received within a specified time.

[0100] In some embodiments, the sequence information does not have corresponding check bits. In some embodiments, the sequence information does not have CRC parity check. Since the input of the encoding model is the measurement result of the first resource set, but the output of the decoding model is the prediction result related to the beam, it can be seen that there is not a one-to-one relationship between the two. Even if the input of the decoding model has a certain degree of deviation from the output of the encoding model, the decoding model has a great possibility to make correct prediction, which is the benefit brought by the robustness of the model itself. The purpose of the CRC check bit is to see whether the receiving end can correctly recover the content consistent with the sending end, so it is not applicable to the scheme in the present application which makes prediction by reporting measurement.

[0101] In some embodiments, the sequence information can only be decoded by the decoding model on the second device side.

[0102] In the above manner, by not including check bits in the reported sequence information, the additional overhead of carrying the transmission content of the sequence information can be saved.

[0103] At step 720, the first device sends the sequence information to the second device, and the sequence information is used as the input of the decoding model, and the decoding model outputs the prediction result related to the spatial domain transmission filter.

[0104] Correspondingly, the second device receives the sequence information sent by the first device.

[0105] At step 730, the second device takes the sequence information as the input of the decoding model, and the decoding model outputs the prediction result related to the spatial domain transmission filter.

[0106] In some embodiments, the decoding model is an AI or ML model. In some embodiments, the decoding model and the encoding model can have the same model structure or different model structures, which are not limited in the present application. Illustratively, the encoding model adopts a fully connected model structure, and the decoding model adopts a CNN model structure. Illustratively, the encoding model and the decoding model both adopt a fully connected model structure, but the number of hidden layers in the two is different.

[0107] In some embodiments, the prediction result is predicted based on at least one of the following domains: spatial domain, time domain, frequency domain. The spatial domain, also known as the space domain, refers to a way of signal processing using spatial position or spatial direction. The time domain, also known as the time domain, refers to a way of managing and allocating time resources by controlling the transmission timing of signals. The frequency domain, also known as the frequency domain, refers to a way of signal processing and resource allocation by allocating and adjusting different frequency resources.

[0108] In some embodiments, the prediction result includes:

[0109] related information of at least one optimal spatial domain transmission filter; or,

[0110] information of at least one optimal spatial domain transmission filter in each time instance; or

[0111] information of at least one optimal spatial domain transmission filter in each carrier or frequency band; or

[0112] information of at least one optimal spatial domain transmission filter in each time instance, each carrier or frequency band;

[0113] The at least one carrier or frequency band is different from the carrier or frequency band corresponding to the first resource set.

[0114] The optimal spatial domain transmission filter, also referred to as an optimal beam, is a predicted beam obtained for the purpose of maximizing the strength or quality of a received signal.

[0115] In some embodiments, the optimal spatial domain transmission filter can be determined in the following ways:

[0116] (1) The optimal spatial domain transmission filter is selected from spatial domain transmission filters corresponding to a second resource set, wherein the second resource set includes at least one reference signal resource.

[0117] (2) The optimal spatial domain transmission filter is predicted by a decoding model.

[0118] For way (1), the second resource set includes at least one reference signal resource, and the decoding model selects at least one optimal spatial domain transmission filter from the at least one reference signal resource after performing inference analysis on the sequence information.

[0119] For way (2), the decoding model directly outputs information of at least one optimal spatial domain transmission filter after performing inference analysis on the sequence information. The at least one optimal spatial domain transmission filter directly outputted can be different from any spatial domain transmission filter included in the second resource set. Since the number of reference signal resources included in the first resource set is not large, it is not necessary to select the optimal beam from the second resource set, because the decoding model can directly predict the optimal beamforming direction according to the measurement result. At this time, the beam direction predicted by the decoding model can be different from the direction of any beam in the second resource set. It should be noted that in the present application, inference, deduction or prediction mean the same.

[0120] In some embodiments, the related information includes at least one of the following: identification information, direction information, and link quality information.

[0121] In some embodiments, the identification information of the optimal spatial domain transmission filter can be indicated by an index of a reference signal resource. In some embodiments, the index of the reference signal resource can be represented by a reference signal index of the first resource set, for example, can be represented by an SSBRI or a CRI. The SSBRI (Synchronization Signal Block Resource Indicator) is used to indicate the resource index of the synchronization signal block. The CRI (CSI-RS Resource Indicator) is used to indicate the index of the CRI-RS resource. In some embodiments, the index of the reference signal resource can also be represented by an index of the second resource set on a carrier (CC), for example, can be represented by a CSI-RS resource ID or an SSB index of the second resource set on the CC. The CSI-RS resource ID is used to distinguish different CSI-RS resources. The SSB index is used to distinguish different synchronization signal blocks.

[0122] In some embodiments, the identification information of the optimal spatial domain transmission filter can also be indicated by a logical index.

[0123] For example, if the total number of beams in the second resource set is 64, the value range of the identification information of the optimal beam is 0 to 63.

[0124] In this way, since the reference signal resource corresponding to the second resource set does not need to be configured and / or transmitted in the feedback information when predicting the result to the first device, the overhead of link transmission is reduced.

[0125] The direction information of the optimal spatial domain transmission filter refers to the direction of one or more optimal beamforming predicted by the decoding model on the second device side. Specifically, the beamforming direction can be expressed as the direction of the main lobe and the coverage direction of the side lobe of the beamforming in the vertical and horizontal dimensions in the three-dimensional space, etc. In some embodiments, the direction information includes at least one of the following: the Azimuth angle and the Zenith Angle of the first device relative to the second device, and the channel state information (CSI).

