Reference signal compensation for training neural networks

Abstract

Example embodiments of the present disclosure relate to compensating reference signals for training neural networks. According to embodiments of the present disclosure, a solution for compensating reference signals for training neural networks is proposed. A second device transmits information to a first device. The information is used to trigger the generation and transmission of a compensated training reference signal. The information indicates a set of receive ports at the first device. The first device compensates the training reference signal based on downlink channel information and transmits the compensated signal to the second device via one or more transmit ports. In this way, CSI can be acquired more accurately. In addition, the processing model at the second device can be improved.

Description

Technical Field

[0001] Embodiments of the present disclosure relate generally to the field of telecommunications, and more particularly to methods, devices, apparatus, and computer-readable storage media for compensating a reference signal for training a neural network. Background Art

[0002] In wireless communications, channel information, such as channel state information (CSI), refers to the known channel properties of a communication link. This information describes how the signal propagates from the transmitter to the receiver and represents the combined effects of, for example, scattering, fading, and power loss over distance. CSI enables adaptation of transmission to current channel conditions, which is crucial for achieving reliable high-data-rate communication in multi-antenna systems. Several solutions have been proposed for acquiring CSI. Therefore, ensuring CSI accuracy is crucial. Summary of the Invention

[0003] In general, example embodiments of the present disclosure provide a solution for compensating a reference signal for training a neural network.

[0004] In a first aspect, a first device is provided. The first device includes at least one processor; and at least one memory including computer program code; the at least one memory and the computer program code are configured to, with the at least one processor, cause the first device to receive information indicating a set of receive ports to be used for compensation at the first device. The first device is further caused to determine, based on downlink channel information, at least one compensation parameter that associates the set of receive ports with a set of transmit ports. The first device is further caused to compensate at least one training reference signal using the at least one compensation parameter. The first device is further caused to transmit the at least one compensated training reference signal to a second device via the set of transmit ports of the first device.

[0005] In a second aspect, a second device is provided. The second device includes at least one processor; and at least one memory including computer program code; the at least one memory and the computer program code are configured to, with the at least one processor, cause the second device to transmit information indicating a set of receive ports to be used for compensation at the first device to a first device. The second device is further caused to receive, from the first device, at least one compensated training reference signal transmitted via a set of transmit ports of the first device, the at least one compensated training reference signal being compensated using at least one compensation parameter, the at least one compensation parameter associating the set of receive ports with a set of transmit ports determined based on downlink channel information.

[0006] In a third aspect, a method is provided. The method includes receiving information indicating a set of receive ports to be used for compensation at a first device. The method also includes determining, based on downlink channel information, at least one compensation parameter that associates the set of receive ports with a set of transmit ports. The method also includes compensating at least one training reference signal using the at least one compensation parameter. The method also includes transmitting the at least one compensated training reference signal to a second device via the set of transmit ports of the first device.

[0007] In a fourth aspect, a method is provided. The method includes transmitting, at a second device, to a first device, information indicating a set of receive ports to be used for compensation at the first device. The method also includes receiving, from the first device, at least one compensated training reference signal transmitted via a set of transmit ports of the first device, the at least one compensated training reference signal compensated with at least one compensation parameter, the at least one compensation parameter associating the set of receive ports with the set of transmit ports determined based on downlink channel information.

[0008] In a fifth aspect, an apparatus is provided. The apparatus includes means for receiving information indicating a set of receive ports to be used for compensation at a first device; means for determining, based on downlink channel information, at least one compensation parameter that associates the set of receive ports with a set of transmit ports; means for compensating at least one training reference signal with the at least one compensation parameter; and means for transmitting the at least one compensated training reference signal to a second device via the set of transmit ports of the first device.

[0009] In a sixth aspect, an apparatus is provided. The apparatus includes means for transmitting, at a second device, to a first device, information indicating a set of receive ports to be used for compensation at the first device; and means for receiving, from the first device, at least one compensated training reference signal transmitted via a set of transmit ports of the first device, the at least one compensated training reference signal compensated with at least one compensation parameter, the at least one compensation parameter associating the set of receive ports with the set of transmit ports determined based on downlink channel information.

[0010] In a seventh aspect, a computer-readable medium is provided, wherein the computer-readable medium includes program instructions for causing a device to at least execute the method according to any one of the third and fourth aspects.

[0011] It should be understood that the invention summary is not intended to determine the key or essential features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Some example embodiments will now be described with reference to the accompanying drawings, in which:

[0013] Figure 1 shows a simplified block diagram of a training model according to some example embodiments of the present disclosure;

[0014] Figure 2 shows a simplified block diagram of a training model according to other example embodiments of the present disclosure;

[0015] Figure 3 An example communication environment is shown in which example embodiments of the present disclosure may be implemented;

[0016] Figure 4 shows a simplified block diagram of a compensation reference signal according to some example embodiments of the present disclosure;

[0017] Figure 5 shows a signaling flow for applying uplink channel information to determine a data processing model to be deployed for the downlink according to some example embodiments of the present disclosure;

[0018] Figure 6 A flowchart illustrating a method implemented at a first device according to some example embodiments of the present disclosure is shown;

[0019] Figure 7 A flowchart illustrating a method implemented at a second device according to some other example embodiments of the present disclosure is shown;

[0020] Figure 8 shows a simplified block diagram of an apparatus suitable for implementing an example embodiment of the present disclosure; and

[0021] Figure 9 A block diagram of an example computer-readable medium is shown, according to some example embodiments of the present disclosure.

[0022] Throughout the drawings, the same or similar reference numerals refer to the same or similar elements. DETAILED DESCRIPTION

[0023] The principles of the present disclosure will now be described with reference to some exemplary embodiments. It should be understood that these embodiments are described only to illustrate and help those skilled in the art understand and implement the present disclosure, and do not represent any limitation on the scope of the present disclosure. The embodiments described herein can be implemented in various other ways except as described below.

[0024] In the following description and claims, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

[0025] In this disclosure, references to "one embodiment," "an embodiment," and "an example embodiment" indicate that the described embodiment may include a particular feature, structure, or characteristic, but not every embodiment must include the particular feature, structure, or characteristic. Furthermore, such phrases do not necessarily refer to the same embodiment. Furthermore, when a particular feature, structure, or characteristic is described in conjunction with one embodiment, those skilled in the art believe that it is within the knowledge of those skilled in the art to affect such feature, structure, or characteristic in conjunction with other embodiments, whether or not explicitly described.

[0026] It should be understood that although the terms "first" and "second" and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element without departing from the scope of the exemplary embodiments. As used herein, the term "and / or" includes any and all combinations of one or more of the listed terms.

[0027] The terms used herein are for describing particular embodiments only and are not intended to limit the example embodiments. As used herein, the singular forms "a," "an," and "the" also include the plural forms, unless the context clearly indicates otherwise. It is further understood that the terms "comprises," "comprising," "has," "having," "includes," and / or "including" when used herein specify the presence of stated features, elements, and / or components, etc., but do not preclude the presence or addition of one or more other features, elements, components, and / or combinations thereof.

