Low-overhead CSI feedback

Abstract

Example embodiments of the present disclosure relate to devices, methods, apparatuses, and computer-readable storage media for low-overhead channel state information (CSI) feedback. The method includes, in response to receiving a reference signal from a second device on a channel between the first device and the second device, obtaining a first component and a second component from the reference signal, the first component and the second component being orthogonal to each other; determining a first transformed component and a second transformed component for characterizing the channel by performing a transform on the first component and the second component; generating a parameter set associated with characteristics of the channel by quantizing the first transformed component and the second transformed component; and transmitting the parameter set to the second device. The proposed CSI feedback can even be flexibly combined with existing overhead reduction schemes (such as beamformed CSI‑RS and CSI compression) to further reduce CSI overhead. In addition, the proposed scheme ensures very fast processing at the UE and reduces latency.

Description

Technical Field

[0001] Embodiments of the present disclosure relate generally to the field of telecommunications, and more particularly to devices, methods, apparatuses, and computer-readable media for low-overhead channel state information (CSI) feedback. Background Art

[0002] Recently, machine learning (ML)-based multiple-input multiple-output (MIMO) has received widespread attention and demonstrated its advantages, especially in physical layer solutions such as beamforming and CSI acquisition. ML-based massive MIMO solutions can provide performance enhancements and reduce computational complexity, overhead, and latency.

[0003] In 6G wireless technology research, one of the key themes in future MIMO technology is MIMO that supports artificial intelligence (AI) / ML. ML for massive MIMO is considered a driving force and a leader in future standardization strategies. One of the open issues is reducing CSI overhead. Summary of the Invention

[0004] In general, example embodiments of the present disclosure provide a low-overhead CSI feedback solution.

[0005] 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, wherein the at least one memory and the computer program code are configured to, together with the at least one processor, cause the first device to at least: in response to receiving a reference signal from the second device on a channel between the first device and the second device, obtain a first component and a second component from the reference signal, the first component and the second component being orthogonal to each other; determine first and second transformed components for characterizing the channel by performing a transform on the first and second components; generate a parameter set associated with a characteristic of the channel by quantizing the first and second transformed components; and transmit the parameter set to the second device.

[0006] 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, wherein the at least one memory and the computer program code are configured to, together with the at least one processor, cause the second device to at least: transmit a reference signal to the first device over a channel between the first device and the second device; receive a parameter set associated with characteristics of the channel from the first device, the parameter set being generated at the first device by transforming and quantizing a first component and a second component obtained from the reference signal, the first component and the second component being orthogonal to each other; and determine channel state information of the channel based on the parameter set.

[0007] In a third aspect, a method is provided. The method includes: in response to receiving a reference signal from the second device on a channel between the first device and the second device, obtaining a first component and a second component from the reference signal, the first component and the second component being orthogonal to each other; determining a first transformed component and a second transformed component for characterizing the channel by performing a transform on the first component and the second component; generating a parameter set associated with a characteristic of the channel by quantizing the first transformed component and the second transformed component; and transmitting the parameter set to the second device.

[0008] In a fourth aspect, a method is provided, comprising: transmitting a reference signal to the first device over a channel between the first device and a second device; receiving, from the first device, a parameter set associated with characteristics of the channel, the parameter set generated at the first device by transforming and quantizing a first component and a second component obtained from the reference signal, the first component and the second component being orthogonal to each other; and determining channel state information of the channel based on the parameter set.

[0009] In a fifth aspect, an apparatus is provided, comprising: a component for obtaining a first component and a second component from a reference signal in response to receiving the reference signal from the second device on a channel between the first device and the second device, the first component and the second component being orthogonal to each other; a component for determining a first transformed component and a second transformed component for characterizing the channel by performing a transform on the first component and the second component; a component for generating a parameter set associated with a characteristic of the channel by quantizing the first transformed component and the second transformed component; and a component for transmitting the parameter set to the second device.

[0010] In a sixth aspect, an apparatus is provided, comprising: a component for transmitting a reference signal to the first device on a channel between the first device and the second device; a component for receiving a set of parameters associated with characteristics of the channel from the first device, the set of parameters being generated at the first device by transforming and quantizing a first component and a second component obtained from the reference signal, the first component and the second component being orthogonal to each other; and a component for determining channel state information of the channel based on the parameter set.

[0011] In a seventh aspect, a computer readable medium is provided, the medium comprising a computer program for causing an apparatus to at least perform the method according to the third aspect.