[0126] In some embodiments, the first device can be a terminal device or a network device. In some embodiments, the second device can be a terminal device or a network device. In some embodiments, the communication between the first device and the second device can be uplink communication, downlink communication, or sidelink communication, which is not limited in the present application. In some embodiments, if the first device is a network device and the second device is a terminal device, the reference signal in step 710 can be implemented as an uplink reference signal. In some embodiments, if the first device and the second device are both terminal devices, the reference signal in step 710 can be implemented as a sidelink reference signal. In some embodiments, if the first device is a terminal device and the second device is a network device, the reference signal in step 710 can be implemented as a downlink reference signal.

[0127] The scheme provided by the embodiments of the present application realizes optimal beam prediction based on a double-end coding and decoding model. The first device encodes the measurement results of the resource set based on the encoding model to obtain sequence information, and sends the sequence information output by the encoding model to the second device. The second device decodes and predicts the received sequence information based on the decoding model to obtain a prediction result related to the beam. By using the encoding model to encode the measurement results at the sending end, the measurement results to be sent are compressed, and the overhead of reporting the measurement results is reduced. Then, by using the decoding model to decode and predict the received sequence information at the receiving end, a prediction result is obtained, and based on the prediction result, it is helpful to decide the beam suitable for the current communication environment.

[0128] In some embodiments, the encoding model has several encoding methods for the measurement results, and different encoding methods will bring different effects, which are classified and explained in the following embodiments.

[0129] Method one, separate source channel coding

[0130] Separate source channel coding (SSCC) is a coding technology that separates source coding and channel coding.

[0131] In some embodiments, the encoding model is used for source coding of the measurement results.

[0132] In some embodiments, after the encoding model performs source coding on the measurement results, the output sequence information is a bit sequence.

[0133] In some embodiments, the above step 720 can be implemented as step 721.

[0134] Step 721, after the first device performs channel coding, modulation, resource mapping and antenna mapping on the sequence information, the sequence information is sent to the second device.

[0135] Correspondingly, after the second device performs radio frequency reception, resource demapping, demodulation and channel decoding on the signal from the first device, the sequence information is obtained.

[0136] In some embodiments, the decoding model is used to source decode the sequence information to obtain the prediction result.

[0137] Exemplarily, refer to FIG. 8, which shows the flowchart of the information sending method and the information receiving method provided by another embodiment of the present application.

[0138] After the first device performs relevant measurement, the measurement result of the first resource set is taken as the input of the encoding model, and the encoding model outputs the sequence information. Then, after the sequence information is subjected to the conventional channel coding (such as LDPC (Low-density Parity-check), Polar, Block Code, etc.), modulation, resource mapping, antenna mapping and other processes, it is sent to the second device. Correspondingly, after the second device performs radio frequency reception, resource demapping, demodulation and channel decoding on the received signal, the sequence information is obtained. The radio frequency signal carrying the sequence information reaches the second device through a wireless channel. After the second device performs beam prediction on the sequence information through the decoding model, the prediction result is obtained.

[0139] In some embodiments, the above method can be applied in different beam prediction scenarios. The following are examples of several possible application scenarios.

[0140] Example 1, beam prediction based on spatial domain

[0141] Taking the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is the measurement result of the first resource set based on the spatial domain, and the output is a bit sequence.

[0142] The input of the decoding model on the NW side is the bit sequence from the UE side, and the output is the related information of at least one optimal spatial domain transmission filter. The at least one optimal spatial domain transmission filter can be selected from the second resource set or directly determined by the encoding model.

[0143] In some embodiments, the number of reference signal resources included in the first resource set is less than the number of reference signal resources included in the second resource set. In this way, the decoding model can predict the optimal spatial domain transmission filter and the link quality from a larger range through a small amount of measurement results.

[0144] In some embodiments, the first set of resources is a subset of the second set of resources. For example, if the first set of resources contains 8 downlink reference signal resources, and the downlink reference signal resources are CSI-RS resources transmitted using narrow beams, and the second set of resources contains 64 downlink reference signal resources, and the downlink reference signal resources are also CSI-RS resources transmitted using narrow beams, then the first set of resources is a subset of the second set of resources.

[0145] In some embodiments, the reference signal resources included in the first set of resources are different from the reference signal resources included in the second set of resources. For example, if the first set of resources contains 4 downlink reference signal resources, and the downlink reference signal resources are SSB resources transmitted using wide beams, and the second set of resources contains 32 downlink reference signal resources, and the downlink reference signal resources are also CSI-RS resources transmitted using narrow beams, then the reference signal resources included in the first set of resources are different from the reference signal resources included in the second set of resources.

[0146] In some embodiments, the encoding model directly outputs the prediction result.

[0147] In some embodiments, the prediction result includes information related to at least one optimal spatial domain transmission filter.

[0148] Example 2, beam prediction based on time domain

[0149] Take the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is the measurement result of the first set of resources based on the time domain, and the output is a bit sequence.

[0150] The input of the decoding model on the NW side is the bit sequence from the UE side, and the output is information related to at least one optimal spatial domain transmission filter at each time.

[0151] In some embodiments, the at least one optimal spatial domain transmission filter is selected from the second set of resources, where the first set of resources and the second set of resources are the same set of resources. Each optimal spatial domain transmission filter corresponds to a downlink reference signal resource index or a logical index.

[0152] In some embodiments, the decoding model on the NW side directly infers the information related to the at least one optimal spatial domain transmission filter. Understandably, at this time, the at least one optimal spatial domain transmission filter does not need to be selected from the second set of resources. The information related to the at least one optimal spatial domain transmission filter includes direction information.