[0028] As used in this application, the term "circuitry" may refer to one or more or all of the following:

[0029] (a) a pure hardware circuit implementation (such as an implementation using only analog and / or digital circuitry), and

[0030] (b) a combination of hardware circuitry and software such as (as applicable):

[0031] (i) a combination of analog and / or digital hardware circuits and software / firmware, and

[0032] (ii) any portion of hardware processor(s) (including digital signal processor(s)) with software, software and memory(s) that work together to cause a device (such as a mobile phone or server) to perform various functions, and

[0033] (c) Hardware circuit(s) and / or processor(s), such as microprocessor(s) or portion(s) of microprocessor(s), which requires software (e.g., firmware)

[0034] The software can be operated, but it can be not saved when no operation is needed.

[0035] This definition of circuitry applies to all uses of the term in this application, including in any claims. As another example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or a portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. For example, if applicable to the particular claim element, the term circuitry also covers a baseband integrated circuit or processor integrated circuit for a mobile device, or a similar integrated circuit in a server, cellular network device, or other computing or network device.

[0036] As used herein, the term "communication network" refers to a network that complies with any suitable communication standard, such as New Radio (NR), Long Term Evolution (LTE), Advanced LTE (LTE-A), Wideband Code Division Multiple Access (WCDMA), High Speed ​​Packet Access (HSPA), Narrowband Internet of Things (NB-IoT), etc. In addition, the communication between the terminal device and the network device in the communication network can be performed according to any suitable generation of communication protocol, including but not limited to first generation (1G), second generation (2G), 2.5G, 2.75G, third generation (3G), fourth generation (4G), 4.5G, fifth generation (5G) communication protocols, future sixth generation (6G), and / or any other protocol currently known or to be developed in the future. The embodiments of the present disclosure can be applied to various communication systems. In view of the rapid development of communications, there will certainly be future types of communication technologies and systems that can embody the present disclosure. The scope of the present disclosure should not be limited to the above-mentioned systems.

[0037] As used herein, the term "network device" refers to a node in a communication network through which a terminal device accesses the network and receives services from the network. A network device may refer to a base station (BS) or an access point (AP), for example, a Node B (NodeB or NB), an evolved NodeB (eNodeB or eNB), a NR NB (also known as a gNB), a remote radio unit (RRU), a radio head (RH), a remote radio head (RRH), a relay, an integrated and access backhaul (IAB) node, a low power node (such as a femto, a pico), a non-terrestrial network (NTN) or a non-terrestrial network device (such as a satellite network device, a low earth orbit (LEO) satellite, and a geosynchronous orbit (GEO) satellite), an aircraft network device, etc., depending on the terminology and technology applied.

[0038] The term "terminal device" refers to any terminal device capable of wireless communication. By way of example and not limitation, a terminal device may also be referred to as a communication device, user equipment (UE), subscriber station (SS), portable subscriber station, mobile station (MS), or access terminal (AT). Terminal devices may include, but are not limited to, mobile phones, cellular phones, smartphones, voice over IP (VoIP) phones, wireless local loop phones, tablet computers, wearable terminal devices, personal digital assistants (PDAs), portable computers, desktop computers, image capture terminal devices (such as digital cameras), gaming terminal devices, music storage and playback devices, in-vehicle wireless terminal devices, wireless endpoints, mobile stations, laptop embedded devices (LEEs), laptop mounted devices (LMEs), USB dongles, smart devices, wireless customer premises equipment (CPE), Internet of Things (IoT) devices, watches or other wearable devices, head-mounted displays (HMDs), vehicles, drones, medical devices and applications (e.g., remote surgery), industrial devices and applications (e.g., robots and / or other wireless devices operating in the context of industrial and / or automated process chains), consumer electronic devices, devices operating on commercial and / or industrial wireless networks, etc. In the following description, the terms "terminal device", "communication device", "terminal", "user equipment" and "UE" may be used interchangeably.

[0039] As mentioned above, how to obtain CSI is crucial for communication performance. Multiple-input, multiple-output (MIMO) based on machine learning (ML) has attracted significant attention and demonstrated its advantages over physical layer solutions such as beamforming and channel state information (CSI) acquisition. ML-based massive MIMO solutions can provide performance enhancements and reduce computational complexity, overhead, and latency.

[0040] For analytical purposes, many ML schemes for CSI acquisition (e.g., including channel estimation, feedback, etc.) and beamforming can use statistical channel models, ray tracing generated data, or measurement data to train neural networks (NNs) offline. Since the datasets used in pre-training cannot fully represent the real environment, when such neural networks are deployed in practice, additional training is required to update the weights of the pre-trained NNs. In some other cases, the NNs can also be trained directly in the real scenarios of deployment. For example, in CSI feedback and channel estimation applications, the training of NNs requires paired datasets (e.g., real or "reference" CSI and compressed CSI), however, it is difficult to obtain real CSI in practice. Some traditional techniques only assume that accurate CSI can be obtained to train or update the NNs in deployment.

[0041] A conventional solution is to use high-resolution CSI feedback to train NNs in real-world deployments. However, data collection incurs significant overhead. Even with high resolution, it is still far from accurate because the UE needs to quantize the estimated CSI. This high-resolution CSI quantization, which is then considered the reference CSI, also degrades NN performance.

[0042] In time division duplex (TDD) systems, uplink CSI can be used directly to train the NN for the downlink due to reciprocity. However, in FDD systems, when reciprocity does not hold, this concept cannot be applied. A solution for channel information reporting has been proposed that uses uplink channel information to determine the data processing model deployed for downlink use. The data processing model is determined using an uplink data set (e.g., an uplink reference signal), and then the data processing model is used to process the downlink CSI.

[0043] Figure 1 A simplified block diagram of an apparatus 100 for training a model according to some example embodiments of the present disclosure is shown. Apparatus 100 may include module 110, which may be configured to compress first CSI to obtain second CSI. The second CSI may contain less information than the first CSI. The apparatus may also include module 120, which may be configured to train a data processing model based on the first CSI and the second CSI. In this manner, by using the data processing model, downlink CSI may be more accurately obtained.

[0044] One issue during NN implementation is that the NN may not always function during its operational phase. Therefore, a monitoring mechanism is needed to detect NN errors or failures. This paper proposes a solution for monitoring channel information of a data processing model. Two parallel measurements performed by the data processing model on both the uplink and downlink channels are used to determine whether the data processing model is functioning properly. Figure 2 A simplified block diagram of an apparatus 200 for training a model according to some example embodiments of the present disclosure is shown.

[0045] Apparatus 200 may include module 210, which may be configured to recover uplink and downlink information using a data processing model. For example, module 210 may recover downlink channel information from processed downlink information and recover uplink channel information from processed uplink information. The recovered downlink channel information may be compared with the processed downlink channel information, and the recovered uplink channel information may be compared with the processed uplink channel information. The data processing model may be updated based on the comparison.