[0012] In an eighth aspect, there is provided a computer readable medium comprising a computer program for causing an apparatus to at least perform the method according to the fourth aspect.

[0013] 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

[0014] Some example embodiments will now be described with reference to the accompanying drawings, in which:

[0015] Figure 1 shows an example communication network in which embodiments of the present disclosure may be implemented;

[0016] Figure 2 An example structure of a MU-MIMO system according to some example embodiments of the present disclosure is shown;

[0017] Figure 3 A signaling diagram illustrating a CSI feedback procedure according to some example embodiments of the present disclosure is shown;

[0018] Figure 4 An example of a subframe for triggering a one-bit quantized CSI report according to some example embodiments of the present disclosure is shown;

[0019] Figure 5 An example structure for generating one-bit quantized CSI feedback according to some example embodiments of the present disclosure is shown;

[0020] Figure 6 An example structure for recovering CSI feedback according to some example embodiments of the present disclosure is shown;

[0021] Figure 7 shows the normalized mean square error (NMSE) performance in example simulations of recovering CSI from one-bit quantized CSI feedback using a deep neural network (NN);

[0022] Figure 8A and Figure 8B shows the throughput performance for different cases in example simulations using deep NNs;

[0023] Figure 9 A flowchart illustrating a method implemented at a first device according to some example embodiments of the present disclosure is shown;

[0024] Figure 10 A flowchart illustrating a method implemented at a second device according to some example embodiments of the present disclosure is shown;

[0025] Figure 11 shows a simplified block diagram of an apparatus suitable for implementing some other embodiments of the present disclosure; and

[0026] Figure 12A block diagram of an example computer-readable medium is shown, according to some example embodiments of the present disclosure.

[0027] Throughout the drawings, the same or similar reference numerals refer to the same or similar elements. DETAILED DESCRIPTION

[0028] 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 disclosure described herein can be implemented in various other ways in addition to the way described below.

[0029] 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.

[0030] 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 an example 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.

[0031] 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 merely used to distinguish the functions of the various elements. As used herein, the term "and / or" includes any and all combinations of one or more of the listed terms.

[0032] 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 the 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.

[0033] As used in this application, the term "circuitry" may refer to one or more or all of the following:

[0034] (a) a pure hardware circuit implementation (such as an implementation using only analog and / or digital circuitry), and

[0035] (b) a combination of hardware circuitry and software such as (as applicable):

[0036] (i) a combination of analog and / or digital hardware circuits and software / firmware, and

[0037] (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

[0038] (c) Hardware circuit(s) and / or processor(s), such as microprocessor(s) or portion(s) of microprocessor(s), that require software (e.g., firmware) to operate, but which may not be present when not required for operation.

[0039] 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.

[0040] As used herein, the term "communication network" refers to a network that complies with any suitable communication standard, such as a fifth generation (5G) system, 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 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, future fifth generation (5G) new radio (NR) communication protocol, 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 considered to be limited to the above-mentioned systems.

[0041] 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 NodeB (NodeB or NB), an evolved NodeB (eNodeB or eNB), a NR next-generation NodeB (gNB), a remote radio unit (RRU), a radio head (RH), a remote radio head (RRH), a relay, a low-power node (such as a femto, pico), etc., depending on the terminology and technology applied. The RAN split architecture includes a gNB-CU (centralized unit, which hosts RRC, SDAP, and PDCP), which controls multiple gNB-DUs (distributed units, which host RLC, MAC, and PHY). A relay node may correspond to the DU portion of an IAB node.

[0042] 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. The terminal device may also correspond to the mobile terminal (MT) part of an integrated access and backhaul (IAB) node (also known as a relay node). In the following description, the terms "terminal device", "communication device", "terminal", "user equipment" and "UE" may be used interchangeably.

[0043] While the functionality described herein may be performed in fixed and / or wireless network nodes in various example embodiments, in other example embodiments, the functionality may be performed in a user equipment device (such as a mobile phone or tablet or laptop or desktop computer or mobile IoT device or fixed IoT device). For example, the user equipment device may be equipped with the corresponding capabilities described in conjunction with (multiple) fixed and / or wireless network nodes (as appropriate). The user equipment device may be a user device and / or a control device (such as a chipset or processor) that is configured to control the user device when installed in the user device. Examples of such functionality include a boot server function and / or a home subscriber server that may be implemented in the user equipment device by providing software to the user equipment device that is configured to cause the user equipment device to perform operations from the perspective of these functions / nodes.