[0153] In some embodiments, the measurement results of the first resource set include measurement results at K historical time instants, K being a positive integer greater than 0. Understandably, the measurement results of the first resource set are obtained by measuring the at least one reference signal resource included in the first resource set at the K historical time instants respectively.

[0154] In some embodiments, the input of the encoding model is all or part of the measurement results of the first resource set. For example, the measurement results at the K historical time instants are input into the encoding model at the UE side together. For another example, the measurement results at the K historical time instants can be input into the encoding model at the UE side in batches (such as K times).

[0155] In this way, when the measurement results at multiple historical time instants have certain correlations, the encoder can compress the correlations, thereby saving the reporting overhead.

[0156] Example 3, beam prediction based on frequency domain

[0157] Taking a terminal device (such as a UE) as the first device and a network device (such as a NW) as the second device as an example. The input of the encoding model at the UE side is the measurement results of the first resource set based on the frequency domain, and the output is a bit sequence.

[0158] The input of the decoding model at the NW side is the bit sequence from the UE side, and the output is related information of at least one optimal spatial domain transmission filter on each carrier or frequency band in at least one carrier or frequency band.

[0159] Similarly, the at least one optimal spatial domain transmission filter can be selected from the second resource set, or can be directly inferred by the decoding model.

[0160] In some embodiments, in the case that the at least one optimal spatial domain transmission filter can be selected from the second resource set, the carrier or frequency band corresponding to the first resource set is different from the carrier or frequency band corresponding to the second resource set.

[0161] In some embodiments, in the case that the decoding model performs beam prediction based on the time domain and the spatial domain, the reference signal resource included in the first resource set is different from the reference signal resource included in the second resource set; or the first resource set is a subset of the second resource set.

[0162] In some embodiments, in the case that the decoding model performs beam prediction based on the time domain, the spatial domain and the frequency domain, the first resource set belongs to a subset of the second resource set; or the reference signal resource included in the first resource set is different from the reference signal resource included in the second resource set, and the carrier or frequency band corresponding to the first resource set is different from the carrier or frequency band corresponding to the second resource set.

[0163] In some embodiments, the prediction result comprises: information about at least one optimal spatial domain transmission filter at each time point, each carrier or frequency band in at least one time point and at least one carrier or frequency band.

[0164] In this case, the beam prediction needs to consider the time domain, the spatial domain and the frequency domain at the same time, and the prediction result obtained is information about the optimal spatial domain transmission filter specific to the time domain, the frequency domain and the spatial domain.

[0165] It should be noted that the decoding model can perform beam prediction based on any two or all of the time domain, the spatial domain and the frequency domain, and the embodiments of the present application do not limit this.

[0166] In the above manner one, the first device transmits the measurement result of the first resource set to the second device after compression by using the encoding model, and the second device decodes and predicts the received measurement result, which can save the overhead of the first device in transmitting the measurement result.

[0167] Manner two, joint source channel coding

[0168] Joint source channel coding (JSCC) is a coding technology that simultaneously processes source coding and channel coding.

[0169] In some embodiments, the encoding model is used to perform joint source channel coding on the measurement result.

[0170] In some embodiments, the encoding model performs joint source channel coding on the measurement result, and the output sequence information is a bit sequence.

[0171] In some embodiments, the above step 720 can be implemented as the following step 722.

[0172] Step 722, the first device transmits the sequence information to the second device after modulation, resource mapping and antenna mapping.

[0173] Correspondingly, the second device obtains the sequence information after radio frequency receiving, resource demapping and demodulation of the signal from the first device.

[0174] In some embodiments, the decoding model is used to perform joint source channel decoding on the sequence information to obtain the prediction result.

[0175] Similar to manner one, manner two also completes the beam prediction function based on the encoding model and the decoding model. However, the difference lies in that the encoding model performs joint source coding and channel coding on the measurement result of the first resource set, and the decoding model decodes the encoding content of the first resource set and performs prediction of the optimal spatial domain transmission filter.

[0176] The design idea of this mode can be understood from two aspects. For source coding, the purpose is to reduce the redundancy of source coding, i.e., to reduce the overhead. In contrast to mode one, the purpose of channel coding is to increase the redundancy to combat the wireless fading channel and / or noise channel. Joint source-channel coding is a reasonable compromise between the two.

[0177] Compared with mode one, the technical advantage of mode two is that it combines channel coding and uses the statistical characteristics of source data for channel coding. In this application, the statistical characteristics of the source are the statistical characteristics of the first resource set. For example, for time domain measurement, there is a certain time domain correlation between the measurement results of the first resource set. This cannot be achieved by traditional channel coding (without source statistical information). Thus, under the condition of a certain number of coded bits, better error protection can be obtained, and the decoder can more accurately predict the optimal beam information and / or link quality.

[0178] Compared with mode one, mode one uses an encoding model and a decoding model to replace source coding and channel coding, with the purpose of reducing the redundancy of source coding, i.e., reducing the overhead. In contrast, mode two uses an encoding model and a decoding model to replace source coding and channel coding.

[0179] Exemplarily, refer to FIG. 9, which shows the flowchart of the information sending method and the information receiving method provided by another embodiment of the application.

[0180] After the first device performs the relevant measurement, the measurement results of the first resource set are taken as the input of the encoding model, and the encoding model outputs sequence information. Then, after the sequence information is modulated, resource mapped, and antenna mapped, etc., it is sent to the second device. Correspondingly, the second device obtains the sequence information after the received signal is radio frequency received, resource demapped, and demodulated. The second device performs source-channel decoding on the sequence information through the decoding model, and then performs beam prediction on the decoded sequence information to obtain the prediction result.