[0046] A key issue that needs to be specified and addressed is how to coordinate network equipment and terminal devices to provide uplink data sets when the terminal device has an asymmetric transceiver, that is, how to generate training reference signals in frequency division duplex (FDD). Traditionally, compensation is performed on the network device side, such as based on reciprocity issues in FDD, and the transceivers are symmetrical. For example, for the downlink transmission processing design at the network device, the uplink channel covariance is compensated, where this compensation should be for instantaneous use. This algorithm is implementation-dependent, and traditional techniques mainly consider it as an algorithm and non-overlapping / non-correlated aspects.

[0047] According to an embodiment of the present disclosure, a solution for compensating reference signals for training neural networks is proposed. A second device transmits information to a first device. This information is used to trigger the generation and transmission of a compensated reference signal. The information indicates a set of receive ports at the first device. The first device compensates the reference signal based on downlink channel information and transmits the compensated signal to the second device via one or more transmit ports. In this way, CSI can be acquired more accurately. In addition, the processing model at the second device can be improved.

[0048] Figure 3 A schematic diagram of a communication environment 300 in which embodiments of the present disclosure may be implemented is shown. The communication environment 100, as part of a communication network, includes devices 310-1, 310-2, ..., 310-N, which may be collectively referred to as "first device(s) 310." The communication environment 300 includes a second device 320. The parameter N may be any suitable number. The first device 310 and the second device 320 may communicate with each other.

[0049] Communication environment 300 may include any suitable number of devices and cells. In communication environment 300, first device 310 and second device 320 may transmit data and control information to each other. When first device 310 is a network device and second device 320 is a terminal device, the link from second device 320 to first device 310 is called an uplink (UL), and the link from first device 310 to second device 320 is called a downlink (DL). Second device 320 and first device 310 are interchangeable.

[0050] It should be understood that Figure 3 The number of first devices and cells and their connections shown is provided for illustration purposes only and is not intended to be limiting. Communication environment 300 may include any suitable number of devices and networks suitable for implementing embodiments of the present disclosure.

[0051] Communications in the communication environment 300 may be implemented according to any suitable communication protocol(s), including but not limited to first generation (1G), second generation (2G), third generation (3G), fourth generation (4G), fifth generation (5G), and sixth generation (6G) cellular communication protocols, wireless local area network communication protocols such as Institute of Electrical and Electronics Engineers (IEEE) 802.11, and / or any other protocol currently known or developed in the future. Furthermore, communications may utilize any suitable wireless communication technology, including but not limited to code division multiple access (CDMA), frequency division multiple access (FDMA), time division multiple access (TDMA), frequency division duplex (FDD), time division duplex (TDD), multiple input multiple output (MIMO), orthogonal frequency division multiplexing (OFDM), discrete Fourier transform spread OFDM (DFT-s-OFDM), and / or any other technology currently known or developed in the future.

[0052] Figure 4 4 is a simplified block diagram of an apparatus 400 for compensating a training reference signal according to some example embodiments of the present disclosure. The apparatus 400 may be implemented at the first device 310. Alternatively, the apparatus 400 may be implemented at the second device 320. It should be noted that the embodiments of the present disclosure are not limited thereto.

[0053] The apparatus 400 may include receiving ports 410-1, 410-2, 410-3, ..., 410-M, which may be collectively referred to as "(plurality) receiving ports 410". The apparatus 400 may also include transmitting ports 420-1, ..., 420-P, which may be collectively referred to as "(plurality) transmitting ports 420". It should be understood that Figure 4 The number of receiving ports and the number of transmitting ports shown are for illustration purposes only and are not limiting. Apparatus 400 may include any suitable number of receiving ports and transmitting ports suitable for implementing embodiments of the present disclosure.

[0054] Apparatus 400 may include a compensation module 230, which may be configured to compensate for a training reference signal. Downlink channel information may be extracted from receiving port 410. Compensation module 230 may compensate for the training reference signal based on the downlink channel information. The training reference signal may be transmitted via transmission port 420-1. The training reference signal may be any suitable reference signal. Embodiments of the present disclosure are not limited thereto.

[0055] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Figure 5 , Figure 5 A signaling flow 500 for training a downlink data processing model using uplink channel information according to an example embodiment of the present disclosure is shown. Figure 3A signaling flow 500 is described. The signaling flow 500 may involve a first device 310-1 and a second device 320.

[0056] The second device 320 transmits information to the first device 310-1. This information may indicate a set of receive ports at the first device 310-1. This information may be used to trigger the first device 310-1 to compensate for one or more training reference signals and transmit the one or more compensated training reference signals. In some embodiments, this information may include an index of the receive ports at the first device 310-1. Alternatively, the second device may understand how the receive ports at the first device 310-1 are divided into different groups. In this case, the information may indicate the groups into which the receive ports at the first device 310-1 are divided. In other embodiments, the information may indicate the number of receive ports at the first device to be used for compensation. Alternatively, the information may indicate a mapping between the set of receive ports and a set of transmit ports at the first device to be used for compensation. The number of ports in the set of transmit ports may be any suitable number. For example, the set of transmit ports may include one or more transmit ports.

[0057] For illustration purposes only, assume that there are four receive ports 410-1, 410-2, 410-3, and 410-4 (not shown) and two transmit ports 420-1 and 420-2 (not shown) at the first device 310-1. The second device 320 may determine two groups of receive ports. For example, receive ports 410-1 and 410-2 may correspond to transmit ports 420-1 and 420-2, and receive ports 410-3 and 410-4 may correspond to transmit ports 420-1 and 420-2. As an example only, the information may indicate receive ports 410-1 and 410-2. Alternatively, the information may also indicate that receive ports 410-1 and 410-2 may correspond to transmit ports 420-1 and 420-2, and receive ports 410-3 and 410-4 may correspond to transmit ports 420-1 and 420-2. It should be noted that the information may indicate any one or more suitable receiving ports at the first device 310 - 1 .

[0058] In some embodiments, as Figure 5 As shown, the second device 320 may broadcast 5005 the information to the first device 310-1. For example, the broadcast information may include triggering the UE's behavior, i.e., compensation of the training reference signal and transmission of the compensated training reference signal. In some embodiments, the information may be sent during the initial access process.

[0059] If the first device 310-1 agrees to send the compensated training reference signal, the first device 310-1 may transmit 5010 an acknowledgement of the information to the second device 320. If the first device 310-1 refuses to send the compensated training reference signal, the first device 310-1 may not transmit any response to the second device 320.

[0060] If an acknowledgment is received from the first device 310-1, the second device 320 may allocate 5015 resource(s) for transmitting the compensated training reference signal. The resource(s) may be any suitable resource for transmitting a signal by the first device 310-1. The second device 320 may transmit 5020 resource information indicating the allocated resources to the first device 310-1.

[0061] Alternatively, the second device 320 may transmit 5025 the information to the first device 310-1 via dedicated signaling. The information may be specific to the first device 310-1. In the event that the second device 320 needs to update the neural network during its operation, for example, in the event that the second device 320 has not previously encountered a specific UE situation, the second device 320 may need to collect such data to train and enhance the neural network. The UE-specific information may indicate compensation of the training reference signal and transmission of the compensated training reference signal from the terminal device 310-1.