[0044] Figure 1 1 shows an example communication network 100 in which embodiments of the present disclosure may be implemented. Figure 1 As shown, the communication network 100 includes terminal devices 110-1 and 110-2 (hereinafter collectively referred to as first device 110 or UE 110) and network device 120 (hereinafter also referred to as second device 120 or gNB 120). Terminal devices 110-1 and 110-2 can communicate with network device 120. It should be understood that Figure 1 The number of network devices and terminal devices shown is given for ease of illustration and is not limiting. Communication network 100 may include any suitable number of network devices and terminal devices.

[0045] Depending on the communication technology, network 100 can be a code division multiple access (CDMA) network, a time division multiple access (TDMA) network, a frequency division multiple access (FDMA) network, an orthogonal frequency division multiple access (OFDMA) network, a single carrier frequency division multiple access (SC-FDMA) network, etc. The communications discussed in network 100 can conform to any suitable standard, including but not limited to New Radio Access (NR), Long Term Evolution (LTE), LTE Evolution, LTE Advanced (LTE-A), Wideband Code Division Multiple Access (WCDMA), Code Division Multiple Access (CDMA), cdma2000, and Global System for Mobile Communications (GSM). In addition, communications can be performed according to any generation of communication protocols currently known or to be developed in the future. Examples of communication protocols include but are not limited to first generation (1G), second generation (2G), 2.5G, 2.75G, third generation (3G), fourth generation (4G), 4.5G, and fifth generation (5G) communication protocols. The techniques described herein can be used for the above-mentioned wireless networks and radio technologies as well as other wireless networks and radio technologies. For clarity, certain aspects of the techniques are described below for LTE, and LTE terminology is used in much of the description below.

[0046] Massive MIMO is a key technology in 5G communication systems due to its enormous potential for further exploring array gain, multiplexing gain, and interference suppression capabilities. A key factor in realizing these advantages is the design of an accurate CSI acquisition scheme for efficient and robust beamforming. As the number of antenna ports increases, the CSI to be estimated at the UE becomes larger in dimensionality. Once the UE acquires downlink CSI, feeding back the entire explicit CSI improves MIMO transmission performance, but also incurs significant overhead.

[0047] In current systems, the number of quantization bits for CSI feedback is, for example, 8 bits. This remains a significant factor limiting CSI feedback overhead. Lower quantization levels lead to inaccurate CSI acquisition, which in turn degrades system performance. Therefore, existing massive MIMO systems require advanced CSI feedback mechanisms to improve performance while significantly reducing overhead.

[0048] Analog CSI feedback is considered a promising alternative solution for further wireless communications. It involves the UE transmitting estimated downlink channel coefficients using unquantized quadrature amplitude modulation. This is a scaled version of common downlink training observations, such as those from the CSI-RS. The gNB estimates the downlink channel based on the UE's forwarded unquantized CSI. Compared to digital feedback, analog feedback does not involve quantization, codebook mapping, or coding, which can reduce UE processing complexity. However, analog feedback suffers from power variation and channel recovery issues. The overhead is also related to the number of antennas and the transmission overhead of the downlink CSI-RS. This is also difficult to implement in systems with larger bandwidths. Furthermore, it would require significant standardization revisions.

[0049] Therefore, the present disclosure proposes a solution for implementing one-bit quantized CSU feedback. On the UE side, the in-phase component and the quadrature component can be extracted from the received reference signal. The in-phase component and the quadrature component can be transformed and quantized respectively to generate one-bit quantized raw CSI feedback. On the gNB side, the one-bit quantized raw CSI feedback can be input to the ML model, where the CSI can be recovered from the one-bit quantized raw CSI feedback. The proposed CSI feedback can even be flexibly combined with existing overhead reduction schemes (such as beamformed CSI-RS and CSI compression) to further reduce CSI overhead. In addition, the proposed scheme ensures very fast processing at the UE and reduces latency.

[0050] The following will refer to Figures 2 to 4 The principles and implementations of the present disclosure are described in detail. Figure 2 An example structure 200 of a MU-MIMO system according to some example embodiments of the present disclosure is shown.

[0051] like Figure 2 As shown, gNB 120 is equipped with M T gNB 120 can simultaneously serve K users, i.e., UEs 110-1 to 110-K, where each UE can have The complete channel 130 in this structure can be expressed as in is the total number of receive antennas from all UEs 110 - 1 to 110 -K. For the k-th UE, the millimeter wave (mmWave) channel is represented by the widely used cluster model H.