[0181] In some embodiments, the above method can be applied in different beam prediction scenarios. The following are examples of several possible application scenarios.

[0182] Example 1, beam prediction based on space domain

[0183] Taking the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is the measurement result of the first resource set based on the space domain, and the output is also a sequence of bit numbers, but has been implicitly channel coded and has error protection capability against fading channels and / or noise channels.

[0184] As the input of the decoding model on the NW side, it is the bit stream which has not been decoded by the channel. The decoding model jointly performs the channel decoding and source decoding operations, and predicts the related information of at least one optimal spatial domain transmission filter in the spatial domain.

[0185] In addition, from the perspective of reporting content compression, in addition to the spatial directivity between beams which can be compressed, different wireless channel environments can use channel coding with different degrees of redundancy for protection. For example, when the SNR (Signal-to-Noise Ratio) is high, fewer redundant bits can be used for implicit channel coding; on the contrary, when the SNR is low, more redundant bits can be used for implicit channel coding. Thus, there can be a compromise between source and channel coding that adapts to the environment.

[0186] Example 2, beam prediction based on time domain

[0187] Taking the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is the measurement result of the first resource set based on the time domain, and the output is also a sequence of bit numbers, but the output of the encoding model implicitly includes the channel coding part and the source coding.

[0188] Similarly, the input of the decoding model on the NW side is the bit stream which has not been decoded by the channel. The decoding model jointly performs the channel decoding and source decoding operations, and predicts the related information of at least one optimal spatial domain transmission filter in each time point in the time domain.

[0189] Example 3, beam prediction based on frequency domain

[0190] Similarly, taking the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is the measurement result of the first resource set based on the frequency domain, and the output is also a sequence of bit numbers, but the output of the encoding model implicitly includes the channel coding part and the source coding.

[0191] Similarly, the input of the decoding model on the NW side is the bit stream which has not been decoded by the channel. The decoding model jointly performs the channel decoding and source decoding operations, and predicts the related information of at least one optimal spatial domain transmission filter on each carrier or frequency band in at least one carrier or frequency band in the frequency domain.

[0192] The above-mentioned mode two, by realizing the source coding and decoding and equivalent channel coding (implicit channel coding) in the encoding model and the decoding model, on the basis of the technical effect of mode one, mode two can also provide error protection for the transmission of the measurement result.

[0193] Mode three, joint source channel coding and modulation

[0194] Joint source channel coding and modulation (JSCCM) is a communication technology that integrates source coding, channel coding and modulation.

[0195] In some embodiments, the encoding model is used to jointly source channel encode and modulate the measurement result.

[0196] In some embodiments, the encoding model outputs a sequence of complex numbers after jointly source channel encoding and modulating the measurement result.

[0197] In some embodiments, the above step 720 can be implemented as the following step 723.

[0198] Step 723, the first device sends the sequence information to the second device after resource mapping and antenna mapping.

[0199] Correspondingly, the second device obtains the sequence information after radio frequency receiving and resource demapping the signal from the first device.

[0200] In some embodiments, the decoding model is used to jointly source channel decode and demodulate the sequence information to obtain the prediction result.

[0201] Compared with mode two, the encoding model and the decoding model of mode three perform modulation and demodulation in addition to source channel coding.

[0202] Compared with mode two, the technical advantage of mode three comes from the joint source channel coding and modulation. This way, the bits after source channel coding can be modulated to different constellation points, so as to allocate different power and error resistance to different information bits, thereby improving the signal-to-noise ratio of important information bits. It should be noted that considering the power control of uplink, the constellation points used in mode three need to ensure that the total power is limited in one OFDM (Orthogonal Frequency Division Multiplexing) transmission symbol. In addition, the output of the encoding model is not a sequence of information, which will not bring quantization error. The last advantage is that each constellation point output by the encoding model can contain the input information, so it is more robust than the traditional scheme. The input of the encoding model is the same as the above-mentioned embodiment mode two. But the output of the encoding model is no longer a sequence of information, but a sequence of complex numbers, which is used to map the constellation points on the complex plane. The input of the decoding model is also a complex number, because the decoding model needs to perform demodulation, channel decoding and source decoding.

[0203] Exemplarily, refer to FIG. 10, which shows flowcharts of the sending method and the receiving method of the measurement result provided by one embodiment of the present application.

[0204] After the first device performs the relevant measurement, the measurement result data is taken as the input of the encoding model, and the encoding model outputs sequence information. After the sequence information is subjected to resource mapping, antenna mapping and other processes, it is sent to the second device. Correspondingly, the second device obtains the sequence information after the received signal is subjected to radio frequency reception and resource demapping. The second device obtains the prediction result by performing beam prediction on the decoded sequence information after the sequence information is subjected to demodulation and source channel decoding by the decoding model.

[0205] In some embodiments, the above method can be applied in different beam prediction scenarios. The following are examples of several possible application scenarios.

[0206] Example 1, beam prediction based on spatial domain

[0207] Taking the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is also the measurement result of the first resource set based on the spatial domain, but the output of the encoder is a string of complex numbers, which is used to map the constellation points on the complex plane.

[0208] The input of the decoding model on the NW side is a complex number that has not been demodulated. The decoding model jointly performs joint channel source decoding and demodulation operation, and predicts the related information of at least one optimal spatial domain transmission filter in the spatial domain.