[0062] In some embodiments, the first device 310-1 may determine 5030 a group of receive ports to be used for compensation based on the information. For example, if the information includes explicit indices of receive ports, the first device 310-1 may determine the group of receive ports based on the explicit indices. Alternatively, if the information indicates partitioned groups, the first device 310-1 may determine the group of receive ports based on the partitioned groups. For example, if the information indicates a first group, the first device 310-1 may determine receive ports 410-1 and 410-2 that belong to the first group.

[0063] The first device 310-1 determines 5035 one or more compensation parameters for associating the set of receive ports with the set of transmit ports based on the downlink channel information. The downlink channel information may include downlink channel covariance. For example, the downlink channel information may include the downlink channel covariance of all receive ports. In some embodiments, the first device 310-1 may extract angle information from the downlink channel covariance. In some embodiments, the first device 310-1 may determine a predefined direction of arrival (DOA) based on the estimated downlink channel information at all receive ports. The downlink channel information may have been previously acquired by the first device 310-1. For example, the first device 310-1 may receive a channel state information (CSI) reference signal from the second device 320 via each receive port 410 at the first device 310-1. The downlink channel information for each receive port 410 may be estimated based on the CSI reference signal received from the second device 320. In some embodiments, the predefined DOA may be a maximum DOA. It should be noted that the predefined DOA may be any suitable DOA. Note that the compensation parameters may be determined using any suitable method, not limited to the above-described embodiments.

[0064] In some embodiments, when an uplink frequency is applied, the first device 310-1 may obtain phase information of a reference signal received at the group of receive ports. When a downlink frequency is applied, the first device 310-1 may obtain phase information of a reference signal received at the group of receive ports. The first device 310-1 may obtain a phase difference at the group of receive ports caused by the difference between the uplink frequency and the downlink frequency. For example, the first device 310-1 may determine a first array response of the group of receive ports 1 at a downlink frequency. A second array response of the group of transmit ports-1 at an uplink frequency may be determined. The compensation parameters may be determined based on the first array response, the second array response, and a predetermined DOA. By way of example only, consider the array response vector a(θ) of a uniform linear antenna ULA having N elements / ports when a single plane wave in the θ direction relative to the array aperture sight strikes the array. This vector may be written as:

[0065]

[0066] where f c denotes the carrier frequency, c denotes the speed of light, and d denotes the distance between two adjacent antenna ports / elements (assumed to be the same for both uplink and downlink).

[0067] For illustration purposes only, assume that there are four receiving ports 410-1, 410-2, 410-3, and 410-4 (not shown) and two transmitting ports 420-1 and 420-2 (not shown) at the first device 310-1. The information may indicate the receiving ports 410-1 and 410-2. In this case, the compensation parameter may be obtained by the following formula:

[0068]

[0069] Where T represents the compensation matrix, UE represents the first device 310-1, "1" represents the first group of receiving ports (ie, receiving ports 410-1 and 410-2), θ max Indicates the maximum DOA, a D (θ max ) represents the array response of the first device 310-1 in the downlink, a U (θ max ) represents the array response of the first device 310-1 in the uplink. [X] n，n represents the nth row and nth column of matrix X, [x] n represents the nth element of the vector x, and n represents a receiving port in the set of receiving ports. In some embodiments, the compensation matrix may be designed to be diagonal to facilitate performing port-specific compensation.

[0070] The first device 310-1 performs 5040 compensation on one or more training reference signals. In some embodiments, the one or more training reference signals may be phase compensated. Alternatively, the one or more training reference signals may be compensated in other aspects (e.g., frequency domain). It should be noted that the one or more training reference signals may be compensated based on a symbol difference between a receiving port and a transmitting port.

[0071] The first device 310-1 transmits 5045 the compensated one or more training reference signals to the second device 320 via the set of transmission ports. For example, the compensated training reference signal can be represented as:

[0072] T UE，1 (θ max )s1, (3)

[0073] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, "1" represents the first group of receiving ports (i.e., receiving ports 410-1 and 410-2), and s1 represents the training reference signal. The compensated training reference signal received at the second device 320 can be represented as:

[0074] Y1=H U，1 T UE，1 (θmax )s1+N1, (4)

[0075] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, "1" represents the first group of receiving ports (ie, receiving ports 410-1 and 410-2), s1 represents a training reference signal, N1 represents the noise on the channel, and H U，1 represents the uplink channel response when the receiving ports 410 - 1 and 410 - 2 are applied, and Y1 represents the received compensated training reference signal.

[0076] In some embodiments, the second device 320 may obtain the compensated uplink channel response by the following formula:

[0077]

[0078] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, "1" represents the first group of receiving ports (ie, receiving ports 410-1 and 410-2), H U，1 represents the uplink channel response when the receiving ports 410-1 and 410-2 are applied, Indicates that the uplink channel response has been compensated.

[0079] In some embodiments, the information may also indicate another set of receiving ports. Alternatively, the second device 320 may transmit 5050 additional information. The additional information may indicate the other set of receiving ports. In this way, the characteristics of the receiving ports can be reflected by the transmitting ports, allowing for more accurate CSI acquisition.

[0080] For illustration purposes only, assume that there are four receive ports 410-1, 410-2, 410-3, and 410-4 (not shown) and two transmit ports 420-1 and 420-2 (not shown) at the first device 310-1. The additional information may indicate receive ports 410-3 and 410-4. An additional compensation parameter associating the additional set of receive ports with the set of transmit ports may be determined based on the downlink channel information. The additional compensation parameter may be determined using a method similar to that used to determine the compensation parameter.

[0081] In some embodiments, when the uplink frequency is applied, the first device 310-1 may obtain phase information of the reference signal received at the other group of receiving ports. When the downlink frequency is applied, the first device 310-1 may obtain phase information of the reference signal received at the other group of receiving ports. The first device 310-1 may obtain the phase difference at the other group of receiving ports caused by the difference between the uplink frequency and the downlink frequency. In some embodiments, the first device 310-1 may determine another predefined direction of arrival (DOA) based on the estimated downlink channel information at all receiving ports. Note also that the maximum DOA used for compensation during different transmission periods may also be calculated differently to ensure accurate compensation or transformation. In this case, additional compensation parameters may be obtained by the following formula:

[0082]

[0083] Where T represents the compensation matrix, UE represents the first device 310-1, "2" represents the second group of receiving ports (ie, receiving ports 410-3 and 410-4), θ max Indicates the maximum DOA, a D (θ max ) represents the array response of the first device 310-1 in the downlink, a U (θ max ) represents the array response of the first device 310-1 in the uplink. [X] n，n represents the nth row and nth column of matrix X, [x] n represents the nth element of the vector x, and n represents a receiving port in the set of receiving ports. In some embodiments, the compensation matrix may be designed to be diagonal to facilitate performing port-specific compensation.

[0084] Compensation can be performed to directly compensate the training reference signal. That is, the first device 310-1 can convert the array response associated with the transmit port in the uplink into the array response associated with the receive port in the downlink and implement it on the training reference signal. The second device 320 can see that the compensated channel will be the pre-compensated uplink channel associated with the transmit port, which represents the statistics of the downlink channel associated with the receive port. From the perspective of the training reference, it can compensate the training reference signal. From the perspective of the channel, it effectively compensates the uplink channel to represent the downlink channel.