[0052] gNB 120 may require H = [H (1) , H (2) ,...,H (K)] Design transmission processing, such as precoding, for the K UEs. In an FDD system, the downlink channel H can be obtained via CSI feedback from UEs 110-1 to 110-K.

[0053] Assume that CSI-RS is The received signal at UE 110 can be written as:

[0054]

[0055] Where P is the power of the transmitted CSI-RS, N RS is the RS length, N is the additive white Gaussian noise (AWGN), and other interference has zero mean and power spectral density N0.

[0056] Figure 3 FIG1 shows a signaling diagram for event-triggered measurement of CSI-RS according to some example embodiments of the present disclosure. Figure 1 Describe process 300. Process 300 may involve Figure 1 UE 110 and gNB 120 are shown. It is worth noting that although process 300 has been Figure 1 The process is described in the communication network 100, but the process can also be applied to other communication scenarios.

[0057] To obtain a one-bit quantized CSI report, gNB 120 may trigger a one-bit quantized CSI report before transmitting a reference signal to UE 110 and send the trigger to a dedicated UE. Figure 3 As shown, gNB 120 transmits 305 a trigger to UE 110 to initiate quantization. The trigger may initiate UE 110 to perform a one-bit quantization process and transmit the one-bit quantization result to gNB 120.

[0058] Figure 4 An example of a subframe for triggering a one-bit quantized CSI report according to some example embodiments of the present disclosure is shown. Figure 4 As shown, a trigger may be included in downlink transmission block 410. An additional downlink transmission block 420 for transmitting a reference signal (CSI-RS) may follow transmission block 410. Based on the trigger, UE 110 may perform one-bit quantization on the CSI-RS and generate a one-bit CSI report. The one-bit CSI report may be included in an uplink transmission block 430.

[0059] In some example embodiments, the trigger may indicate a one-bit quantization scheme for the reference signal.For example, the trigger may be transmitted to UE 110 via downlink control information (DCI) or a medium access control element (MAC CE).

[0060] Reference again Figure 3 , when the UE 110 has received a trigger for reporting one-bit quantized CSI, the UE 110 may perform 310 one-bit quantization if the UE receives a reference signal, such as a CSI-RS.

[0061] Figure 5 An example structure for generating one-bit quantized CSI feedback according to some example embodiments of the present disclosure is shown. Figure 5 As shown, the received reference signal may be processed by an extractor 510 to extract an in-phase component and a quadrature component that are orthogonal to each other. The in-phase component and the quadrature component may be converted by a first analog-to-digital converter (ADC) 511 and a second ADC 512, respectively.

[0062] UE 110 may then perform a transform of the in-phase component and the quadrature component to determine a transformed in-phase component and a transformed quadrature component for characterizing the channel.

[0063] For example, the in-phase component and the quadrature component can be transformed into the code correlation domain by the first code matched filter (CMF) 521 and the second CMF 522 associated with the CSI-RS sequence, respectively. Hereinafter, the term "code correlation domain" may refer to the code domain in which the characteristics of the channel can be maximized. The resulting signal is represented as:

[0064]

[0065] The transmission reference signal X is designed to satisfy The transmission SNR of CSI-RS can be given by the following formula:

[0066]

[0067] After the transformation, the transformed in-phase component and the transformed quadrature component may be quantized by a quantizer 530 (e.g., a one-bit quantizer 630). Then, the one-bit quantized original CSI may be obtained and given by the following formula:

[0068]

[0069] One-bit quantization can be performed by binarizing the values ​​of a set of elements in the transformed in-phase component into a set of bits, and binarizing the values ​​of a set of elements in the transformed quadrature component into a set of bits. (In-phase component and quadrature component) may take values ​​in {±1} or {1,0}. UE 110 may generate a parameter set associated with characteristics of a channel through one-bit quantization.

[0070] It should be understood that for executing Figure 5 The illustrated one-bit quantization structure 500 is merely one embodiment. Any suitable units and modules may be added to the structure 500. For example, in some exemplary embodiments, when the downlink channel quality is poor, a denoising unit may be included in the structure. The transformed in-phase component and the transformed quadrature component may be denoised prior to quantization.

[0071] Reference again Figure 3 , UE 110 may transmit to gNB 120 a parameter set 315 associated with the characteristics of the channel, i.e., a one-bit quantized raw CSI feedback Depending on the CSI feedback configuration, one-bit quantized raw CSI feedback can be transmitted in the Physical Uplink Control Channel (PUCCH) or the Physical Uplink Shared Channel (PUSCH).