[0209] Example 2, beam prediction based on time domain

[0210] Similarly, taking the first device as a terminal device (such as UE) and the second device as a network device (such as NW) as an example. The input of the encoding model on the UE side is also the measurement result of the first resource set based on the spatial domain, but the output of the encoder is a string of complex numbers, which is used to map the constellation points on the complex plane.

[0211] Similarly, the input of the decoding model on the NW side is a complex number that has not been demodulated. The decoding model jointly performs joint channel source decoding and demodulation operation, and predicts the related information of at least one optimal spatial domain transmission filter in each time point in at least one time point in the time domain.

[0212] Example 3, beam prediction based on frequency domain

[0213] Similarly, taking the first device as a terminal device (such as a UE) and the second device as a network device (such as a NW) as an example. The input of the encoding model on the UE side is also the measurement result of the first resource set based on the space domain, but the output of the encoder is a sequence of complex numbers, which is used to map the constellation points on the complex plane.

[0214] Similarly, the input of the decoding model on the NW side is a complex number that has not been demodulated. The decoding model jointly performs joint channel source decoding and demodulation operations, and predicts the related information of at least one optimal space domain transmission filter on each carrier or frequency band in at least one carrier or frequency band in the frequency domain.

[0215] The above-mentioned mode three, the encoding model and the decoding model realize the modulation and demodulation process on the basis of mode two. On the basis of the technical effects of mode two, mode three can also directly map the output of the decoding model to the constellation points on the complex plane, and balance the allocation of transmit power.

[0216] It should be noted that, in the above method embodiment, the steps performed by the first device can be implemented separately as a signal sending method on the first device side; the steps performed by the second device can be implemented separately as a signal receiving method on the second device side.

[0217] The following is a device embodiment of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the device embodiments of the present application, please refer to the method embodiments of the present application.

[0218] Please refer to FIG. 11, which shows a block diagram of an information sending device according to an embodiment of the present application. The device has the functions of implementing the above-mentioned information sending method examples, which can be implemented by hardware or by hardware executing corresponding software. The device can be the first device introduced above or can be arranged in the first device. As shown in FIG. 11, the device 1100 can include a processing module 1110 and a sending module 1120.

[0219] The processing module 1110 is configured to input the measurement result for the first resource set as the input of the encoding model, and output sequence information from the encoding model, wherein the first resource set includes at least one reference signal resource.

[0220] The sending module 1120 is configured to send the sequence information to the second device, wherein the sequence information is used as the input of the decoding model, and the decoding model outputs a prediction result related to the space domain transmission filter.

[0221] In some embodiments, the prediction result is predicted based on at least one of the following domains: spatial domain, time domain, and frequency domain.

[0222] In some embodiments, the predicted result comprises: information about at least one optimal spatial domain transmission filter; or, information about at least one optimal spatial domain transmission filter at each of at least one time instant; or, information about at least one optimal spatial domain transmission filter on each of at least one carrier or frequency band; or, information about at least one optimal spatial domain transmission filter on each of at least one time instant, each of at least one carrier or frequency band; wherein the at least one carrier or frequency band is different from a carrier or frequency band corresponding to the first resource set.

[0223] In some embodiments, the information comprises at least one of: identification information, direction information, and link quality information.

[0224] In some embodiments, the optimal spatial domain transmission filter is selected from spatial domain transmission filters corresponding to a second resource set, wherein the second resource set comprises at least one reference signal resource.

[0225] In some embodiments, a relationship between the first resource set and the second resource set satisfies at least one of the following conditions: a number of reference signal resources included in the first resource set is less than a number of reference signal resources included in the second resource set; the first resource set is a subset of the second resource set; the reference signal resources included in the first resource set are different from the reference signal resources included in the second resource set; and a carrier or frequency band corresponding to the first resource set is different from a carrier or frequency band corresponding to the second resource set.

[0226] In some embodiments, the optimal spatial domain transmission filter is predicted by the decoding model.

[0227] In some embodiments, the encoding model is configured to source encode the measurement result.

[0228] In some embodiments, the sending module 1120 is configured to send the sequence information to the second device after channel encoding, modulation, resource mapping, and antenna mapping.

[0229] In some embodiments, the encoding model is configured to jointly source and channel encode the measurement result.

[0230] In some embodiments, the sending module 1120 is configured to send the sequence information to the second device after modulation, resource mapping, and antenna mapping.

[0231] In some embodiments, the encoding model is configured to jointly source and channel encode and modulate the measurement result.

[0232] In some embodiments, the sending module 1120 is configured to send the sequence information to the second device after resource mapping and antenna mapping are performed on the sequence information.

[0233] In some embodiments, the sequence information does not have corresponding check bits.

[0234] In some embodiments, the encoding model is an AI or ML model.

[0235] The scheme provided by the embodiments of the present application realizes optimal beam prediction based on a double-end encoding and decoding model. The first device encodes the measurement result of the resource set based on the encoding model to obtain sequence information, and sends the sequence information output by the encoding model to the second device. The second device decodes and predicts the received sequence information based on the decoding model to obtain a prediction result related to a beam. By using the encoding model to encode the measurement result at the sending end, the measurement result to be sent is compressed, and the overhead of reporting the measurement result is reduced. Then, by using the decoding model to decode and predict the received sequence information at the receiving end, a prediction result is obtained, and based on the prediction result, it is helpful to decide a beam suitable for the current communication environment.

[0236] Please refer to FIG. 12, which shows a block diagram of an information receiving device provided by an embodiment of the present application. The device has the functions of implementing the above-mentioned information receiving method examples, which can be implemented by hardware, or by hardware executing corresponding software. The device can be the second device introduced above, or can be arranged in the second device. As shown in FIG. 12, the device 1200 can include a receiving module 1210 and a processing module 1220.