[0085] In some embodiments, the first device 310-1 may transmit 5055 one or more additional compensated training reference signals to the second device 320. For example, the additional compensated training reference signals may be represented as:

[0086] T UE，2 (θmax )s2, (7)

[0087] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, "2" represents the second set of receive ports (i.e., receive ports 410-3 and 410-4), and s2 represents an additional training reference signal. The additional compensated training reference signal received at the second device 320 can be represented as:

[0088] Y2=H U，2 T UE，2 (θ max )s2+N2, (8)

[0089] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, "2" represents the second group of receiving ports (ie, receiving ports 410-3 and 410-4), s2 represents an additional training reference signal, N2 represents the noise on the channel, H U，2 represents an additional response of the uplink channel when the receiving ports 410 - 4 and 410 - 3 are applied, and Y2 represents an additional received compensated training reference signal.

[0090] In some embodiments, the second device 320 may obtain another compensated uplink channel response by the following formula:

[0091]

[0092] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, "2" represents the second group of receiving ports (ie, receiving ports 410-3 and 410-4), H U，2 represents another response of the uplink channel when the receiving ports 410-3 and 410-4 are applied, Indicates the additional response of the uplink channel has been compensated.

[0093] In some embodiments, two operations of the UE's transmission of the transmission port can be identified. In one case, the UE has antenna switching capability. Two rounds of compensated training reference signals can be transmitted via transmission ports 420-1 and 420-2 switched to receive ports (410-1, 410-2) and (410-3, 410-4), respectively. In another case, when the first device 310-1 does not have antenna switching capability, the transmission port cannot be switched to the receive port, but a compensation operation related to two groups of receive ports is required. Two rounds of compensated training reference signals can be transmitted only via transmission ports 420-1 and 420-2.

[0094] In some embodiments, for example, in NR, the first device 310-1 may have a beamformed training reference signal to ensure good transmission power. In this case, the first device 310-1 may use the beamforming matrix W R The compensated training reference signal is beamformed. The compensated training reference signal received by the second device 320 can be written as:

[0095] Y=H U W R T UE (θ max )s+N, (10)

[0096] Where T represents the compensation matrix, θ max represents the maximum DOA, UE represents the first device 310-1, s represents the training reference signal, N represents the noise on the channel, H U represents the response of the uplink channel, Y represents the received compensated training reference signal, and W R represents the beamforming matrix.

[0097] In some embodiments, the second device 320 may determine 5060 an estimate of a compensated uplink channel between the first device 310-1 and the second device 320 based on the received training reference signals. The second device 320 may collect all sets of compensated reference signals from all receive ports. The complete compensated uplink channel used as input data for the neural network may be constructed as:

[0098]

[0099] in represents the total response of the compensated uplink channel, represents the compensated uplink channel response when receiving ports 410-1 and 410-2 are applied, This indicates that the additional response of the uplink channel has been compensated when the receiving ports 410 - 3 and 410 - 4 are applied.

[0100] In some embodiments, the second device 320 may update 5065 a data processing model based on the estimation, the data processing model to be used to recover channel information on the downlink channel between the first device 310-1 and the second device 320. For example, the complete compensated uplink channel may be used as input data for a neural network to update the data processing model.

[0101] According to the embodiments of the present disclosure, the data processing model can be trained more accurately. In addition, more complete channel information can be obtained, and CSI can be estimated more accurately. The training reference signal is not always required as an additional reference signal in the system. In the early stages of deploying a neural network, since many situations are unknown and the neural network needs to be fine-tuned, a training reference signal is more often required. After sufficient training data is collected, the neural network is more stable and only needs to be updated when necessary. Once the neural network reaches relative robustness, little training is required in the current scenario.

[0102] Figure 6 A flow chart of an example method 600 implemented at a first device 310 according to some example embodiments of the present disclosure is shown. For purposes of discussion, the method 600 will be described from the perspective of the first device 310.

[0103] In block 610, the first device 310-1 receives information from the second device 320. The information may indicate a set of receive ports on the first device 310-1. This information may be used to trigger the first device 310-1 to compensate for at least one training reference signal and transmit at least one compensated training reference signal. In some embodiments, the information may include an index of the receive ports on the first device 310-1. Alternatively, the second device may understand how the receive ports on the first device 310-1 are divided into different groups. In this case, the information may indicate the grouping of the receive ports on the first device 310-1. In other embodiments, the information may indicate the number of receive ports. Alternatively, the information may indicate a mapping between the set of receive ports and a set of transmit ports on the first device. The number of ports in the set of transmit ports may be any suitable number. For example, the set of transmit ports may include one or more transmit ports.

[0104] In some embodiments, the information may be broadcast to the first device 310-1. For example, the broadcast information may include triggering the UE's behavior, i.e., compensation of the training reference signal and transmission of the compensated training reference signal. In some embodiments, the information may be sent during the initial access process.

[0105] If the first device 310-1 agrees to transmit the compensated training reference signal, the first device 310-1 may transmit an acknowledgement of the information to the second device 320. If the first device 310-1 refuses to transmit the compensated training reference signal, the first device 310-1 may not transmit any response to the second device 320. In some embodiments, the first device 310-1 may receive resource information from the second device 320, the resource information indicating allocated resources to be used for transmission of the compensated training reference signal.

[0106] Alternatively, the information may be transmitted to the first device 310-1 via dedicated signaling.The information may be specific to the first device 310-1.The UE-specific information may indicate compensation of training reference signals and transmission of compensated training reference signals from the terminal device 310-1.

[0107] In some embodiments, the first device 310-1 may determine the group of receiving ports based on the information. For example, if the information includes an explicit index of the receiving port, the first device 310-1 may determine the group of receiving ports based on the explicit index. Alternatively, if the information indicates a partition group, the first device 310-1 may determine the group of receiving ports based on the partition group.

[0108] In block 620, the first device 310-1 determines at least one compensation parameter that associates the group of receive ports with a group of transmit ports based on the downlink channel information. The downlink channel information may include downlink channel covariance. In some embodiments, the first device 310-1 may extract angle information from the downlink channel covariance. In some embodiments, the first device 310-1 may determine a predefined direction of arrival (DOA) based on the estimated downlink channel information at all receive ports. The downlink channel information may be previously acquired by the first device 310-1. The downlink channel information of each receive port 410 may be determined based on the downlink CSI-RS from the second device 320. In some embodiments, the predefined DOA may be a maximum DOA. It should be noted that the predefined DOA may be any suitable DOA. Note that the compensation parameter may be determined using any suitable method not limited to the above-described embodiments.

[0109] In some embodiments, when an uplink frequency is used, the first device 310-1 may obtain phase information of a reference signal received at the group of receive ports. When a downlink frequency is used, the first device 310-1 may obtain phase information of a reference signal received at the group of receive ports. The first device 310-1 may obtain a phase difference caused by the difference between the uplink frequency and the downlink frequency. For example, the first device 310-1 may determine a first array response of the group of receive ports at the downlink frequency. A second array response of the group of transmit ports at the uplink frequency may be determined. The compensation parameter may be determined based on the first array response, the second array response, and a predetermined DOA.