[0072] gNB 120 can then select the raw CSI feedback from a parameter set associated with the characteristics of the channel (i.e., one-bit quantized ) 320 CSI H. gNB 120 can recover CSI using a one-bit CSI recovery neural network. The recovery neural network can characterize the association between channel state information and reference parameters describing the channel. Based on this association and a parameter set associated with the characteristics of the channel, channel state information can be determined.

[0073] Figure 6 An example structure 600 for recovering CSI feedback according to some example embodiments of the present disclosure is shown. In the structure 600, for example, the recovery neural network 620 may include layers 621 and 622. It should be understood that Figure 6 The restoration neural network 620 shown is only an example. The restoration neural network 620 may include any other layers.

[0074] The input 610 received from the UE 110 (i.e., one-bit quantized raw CSI feedback ) may be input into the restoration neural network 620, and the output 630 of the restoration neural network 620 may be the restored CSI H. In the restoration neural network 620, the input is represented as a three-dimensional tensor symbol, where the third dimension corresponds to the in-phase component and the quadrature component of the one-bit quantized CSI.

[0075] Restoration neural network 620 may be well-trained. For example, gNB 120 may generate a simulated reference signal for transmission on a simulated channel between UE 110 and gNB 120, and determine a simulated receive signal based on the simulated reference signal and predetermined parameters associated with channel state information of the simulated channel.

[0076] Based on one-bit quantization similar to the one-bit quantization performed at the UE, gNB 120 can determine a set of simulated parameters representing the simulated channel based on the simulated reference signal. Recovery neural network 620 can be trained using the set of simulated parameters representing the simulated channel and predetermined parameters associated with channel state information of the simulated channel.

[0077] To accommodate one-bit raw CSI feedback with varying SNRs, an enhanced neural network can be trained for varying SNRs with a certain granularity. Since the input data has similar characteristics but only varies with SNR, there is no need to train the entire deep neural network.

[0078] The neural network consists of a concatenated CNN unit followed by a fully connected and rectified linear unit (ReLU) layer. The CNN unit acts as a feature extractor, and the fully connected and ReLU layers attempt to recover the data based on the extracted features.

[0079] Therefore, a training strategy for an enhanced neural network is proposed. The enhanced neural network consists of a feature extraction layer and a key recovery layer. In this example, the feature extraction layer includes a series of CNN units, and the key recovery layer is a fully connected layer and a Reluctant Unit (ReLU) layer. Using input data with a fixed SNR as the training set, the enhanced neural network is trained from scratch, and a pre-trained neural network is obtained. Input data with different SNRs can be added to the training set, but only the key recovery layer is trained. Thus, an upgraded neural network can be obtained. This method can reduce a lot of work, resulting in a fast and simple training process.

[0080] In addition, the proposed scheme can also be applied to the case of compressed CSI. Then, the compressed noisy The original compression is passed through a one-bit quantizer and quantized by one bit Can be obtained.

[0081] To recover CSI, unlike the uncompressed CSI case, a pre-processing unit is required to reconstruct the received compressed CSI based on the compression information reported at the UE. For example, the UE can perform CSI compression based on a common codebook and feed back the selected codebook index to inform the gNB of the compression so that the gNB can perform pre-reconstruction. Therefore, the input of the neural network should be pre-reconstructed Neural networks should be trained by including this compression case.

[0082] The proposed solution of this disclosure has minimal requirements on current systems, but can provide promising performance through this differentiation by leveraging the advantages of deep neural networks. The proposed CSI feedback can even be flexibly combined with existing overhead reduction schemes (such as beamformed CSI-RS and CSI compression) to further reduce CSI overhead.

[0083] The proposed scheme supports various UEs to achieve one-bit quantization on non-estimated channels (noisy CSI-RS) in the code correlation domain, and denoises the received CSI-RS after transforming it into the code correlation domain.

[0084] Compared to current digital CSI feedback, the proposed CSI is raw CSI, significantly reducing the signal processing workload at the UE. Since precise channel estimation is not required at the UE, it ensures very fast processing at the UE and reduces latency. The UE only needs to perform "code-matched filtering" (with respect to the CSI-RS sequence), one-bit quantization, and forwarding.

[0085] Compared to analog feedback, the proposed CSI is digital and does not suffer from power variations due to power scaling in the uplink. Feedback overhead is also reduced, as the overhead of analog feedback increases with the transmit antenna (port). The channel recovery performance of analog feedback is affected by channel variations in both the downlink and uplink, while deep learning-based neural networks are more robust in processing noisy CSI.