[0237] The receiving module 1210 is configured to receive sequence information sent by the first device, the sequence information being obtained by an encoding model based on a measurement result of a first resource set, the first resource set including at least one reference signal resource.

[0238] The processing module 1220 is configured to take the sequence information as an input of a decoding model, and output a prediction result related to a spatial domain transmission filter by the decoding model.

[0239] In some embodiments, the prediction result is obtained based on at least one of the following domains: a spatial domain, a time domain, and a frequency domain.

[0240] In some embodiments, the prediction result comprises: information about at least one optimal spatial domain transmission filter; or information about at least one optimal spatial domain transmission filter in each of at least one time instant; or information about at least one optimal spatial domain transmission filter on each of at least one carrier or frequency band; or information about at least one optimal spatial domain transmission filter on each of at least one time instant, each of at least one carrier or frequency band; wherein the at least one carrier or frequency band is different from a carrier or frequency band corresponding to the first resource set.

[0241] In some embodiments, the information comprises at least one of: identification information, direction information, and link quality information.

[0242] In some embodiments, the optimal spatial domain transmission filter is selected from spatial domain transmission filters corresponding to a second resource set, wherein the second resource set comprises at least one reference signal resource.

[0243] In some embodiments, a relationship between the first resource set and the second resource set satisfies at least one of the following conditions: a number of reference signal resources included in the first resource set is less than a number of reference signal resources included in the second resource set; the first resource set is a subset of the second resource set; the reference signal resources included in the first resource set are different from the reference signal resources included in the second resource set; and a carrier or frequency band corresponding to the first resource set is different from a carrier or frequency band corresponding to the second resource set.

[0244] In some embodiments, the optimal spatial domain transmission filter is predicted by the decoding model.

[0245] In some embodiments, the encoding model is used for source encoding the measurement result.

[0246] In some embodiments, the receiving module 1210 is configured to obtain the sequence information after performing radio frequency receiving, resource demapping, demodulation, and channel decoding on a signal from the first device.

[0247] In some embodiments, the decoding model is used for joint source channel decoding of the sequence information to obtain the prediction result.

[0248] In some embodiments, the receiving module 1210 is configured to obtain the sequence information after performing radio frequency receiving, resource demapping, and demodulation on a signal from the first device.

[0249] In some embodiments, the decoding model is used for joint source channel decoding and demodulation of the sequence information to obtain the prediction result.

[0250] In some embodiments, the receiving module 1210 is configured to obtain the sequence information after performing radio frequency receiving and resource demapping on the signal from the first device.

[0251] In some embodiments, the sequence information does not have corresponding check bits.

[0252] In some embodiments, the encoding model is an AI or ML model.

[0253] The scheme provided by the embodiments of the present application realizes optimal beam prediction based on a double-end encoding and decoding model. The first device encodes the measurement result of the resource set based on the encoding model to obtain sequence information, and sends the sequence information output by the encoding model to the second device. The second device decodes and predicts the received sequence information based on the decoding model to obtain a prediction result related to the beam. By using the encoding model to encode the measurement result at the sending end, the measurement result to be sent is compressed, and the overhead of reporting the measurement result is reduced. Then, by using the decoding model to decode and predict the received sequence information at the receiving end, a prediction result is obtained, and based on the prediction result, it is helpful to decide the beam suitable for the current communication environment.

[0254] It should be noted that the apparatus provided by the above embodiments is only exemplified by the above division of each functional module when implementing its functions, and in actual application, the above functions can be completed by different functional modules according to actual needs, that is, the content structure of the device is divided into different functional modules to complete all or part of the above described functions.

[0255] As for the apparatus in the above embodiments, the specific manner in which each module performs operations has been described in detail in the embodiments related to the method, and will not be described in detail here.

[0256] Please refer to FIG. 13, which shows a structural schematic diagram of a communication device provided by an embodiment of the present application. The communication device can be the first device or the second device introduced above. The communication device 1300 can include a processor 1301, a transceiver 1302, and a memory 1303. The transceiver 1302 is configured to implement the sending or receiving function, such as the function of the sending module 1120 or the function of the receiving module 1210. The processor 1301 can be configured to implement other processing functions or control the sending and / or receiving, such as the function of the processing module 1110 or the function of the processing module 1220.

[0257] The processor 1301 includes one or more processing cores, and performs various function applications and information processing by running software programs and modules.

[0258] The transceiver 1302 can include a receiver and a transmitter, which can be implemented as the same wireless communication component, and can include a wireless communication chip and a radio frequency antenna.

[0259] The memory 1303 can be connected to the processor 1301 and the transceiver 1302.

[0260] The memory 1303 can be used to store a computer program executed by the processor 1301, and the processor 1301 is configured to execute the computer program to implement various steps in the above method embodiments.

[0261] In some embodiments, when the communication device 1300 is a first device, the processor 1301 is configured to use a measurement result of a first resource set as an input of an encoding model, and output sequence information from the encoding model, the first resource set including at least one reference signal resource. The transceiver 1302 is configured to send the sequence information to a second device, and the sequence information is used as an input of a decoding model to output a prediction result related to a spatial domain transmission filter from the decoding model.

[0262] In some embodiments, when the communication device 1300 is a second device, the transceiver 1302 is configured to receive sequence information sent by a first device, the sequence information being obtained by an encoding model based on a measurement result of a first resource set, the first resource set including at least one reference signal resource. The processor 1301 is configured to use the sequence information as an input of a decoding model to output a prediction result related to a spatial domain transmission filter from the decoding model.