[0110] At block 630, the first device 310-1 compensates the at least one training reference signal using at least one compensation parameter. In some embodiments, the at least one training reference signal may be phase compensated. Alternatively, the at least one training reference signal may be compensated in other aspects (e.g., frequency domain). It should be noted that the at least one training reference signal may be compensated based on a character difference between a receiving port and a transmitting port of the first device 310-1.

[0111] At block 640, the first device 310-1 transmits at least one compensated training reference signal to the second device 320 via the set of transmit ports. In some embodiments, the information may also indicate another set of receive ports. Alternatively, the first device 310-1 may receive additional information. The additional information may indicate another set of receive ports to be used for compensation.

[0112] At least one additional compensation parameter that associates the other set of receive ports with the set of transmit ports can be determined based on the downlink channel information. The at least one additional compensation parameter can be determined using a method similar to determining the at least one compensation parameter. In some embodiments, the first device 310-1 transmits at least one additional compensated training reference signal to the second device 320.

[0113] Figure 7 1 is a flow chart illustrating an example method 700 implemented at the second device 320 according to some example embodiments of the present disclosure. For discussion purposes, the method 700 will be described from the perspective of the second device 320. It should be noted that dashed boxes are optional.

[0114] In block 710, the second device 320 transmits information to the first device 310-1. This information may indicate a set of receive ports on the first device 310-1 to be used for compensation. This information may be used to trigger the first device 310-1 to compensate for at least one training reference signal and transmit at least one compensated training reference signal. In some embodiments, the information may include an index of a receive port on the first device 310-1. Alternatively, the second device may understand how the receive devices on the first device 310-1 are divided into different groups. In this case, the information may indicate the groups of receive ports to be used for compensation. In other embodiments, the information may indicate the number of receive ports. Alternatively, the information may indicate a mapping between the set of receive ports and a set of transmit ports on the first device. The number of ports in the set of transmit ports may be any suitable number. For example, the set of transmit ports may include one or more transmit ports.

[0115] In some embodiments, the second device 320 may broadcast the information to the first device 310-1. For example, the broadcast information may include triggering the UE's behavior, i.e., compensation of the training reference signal and transmission of the compensated training reference signal. In some embodiments, the information may be sent during the initial access process.

[0116] If the confirmation is received from the first device 310-1, the second device 320 may allocate (multiple) resources for transmitting the compensated training reference signal. The resources may be any suitable resources for transmitting signals. The second device 320 may transmit resource information indicating the allocated resources to the first device 310-1.

[0117] Alternatively, the second device 320 may transmit this information to the first device 310-1 via dedicated signaling. This information may be specific to the first device 310-1. In the event that the second device 320 needs to update the neural network during its operation, for example, due to a specific UE situation that the second device 320 has not encountered before, the second device 320 may need to collect such data to train and enhance the neural network. This UE-specific information may indicate compensation of the training reference signal and transmission of the compensated training reference signal from the terminal device 310-1.

[0118] At block 720, the second device 320 receives at least one compensated training reference signal from the first device 310-1. In some embodiments, the second device 320 may transmit additional information. This additional information may indicate another set of receive ports of the first device. In this way, the characteristics of the receive ports can be reflected by the transmit ports, allowing for more accurate CSI acquisition. In some embodiments, the second device 320 receives at least one additional compensated training reference signal from the first device 310-1.

[0119] In some embodiments, the information may also indicate an additional set of receive ports. Alternatively, the second device 320 may transmit additional information. The additional information may indicate an additional set of receive ports. At least one additional compensation parameter associating the additional set of receive ports with the set of transmit ports may be determined based on the downlink channel information. The at least one additional compensation parameter may be determined using a method similar to determining the at least one compensation parameter. The second device 320 may receive at least one additional compensated training reference signal transmitted via the set of transmit ports of the first device. The additional compensated training reference signal may be compensated using the at least one additional compensation parameter determined based on the downlink channel information.

[0120] In some embodiments, the second device 320 may determine an estimate of a compensated uplink channel between the first device 310-1 and the second device 320 based on the compensated training reference signals.The second device 320 may collect all sets of compensated reference signals from all receive ports.

[0121] In some embodiments, the second device 320 may update a data processing model based on the estimation, the data processing model being used to recover channel information on the downlink channel between the first device 310-1 and the second device 320. For example, the complete compensated uplink channel may be used as input data for a neural network to update the data processing model.

[0122] In some example embodiments, a first device (e.g., first device 310) capable of performing any of method 600 may include a component for performing the corresponding operations of method 600. The component may be implemented in any suitable form. For example, the component may be implemented in a circuit system or a software module. The first device may be implemented as first device 310 or included in first device 310. In some example embodiments, the component may include at least one processor and at least one memory including computer program code. The at least one memory and the computer program code are configured to cause the operation of the device together with the at least one processor.

[0123] In some embodiments, the apparatus includes means for receiving information indicating a set of receive ports at a first device to be used for compensation; means for determining, based on downlink channel information, at least one compensation parameter associating the set of receive ports with a set of transmit ports; means for compensating at least one training reference signal with the at least one compensation parameter; and means for transmitting the at least one compensated training reference signal to a second device via the set of transmit ports of the first device.

[0124] In some embodiments, the apparatus further comprises means for receiving at least one channel state information reference signal (CSI-RS) from the second device; and means for estimating downlink channel information at all receive ports of the first device based on the received CSI-RS, the downlink channel information comprising downlink channel covariance.

[0125] In some embodiments, the information further indicates a mapping between the set of receive ports and the set of transmit ports to be used for compensation at the first device.

[0126] In some embodiments, the information is used to trigger the first device to compensate for at least one training reference signal and transmit at least one compensated training reference signal.

[0127] In some embodiments, the means for determining at least one compensation parameter includes: means for determining a predefined direction of arrival based on estimated downlink channel information at all receive ports of the first device; means for determining a first array response of the group of receive ports at a downlink frequency; means for determining a second array response of the group of transmit ports at an uplink frequency; and means for determining at least one compensation parameter based on the predefined direction of arrival, the first array response, and the second array response.

[0128] In some embodiments, the predetermined direction of arrival is a maximum direction of arrival.

[0129] In some embodiments, the means for compensating the at least one training reference signal comprises means for phase compensating the at least one training reference signal based on at least one compensation parameter.

[0130] In some embodiments, the information indicates an additional set of receive ports for compensation, and the apparatus further comprises means for determining, based on the downlink channel information, at least one additional compensation parameter associating the additional set of receive ports with the set of transmit ports; means for compensating at least one additional training reference signal with the at least one additional compensation parameter; and means for transmitting the at least one additional compensated training reference signal to the second device via the set of transmit ports of the first device.

[0131] In some embodiments, the information is received via broadcast signaling.The apparatus further comprises means for transmitting an acknowledgement of the information to the second device; and means for receiving resource information from the second device indicating an allocation of resources for transmitting at least one compensated training reference signal.