[0086] As described above, simulations were performed to test one-bit quantization. A millimeter wave multi-user MIMO system can be considered as an environment for simulation. Figure 7 The normalized mean square error (NMSE) performance in example simulations of recovering CSI from one-bit quantized CSI feedback using a deep neural network (NN) is shown. Figure 8A and Figure 8B The throughput performance of different cases in an example simulation using a deep NN is shown. The simulation setup is as follows:

[0087] Table 1: Simulation settings

[0088]

[0089] In this simulation, deep CNN is implemented by MATLAB2018a. It can be considered to use 4 CNN units for feature extraction, and in the convolution layer, L of consecutive units are applied respectively. F= [7, 16, 32, 64] filters. Two fully connected layers were applied after the cascaded CNN units with 512 and 1024 neurons, respectively, as the key recovery component. The total dataset consisted of 50,000 samples, 85% of which were used for training and the remainder for validation. The training rate was 0.0002, and the gradient descent rate was 0.9. We used a mini-batch size of 512 for gradient descent.

[0090] Figure 7 The normalized mean square error (NMSE) performance of the one-bit quantization CSI acquisition scheme is shown (case 1: curves 701-704 and case 2: 705-708), and the performance is compared with different SNR Tx (i.e., SNR Tx =20dB, SNR Tx =10dB, SNR Tx =5dB and SNR Tx =2dB) and simulated feedback. Tx is the transmission signal-to-noise ratio of CSI-RS. Tx = 20dB means that relatively clean CSI is obtained at the UE, i.e., very good channel conditions, or the estimated channel is performed, or the CSI-RS is received after denoising and correlation. It can be observed that the performance of the estimation using the proposed deep NN decreases with the SNR Tx The performance is better than the analog feedback, which depends on both the uplink and downlink SNR of the RS.

[0091] The simulated CSI feedback can also be considered as a reference solution for comparison. Least squares (LS) channel estimation is performed based on the knowledge of the uplink channel and the power level of the simulated signal. The uplink channel can be estimated via the uplink reference signal, but in this simulation, it is assumed to be perfect. Since the UE needs to amplify and forward the received CSI-RS, the downlink SNR and uplink SNR of the RS will affect the performance. Therefore, consider (SNR DL , SNR UL )=(10,10) dB and (5,5) dB were used for the simulation.

[0092] The CDF of throughput performance is as follows Figure 8A ( Figure 7 Case 1) and Figure 8B ( Figure 7 It can be observed that in case 1, the performance of the proposed scheme (curves 822-825) is very close to that with excellent (curve 821) CSI (≈98% at the median), and the performance degrades slightly when the channel quality is poor, with the median SNR Tx=5dB, the loss is 7%. In case 2, when there are more scatterers in the channel, the performance (curves 812-815) degrades only <8% at the median compared to the excellent (curve 811) case. The UE can adjust its implementation based on the channel quality.

[0093] In addition, the overhead of the CSI feedback scheme is calculated as follows:

[0094] Table 2: Cost Analysis

[0095]

[0096] where {·} denotes the calculation of our example. It is assumed that the training overhead caused by CSI-RS is the same in both cases. The CSI feedback is determined by the dimension of the channel to be fed back. For the proposed one-bit quantized feedback, the overhead can be calculated by (Q = 1 calculation), where Q is the number of quantized bits. In current systems, CSI feedback usually requires 8 bits. For analog feedback, since the UE directly forwards the received CSI-RS, the overhead is determined by the CSI-RS and is related to the channel size. is proportional to the downlink channel [2]. In addition, analog feedback schemes require the uplink channel to recover the downlink channel, and therefore the uplink training overhead depends on the number of antennas at the UE [2]. The scaling O(·) in analog CSI feedback can be modeled by multiplying by a quota, which is an integer and usually greater than 2. Therefore, the overhead of one-bit quantized CSI is smaller than that of analog feedback.

[0097] Figure 9 A flow chart illustrating an example method 900 for service management in a communication system according to some example embodiments of the present disclosure is shown. The method 900 may be Figure 1 For ease of discussion, reference will be made to the first device 110. Figure 1 Describe method 90.

[0098] like Figure 9 As shown, at 910, if the first device receives a reference signal from the second device on a channel between the first device and the second device, the first device obtains a first component and a second component from the reference signal, and the first component and the second component are orthogonal to each other.