[0263] For details not described in the present embodiment, refer to the above embodiments, which will not be repeated here.

[0264] In addition, the memory can be implemented by any type of volatile or non-volatile storage devices or a combination thereof, including but not limited to: magnetic or optical disks, electrically erasable programmable read-only memories, erasable programmable read-only memories, static random access memories, read-only memories, magnetic memories, flash memories, programmable read-only memories.

[0265] The embodiment of the present application further provides a computer readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to be executed by a processor to implement the information sending method of the first device side or the information receiving method of the second device side. Optionally, the computer readable storage medium can include a ROM (Read-Only Memory), a RAM (Random-Access Memory), a SSD (Solid State Drives) or an optical disc, and the like. The RAM can include a ReRAM (Resistance Random Access Memory) and a DRAM (Dynamic Random Access Memory).

[0266] The embodiment of the present application further provides a chip, wherein the chip includes a programmable logic circuit and / or program instructions, and when the chip is running, the programmable logic circuit and / or the program instructions are used to implement the information sending method of the first device side or the information receiving method of the second device side.

[0267] The embodiment of the present application further provides a computer program product, wherein the computer program product includes a computer program, the computer program is stored in a computer readable storage medium, and a processor reads and executes the computer program from the computer readable storage medium to implement the information sending method of the first device side or the information receiving method of the second device side.

[0268] It should be understood that the "indication" mentioned in the embodiments of the present application can be direct indication, indirect indication or an indication representing an associated relationship. For example, A indicates B, which can mean that B can be obtained by A directly; or A indirectly indicates B, for example, A indicates C, and B can be obtained by C; or A and B have an associated relationship.

[0269] In the description of the embodiments of the present application, the term "corresponding" can mean that there is a direct or indirect corresponding relationship between the two, or can mean that there is an associated relationship between the two, or can mean an indication and being indicated, configuration and being configured, and the like.

[0270] In some embodiments of the present application, "predefined" can be implemented by pre-storing corresponding codes, tables or other information indicating manners in devices (for example, including terminal devices and APs), and the present application does not limit the specific implementation manners. For example, predefined can mean defined in a protocol.

[0271] In some embodiments of the present application, the "protocol" can refer to a standard protocol in the communication field, which can include the LTE protocol, the NR protocol, and related protocols applied in future communication systems, and the present application is not limited thereto.

[0272] "Multiple" mentioned in the present application refers to two or more. "And / or" describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents that the associated objects before and after it are in an "or" relationship.

[0273] "Greater than or equal to" mentioned in the present application can mean greater than or equal to, and "less than or equal to" can mean less than or equal to.

[0274] In addition, the step numbers described in the present application only exemplarily show a possible execution order between steps, and in some other embodiments, the above steps can also be executed in a sequence different from the number, such as two steps with different numbers are executed at the same time, or two steps with different numbers are executed in an order opposite to the illustration, and the embodiments of the present application are not limited thereto.

[0275] Those skilled in the art should realize that in one or more of the above examples, the functions described in the embodiments of the present application can be realized by hardware, software, firmware or any combination thereof. When realized by software, these functions can be stored in a computer readable medium or transmitted as one or more instructions or codes on a computer readable medium. The computer readable medium includes computer storage medium and communication medium, wherein the communication medium includes any medium facilitating the transmission of computer programs from one place to another. The storage medium can be any available medium accessible by a general or special purpose computer.

[0276] The above only describes exemplary embodiments of the present application, and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An information transmission method characterized by comprising: The method is performed by a first device, and the method comprises: outputting, by a coding model, sequence information as an input of the coding model, the sequence information being based on measurement results of a first resource set, the first resource set comprising at least one reference signal resource; sending, to a second device, the sequence information, the sequence information being used as an input of a decoding model, and a prediction result related to a spatial domain transmission filter being output by the decoding model.

2. The method of claim 1, wherein, The prediction result is predicted based on at least one of the following domains: a spatial domain, a time domain, and a frequency domain.

3. The method according to claim 1 or 2, characterized in that, The prediction result comprises: related information of at least one optimal spatial domain transmission filter; or related information of at least one optimal spatial domain transmission filter in each time instance; or related information of at least one optimal spatial domain transmission filter on each carrier or frequency band in at least one carrier or frequency band; or related information of at least one optimal spatial domain transmission filter in each time instance, each carrier or frequency band in at least one time instance and at least one carrier or frequency band; wherein the at least one carrier or frequency band is different from a carrier or frequency band corresponding to the first resource set.

4. The method of claim 3, wherein, The related information comprises at least one of the following: identification information, direction information, and link quality information.

5. The method according to claim 3 or 4, characterized in that, The optimal spatial domain transmission filter is selected from spatial domain transmission filters corresponding to a second resource set, wherein the second resource set comprises at least one reference signal resource.

6. The method of claim 5, wherein, The relationship between the first resource set and the second resource set satisfies at least one of the following conditions: a number of reference signal resources included in the first resource set is less than a number of reference signal resources included in the second resource set; the first resource set is a subset of the second resource set; the reference signal resources included in the first resource set are different from the reference signal resources included in the second resource set; a carrier or frequency band corresponding to the first resource set is different from a carrier or frequency band corresponding to the second resource set.

7. The method according to any one of claims 3 to 6, characterized in that, The optimal spatial domain transmission filter is predicted by the decoding model.

8. The method according to any one of claims 1 to 7, characterized in that, The coding model is used for source coding of the measurement results.