[0132] In some embodiments, this information is specific to the terminal device.

[0133] In some example embodiments, a second first device (e.g., second device 320) capable of performing any of method 700 may include a component for performing the corresponding operations of method 700. The component may be implemented in any suitable form. For example, the component may be implemented in a circuit system or a software module. The first device may be implemented as second device 320 or included in second device 320. In some example embodiments, the component may include at least one processor and at least one memory including computer program code. The at least one memory and the computer program code are configured to cause the operation of the device together with the at least one processor.

[0134] In some embodiments, the apparatus includes a component for transmitting, at a second device, to a first device, information indicating a set of receive ports to be used for compensation at the first device; and a component for receiving, from the first device, at least one compensated training reference signal transmitted via a set of transmit ports of the first device, the at least one compensated training reference signal being compensated with at least one compensation parameter, the at least one compensation parameter associating the set of receive ports with the set of transmit ports determined based on downlink channel information.

[0135] In some embodiments, the apparatus further comprises a component for transmitting at least one channel state information reference signal (CSI-RS) to the first and second devices, the at least one CSI-RS being used by the first device to estimate downlink channel information at all receiving ports of the first device.

[0136] In some embodiments, the information further indicates a mapping between the set of receiving ports and the set of transmitting ports of the first device.

[0137] In some embodiments, the information is used to trigger the first device to compensate for at least one training reference signal and transmit at least one compensated training reference signal.

[0138] In some embodiments, the information indicates an additional set of receive ports for compensation. The apparatus further comprises means for receiving, from the first device, at least one additional compensated training reference signal transmitted via the set of transmit ports of the first device, the at least one additional compensated training reference signal compensated with at least one additional compensation parameter, the at least one additional compensation parameter associating the additional set of receive ports with the set of transmit ports determined based on the downlink channel information.

[0139] In some embodiments, the information is transmitted via broadcast signaling.The apparatus further comprises means for receiving an acknowledgement of the information from the first device; and means for transmitting resource information indicating an allocation of resources for transmitting the at least one compensated training reference signal to the second device.

[0140] In some embodiments, the apparatus further comprises means for determining an estimate of a compensated uplink channel between the first device and the second device based on at least one compensated received training reference signal; and means for updating a data processing model to be used to recover channel information on a downlink channel between the first device and the second device based on the estimate.

[0141] In some embodiments, this information is specific to the terminal device.

[0142] Figure 8 is a simplified block diagram of a device 800 suitable for implementing an example embodiment of the present disclosure. The device 800 may be provided to implement a communication device, such as Figure 3 The first device 310 or the second device 320 is shown. As shown, the device 800 includes one or more processors 810, one or more memories 820 coupled to the processors 810, and one or more communication modules 840 coupled to the processors 810.

[0143] The communication module 840 is configured for bidirectional communication. The communication module 840 has one or more communication interfaces to facilitate communication with one or more other modules or devices. A communication interface may represent any interface necessary for communication with other network elements. In some example embodiments, the communication module 840 may include at least one antenna.

[0144] Processor 810 can be of any type suitable for the local technology network and, as non-limiting examples, can include one or more of the following: a general-purpose computer, a special-purpose computer, a microprocessor, a digital signal processor (DSP), and a processor based on a multi-core processor architecture. Device 800 can have multiple processors, such as application-specific integrated circuit chips that are time-slave to a clock synchronized with a main processor.

[0145] The memory 820 may include one or more non-volatile memories and one or more volatile memories. Examples of non-volatile memories include, but are not limited to, read-only memory (ROM) 824, electrically programmable read-only memory (EPROM), flash memory, hard disks, compact disks (CDs), digital video disks (DVDs), optical disks, laser disks, and other magnetic and / or optical storage. Examples of volatile memories include, but are not limited to, random access memory (RAM) 722 and other volatile memories that do not persist during power outages.

[0146] Computer program 830 includes computer-executable instructions executed by associated processor 810. Program 830 may be stored in a memory (e.g., ROM 824). Processor 810 may perform any suitable actions and processes by loading program 830 into RAM 822.

[0147] The exemplary embodiments of the present disclosure may be implemented by the program 830 so that the device 800 may execute the reference Figures 5 to 7 Any process of the present disclosure discussed. Example embodiments of the present disclosure may also be implemented by hardware or a combination of software and hardware.

[0148] In some example embodiments, the program 830 may be tangibly embodied in a computer-readable medium that may be included in the device 800 (such as in the memory 820) or in other storage devices accessible to the device 800. The device 800 may load the program 830 from the computer-readable medium into the RAM 822 for execution. The computer-readable medium may include any type of tangible, non-volatile memory, such as ROM, EPROM, flash memory, hard disks, CDs, DVDs, and other magnetic and / or optical storage devices. Figure 9 An example of a computer readable medium 900 in the form of an optical storage disc is shown. The computer readable medium has a program 830 stored thereon.

[0149] In general, various embodiments of the present disclosure may be implemented using hardware or dedicated circuits, software, logic, or any combination thereof. Some aspects may be implemented using hardware, while other aspects may be implemented using firmware or software that may be executed by a controller, microprocessor, or other computing device. Although various aspects of the embodiments of the present disclosure are illustrated and described as block diagrams, flow charts, or using some other graphical representations, it should be understood that, as non-limiting examples, the blocks, devices, systems, techniques, or methods described herein may be implemented using hardware, software, firmware, dedicated circuits or logic, general-purpose hardware or a controller or other computing device, or some combination thereof.

[0150] The present disclosure also provides at least one computer program product tangibly stored on a non-transitory computer readable storage medium. The computer program product includes computer executable instructions, such as instructions included in a program module, which are executed in a device on a target physical or virtual processor to perform the above-referenced Figures 5 to 7 Any method described herein. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, etc. that perform specific tasks or implement specific abstract data types. In various embodiments, the functionality of program modules can be combined or split between program modules as needed. The machine-executable instructions of program modules can be executed on local or distributed devices. In distributed devices, program modules can be located in both local and remote storage media.

[0151] The program code for executing the disclosed method can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device so that the program code causes the function / operation specified in the flow chart and / or block diagram to be realized when executed by the processor or controller. The program code can be executed entirely on the machine, partially on the machine, as an independent software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0152] In the context of the present disclosure, computer program codes or related data may be carried by any suitable carrier to enable a device, apparatus or processor to perform various processes and operations as described above. Examples of carriers include signals, computer-readable media, etc.

[0153] Computer readable media can be computer readable signal media or computer readable storage media.Computer readable media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing.More specific examples of computer readable storage media will include electrical connections with one or more wires, portable computer floppy disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disc read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0154] In addition, although operations are described in a particular order, this should not be understood as requiring such operations to be performed in the particular order shown or in sequence or performing all of the operations shown to obtain the desired result. In some cases, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details have been included in the above discussion, these should not be interpreted as limiting the scope of this disclosure, but rather as descriptions of features that may be specific to a particular embodiment. Certain features described in the context of a separate embodiment may also be implemented in combination in a single embodiment. On the contrary, the various features described in the context of a single embodiment may also be implemented in multiple embodiments individually or in any suitable sub-combination.