[0099] In some example embodiments, the first device may receive a trigger from the second device for initiating quantization and transmission of the parameter set, the trigger indicating at least a predetermined quantization scheme for the reference signal.

[0100] In some example embodiments, the trigger is included in downlink control information or in a control element for medium access control.

[0101] In some example embodiments, the first component is an in-phase component of the reference signal, and the second component is a quadrature component of the reference signal.

[0102] At 920 , the first device determines first and second transformed components for characterizing a channel by performing a transform on the first and second components.

[0103] In some example embodiments, the first device may perform a transform on the first component and the second component in a code-dependent domain. The first device may determine a transform result associated with the first component that represents a channel in the code-dependent domain as a first transformed component, and determine a transform result associated with the second component that represents a channel in the code-dependent domain as a second transformed component.

[0104] At 930 , the first device generates a parameter set associated with characteristics of a channel by quantizing the first transformed component and the second transformed component.

[0105] In some example embodiments, the first device may binarize the values ​​of the first group of elements in the first transformed component into a first group of bits and binarize the values ​​of the second group of elements in the second transformed component into a second group of bits. The first device may generate a parameter set based on the first group of bits and the second group of bits.

[0106] At 940, the first device transmits the parameter set to the second device.

[0107] In some example embodiments, the first device comprises a terminal device, and the second device comprises a network device.

[0108] Figure 10 A flow chart of an example method 1000 for service management in a communication system according to some example embodiments of the present disclosure is shown. The method 1000 may be Figure 1 For ease of discussion, reference will be made to the second device 120. Figure 1 Method 1000 is described.

[0109] like Figure 10 As shown, at 1010 , the second device 120 transmits a reference signal to the first device on a channel between the first device and the second device.

[0110] At 1020, the second device 120 receives a parameter set associated with characteristics of a channel from the first device. The parameter set is generated at the first device by transforming and quantizing a first component and a second component obtained from a reference signal, and the first component and the second component are orthogonal to each other.

[0111] At 1030 , the second device 120 determines channel state information of the channel based on the parameter set.

[0112] In some example embodiments, second device 120 may obtain an association between the channel state information and a reference parameter characterizing the channel, and determine the channel state information based on the parameter set and the association.

[0113] In some example embodiments, second device 120 may generate an additional reference signal for analog transmission on the analog channel between the first device and the second device, and determine an analog receive signal based on the additional reference signal and a predetermined parameter set indicating channel state information of the analog channel. Second device 120 may determine an analog parameter set characterizing the analog channel based on the analog receive signal, and determine the association based on the analog parameter set and the predetermined parameter set.

[0114] In some example embodiments, the second device 120 may obtain a first analog component and a second analog component from an analog received signal, the first analog component and the second analog component being orthogonal to each other; determine a first transformed analog component and a second transformed analog component for characterizing an analog channel by performing a transform on the first analog component and the second analog component; and determine a set of analog parameters associated with characteristics of the analog channel by quantizing the first transformed analog component and the second transformed analog component.

[0115] In some example embodiments, the second device 120 may train a machine learning model indicating an association with the simulation parameter set and the predetermined parameter set.

[0116] In some example embodiments, the first device comprises a terminal device, and the second device comprises a network device.

[0117] In some example embodiments, an apparatus capable of performing method 900 (e.g., implemented at first device 110) may include a component for performing the corresponding steps of method 900. The component may be implemented in any suitable form. For example, the component may be implemented by a circuit system or a software module.

[0118] In some example embodiments, the apparatus includes: means for obtaining a first component and a second component from a reference signal in response to receiving the reference signal from the second device on a channel between the first device and the second device, the first component and the second component being orthogonal to each other; means for determining first and second transformed components for characterizing the channel by performing a transform on the first and second components; means for generating a parameter set associated with characteristics of the channel by quantizing the first and second transformed components; and means for transmitting the parameter set to the second device.

[0119] In some example embodiments, an apparatus capable of performing method 1000 (e.g., implemented at second device 120) may include a component for performing the corresponding steps of method 1000. The component may be implemented in any suitable form. For example, the component may be implemented by a circuit system or a software module.

[0120] In some example embodiments, the apparatus includes means for transmitting a reference signal to the first device over a channel between the first device and the second device; means for receiving, from the first device, a parameter set associated with characteristics of the channel, the parameter set being generated at the first device by transforming and quantizing first and second components obtained from the reference signal, the first and second components being orthogonal to each other; and means for determining channel state information of the channel based on the parameter set.