9. The method of claim 8, wherein, The sending of the sequence information to the second device comprises: after channel coding, modulation, resource mapping, and antenna mapping of the sequence information, the sequence information is sent to the second device.

10. The method according to any one of claims 1 to 7, characterized in that, The coding model is used for joint source and channel coding of the measurement results.

11. The method of claim 10, wherein, The sending of the sequence information to the second device comprises: after modulation, resource mapping, and antenna mapping of the sequence information, the sequence information is sent to the second device.

12. The method according to any one of claims 1 to 7, characterized in that, The coding model is used for joint source and channel coding and modulation of the measurement results.

13. The method of claim 12, wherein, The sending of the sequence information to the second device comprises: after resource mapping and antenna mapping of the sequence information, the sequence information is sent to the second device.

14. The method according to any one of claims 1 to 13, characterized in that, The sequence information does not have corresponding check bits.

15. The method according to any one of claims 1 to 14, characterized in that, The coding model is an artificial intelligence (AI) or machine learning (ML) model.

16. An information receiving method, comprising: The method is performed by a second device, and the method comprises: receive sequence information sent by a first device, the sequence information being obtained by an encoding model based on measurement results for a first resource set, the first resource set including at least one reference signal resource; input the sequence information as input of a decoding model, and output prediction results related to a spatial domain transmission filter by the decoding model.

17. The method of claim 16, wherein, The prediction results are predicted based on at least one of the following domains: spatial domain, time domain, and frequency domain.

18. The method of claim 16 or 17, wherein, The prediction results include: related information of at least one optimal spatial domain transmission filter; or related information of at least one optimal spatial domain transmission filter in each time instance; or related information of at least one optimal spatial domain transmission filter on each carrier or frequency band in at least one carrier or frequency band; or related information of at least one optimal spatial domain transmission filter in each time instance, each carrier or frequency band in at least one time instance and at least one carrier or frequency band; The at least one carrier or frequency band is different from a carrier or frequency band corresponding to the first resource set.

19. The method of claim 18, wherein, The related information includes at least one of the following: identification information, direction information, and link quality information.

20. The method of claim 18 or 19, wherein, The optimal spatial domain transmission filter is selected from spatial domain transmission filters corresponding to a second resource set, wherein the second resource set includes at least one reference signal resource.

21. The method of claim 20, wherein, The relationship between the first resource set and the second resource set satisfies at least one of the following conditions: The number of reference signal resources included in the first resource set is less than the number of reference signal resources included in the second resource set; The first resource set is a subset of the second resource set; The reference signal resources included in the first resource set are different from the reference signal resources included in the second resource set; The carrier or frequency band corresponding to the first resource set is different from the carrier or frequency band corresponding to the second resource set.

22. The method according to any one of claims 18 to 21, characterized in that, The optimal spatial domain transmission filter is predicted by the decoding model.

23. The method according to any one of claims 16 to 22, characterized in that, The decoding model is used for source decoding of the sequence information to obtain the prediction results.

24. The method of claim 23, wherein, The receiving of the sequence information sent by the first device includes: The sequence information is obtained after radio frequency reception, resource demapping, demodulation, and channel decoding of a signal from the first device.

25. The method according to any one of claims 16 to 22, characterized in that, The decoding model is used for joint source and channel decoding of the sequence information to obtain the prediction results.

26. The method of claim 25, wherein, The receiving of the sequence information sent by the first device includes: The sequence information is obtained after radio frequency reception, resource demapping, and demodulation of a signal from the first device.

27. The method of any one of claims 16 to 22, wherein, The decoding model is used for joint source and channel decoding and demodulation of the sequence information to obtain the prediction results.

28. The method of claim 27, wherein, The receiving of the sequence information sent by the first device includes: The sequence information is obtained after radio frequency reception and resource demapping of a signal from the first device.

29. The method according to any one of claims 16 to 28, characterized in that, The sequence information does not have corresponding check bits.

30. The method of any one of claims 16 to 29, wherein, The decoding model is an artificial intelligence (AI) or machine learning (ML) model.

31. A transmission apparatus of a measurement result, characterized by, The apparatus includes: The processing module is configured to use the measurement result for the first resource set as an input of an encoding model, and output sequence information from the encoding model. The sending module is configured to send the sequence information to a second device, and use the sequence information as an input of a decoding model, and output a prediction result related to the spatial domain transmission filter from the decoding model.

32. A receiving device of a measurement result, characterized by, The apparatus comprises: The receiving module is configured to receive sequence information sent by a first device, and the sequence information is obtained by an encoding model based on a measurement result for a first resource set, and the first resource set comprises at least one reference signal resource. The processing module is configured to use the sequence information as an input of a decoding model, and output a prediction result related to the spatial domain transmission filter from the decoding model.

33. A communications device, characterized by The communication device comprises a processor and a memory, and the memory stores a computer program, and the processor executes the computer program to implement the method in any one of claims 1 to 15, or implement the method in any one of claims 16 to 30.

34. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and the computer program is used to be executed by a processor to implement the method in any one of claims 1 to 15, or implement the method in any one of claims 16 to 30.

35. A chip, comprising: The chip comprises a programmable logic circuit and / or program instructions, and when the chip is running, the programmable logic circuit and / or program instructions are used to implement the method in any one of claims 1 to 15, or implement the method in any one of claims 16 to 30.

36. A computer program product, characterised in that, The computer program product comprises computer instructions stored in a computer readable storage medium, and a processor reads and executes the computer instructions from the computer readable storage medium to implement the method in any one of claims 1 to 15, or implement the method in any one of claims 16 to 30.

Images