[0155] Although the disclosure has been described in language specific to structural features and / or methodological acts, it should be understood that the disclosure defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

1. A method for compensating a reference signal for training a neural network, comprising: receiving information indicating a set of receive ports to be used for compensation at a first device, the information further indicating a mapping between the set of receive ports to be used for compensation and a set of transmit ports at the first device, the first device being operated in a frequency division duplex mode; Determining at least one compensation parameter that associates the set of receive ports with a set of transmit ports based on downlink channel information, comprising: determining a predefined direction of arrival based on the downlink channel information estimated at all receive ports of the first device; determining a first array response of the set of receive ports at a downlink frequency; determining a second array response of the set of transmit ports at an uplink frequency; and determining the at least one compensation parameter based on the predefined direction of arrival, the first array response, and the second array response; compensating at least one training reference signal with the at least one compensation parameter for use in training a neural network at a second device; and At least one compensated training reference signal is transmitted to a second device via the set of transmission ports, so that the second device determines an estimate of a compensated uplink channel between the first device and the second device based on the at least one compensated training reference signal, and updates a data processing model of the neural network of the second device based on the estimate of the compensated uplink channel, wherein the data processing model is used to recover channel information on a downlink channel between the first device and the second device.

2. The method according to claim 1, further comprising: receiving at least one channel state information reference signal (CSI-RS) from the second device; as well as The downlink channel information at all reception ports of the first device is estimated based on the received CSI-RS, the downlink channel information including downlink channel covariance. 3 . The method according to claim 1 , wherein the information is used to trigger the first device to compensate for the at least one training reference signal and transmit the at least one compensated training reference signal. The method according to claim 1 , wherein the predefined direction of arrival is a maximum direction of arrival.

5. The method of claim 1 , wherein compensating the at least one training reference signal comprises: The at least one training reference signal is phase compensated based on the at least one compensation parameter.

6. The method of claim 1 , wherein the information indicates an additional set of receive ports for compensation, and wherein the method further comprises: determining at least one further compensation parameter associating the further set of receive ports with the set of transmit ports based on the downlink channel information; compensating at least one further training reference signal with the at least one further compensation parameter; as well as The at least one further compensated training reference signal is transmitted to the second device via the set of transmission ports of the first device of the first device.

7. The method of claim 1 , wherein the information is received via broadcast signaling, and wherein the method further comprises: transmitting a confirmation of the information to the second device; as well as Resource information indicating an allocation of resources for transmitting the at least one compensated training reference signal is received from the second device. The method of claim 1 , wherein the information is specific to the first device.

9. The method of claim 1, wherein the first device comprises a terminal device, and the second device comprises a network device.

10. A method for compensating a reference signal for training a neural network, comprising: At the second device, in response to a need to update a neural network of the second device during operation, transmitting to the first device information indicating a set of receive ports to be used for compensation at the first device, a data processing model of the neural network being used to recover channel information on a downlink channel between the first device and the second device; as well as receiving, from the first device, at least one compensated training reference signal transmitted via a set of transmit ports of the first device, the at least one compensated training reference signal compensated with at least one compensation parameter, the at least one compensation parameter associating the set of receive ports with the set of transmit ports determined based on downlink channel information and determined by the first device using the following operations: determining a predefined direction of arrival based on the downlink channel information estimated at all receive ports of the first device; determining a first array response of the set of receive ports at a downlink frequency; determining a second array response of the set of transmit ports at an uplink frequency; and determining the at least one compensation parameter based on the predefined direction of arrival, the first array response, and the second array response; determining an estimate of a compensated uplink channel between the first device and the second device based on the at least one compensated training reference signal; as well as The data processing model of the second device is updated based on the estimate of the compensated uplink channel.

11. The method according to claim 10, further comprising: At least one channel state information reference signal (CSI-RS) is transmitted to a first and a second device, where the at least one CSI-RS is used by the first device to estimate the downlink channel information at all receiving ports of the first device.

12. The method of claim 10, wherein the information is used to trigger the first device to compensate for the at least one training reference signal and transmit the at least one compensated training reference signal.

13. The method of claim 10, wherein the information indicates an additional set of receive ports for compensation, and wherein the method further comprises: and receiving, from the first device, at least one additional compensated training reference signal transmitted via the set of transmit ports of the first device, the at least one additional compensated training reference signal compensated with at least one additional compensation parameter associating the additional set of receive ports with the set of transmit ports determined based on the downlink channel information.

14. The method of claim 10, wherein the information is transmitted via broadcast signaling, and wherein the method further comprises: receiving a confirmation of the information from the first device; as well as Resource information indicating an allocation of resources for transmitting the at least one compensated training reference signal is transmitted to the second device. The method of claim 10 , wherein the information is specific to the first device.

16. The method of claim 10, wherein the first device comprises a terminal device, and the second device comprises a network device.

17. An apparatus for compensating a reference signal for training a neural network, comprising means for: receiving information indicating a set of receive ports to be used for compensation at a first device, the information further indicating a mapping between the set of receive ports to be used for compensation and a set of transmit ports at the first device, the first device being operated in a frequency division duplex mode; Determining at least one compensation parameter that associates the set of receive ports with a set of transmit ports based on downlink channel information, comprising: determining a predefined direction of arrival based on the downlink channel information estimated at all receive ports of the first device; determining a first array response of the set of receive ports at a downlink frequency; determining a second array response of the set of transmit ports at an uplink frequency; and determining the at least one compensation parameter based on the predefined direction of arrival, the first array response, and the second array response; compensating at least one training reference signal with the at least one compensation parameter for use in training a neural network at a second device; and At least one compensated training reference signal is transmitted to a second device via the set of transmission ports, so that the second device determines an estimate of a compensated uplink channel between the first device and the second device based on the at least one compensated training reference signal, and updates a data processing model of the neural network of the second device based on the estimate of the compensated uplink channel, wherein the data processing model is used to recover channel information on a downlink channel between the first device and the second device.

18. An apparatus for compensating a reference signal for training a neural network, comprising means for: At the second device, in response to a need to update a neural network of the second device during operation, transmitting to the first device information indicating a set of receive ports to be used for compensation at the first device, a data processing model of the neural network being used to recover channel information on a downlink channel between the first device and the second device; as well as receiving, from the first device, at least one compensated training reference signal transmitted via a set of transmit ports of the first device, the at least one compensated training reference signal compensated with at least one compensation parameter, the at least one compensation parameter associating the set of receive ports with the set of transmit ports determined based on downlink channel information and determined by the first device using the following operations: determining a predefined direction of arrival based on the downlink channel information estimated at all receive ports of the first device; determining a first array response of the set of receive ports at a downlink frequency; determining a second array response of the set of transmit ports at an uplink frequency; and determining the at least one compensation parameter based on the predefined direction of arrival, the first array response, and the second array response; determining an estimate of a compensated uplink channel between the first device and the second device based on the at least one compensated training reference signal; as well as The data processing model of the second device is updated based on the estimate of the compensated uplink channel.

Images