[0121] Figure 11 is a simplified block diagram of a device 1100 suitable for implementing embodiments of the present disclosure. The device 1100 may be provided to implement a communication device, such as Figure 1 UE 110 and gNB 120 are shown. As shown, device 1100 includes one or more processors 1110, one or more memories 1140 coupled to processor 1110, and one or more transmitters and / or receivers (TX / RX) 1140 coupled to processor 1110.

[0122] The TX / RX 1140 is used for bidirectional communication. The TX / RX 1140 has at least one antenna to facilitate communication. The communication interface can represent any interface required to communicate with other network elements.

[0123] Processor 1110 may be of any type suitable for the local technology network and, as non-limiting examples, may 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 1100 may have multiple processors, such as application-specific integrated circuit chips that are time-slave to a clock synchronized with a main processor.

[0124] Memory 1120 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) 1124, electrically programmable read-only memory (EPROM), flash memory, hard disks, compact disks (CDs), digital video disks (DVDs), and other magnetic and / or optical storage. Examples of volatile memories include, but are not limited to, random access memory (RAM) 1122 and other volatile memories that do not persist during power outages.

[0125] Computer program 1130 includes computer-executable instructions executed by associated processor 1110. Program 1130 may be stored in ROM 1120. Processor 1110 may perform any suitable actions and processes by loading program 1130 into RAM 1120.

[0126] The embodiment of the present disclosure can be implemented by the program 1130 so that the device 1100 can execute the reference Figures 9 and 10 The embodiments of the present disclosure may also be implemented by hardware or a combination of software and hardware.

[0127] In some embodiments, the program 1130 may be tangibly embodied in a computer-readable medium that may be included in the device 1100 (such as in the memory 1120) or in other storage devices accessible to the device 1100. The device 1100 may load the program 1130 from the computer-readable medium into the RAM 1122 for execution. The computer-readable medium may include any type of tangible non-volatile memory, such as ROM, EPROM, flash memory, hard disk, CD, DVD, etc. Figure 12 An example of a computer readable medium 1200 in the form of a CD or DVD is shown. The computer readable medium has a program 1130 stored thereon.

[0128] 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.

[0129] 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 program modules, which are executed in a device on a target real or virtual processor to perform the above-referenced Figures 9 and 10 Methods 900 and 1000 are described. 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.

[0130] 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.

[0131] 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.

[0132] 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.

[0133] 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 separate embodiments may also be combined in a single embodiment. Conversely, the various features described in the context of a single embodiment may also be separated or implemented in any suitable sub-combination in multiple embodiments.

[0134] 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 first device, comprising: at least one processor; as well as 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 at least: In response to receiving a reference signal from the second device on a channel between the first device and the second device, obtaining a first component and a second component from the reference signal, the first component and the second component being orthogonal to each other; determining a first transformed component and a second transformed component for characterizing the channel by performing a transform on the first component and the second component in a code dependent domain; generating a set of parameters associated with characteristics of the channel by quantizing the first transformed component and the second transformed component; and The parameter set is transmitted to the second device.

2. The first device according to claim 1, wherein the first device is further caused to: A trigger for initiating the quantization and transmitting the parameter set is received from the second device, the trigger indicating at least a predetermined quantization scheme for the reference signal. 3 . The first device of claim 2 , wherein the trigger is included in downlink control information or in a control element for medium access control. 4 . The first device of claim 1 , wherein the first component is an in-phase component of the reference signal, and the second component is a quadrature component of the reference signal.

5. The first apparatus of claim 1 , wherein the first apparatus is caused to determine the first transformed component and the second transformed component by: performing the transform on the first component and the second component in the code dependent domain; determining a transformation result associated with the first component representing the channel in the code-related domain as the first transformed component; as well as A transformation result associated with the second component, which represents the channel in the code-dependent domain, is determined as the second transformed component.

6. The first device of claim 1 , wherein the first device is caused to generate the parameter set by: binarizing values ​​of a first group of elements in the first transformed component into a first group of bits; binarizing values ​​of a second group of elements in the second transformed component into a second group of bits; as well as The parameter set is generated based on the first set of bits and the second set of bits.

7. The first device according to claim 1, wherein the first device comprises a terminal device, and the second device comprises a network device.

8. A second device comprising: at least one processor; as well as 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 at least: transmitting a reference signal to the first device over a channel between the first device and the second device; receiving, from the first device, a parameter set associated with characteristics of the channel, the parameter set being generated at the first device by transforming and quantizing, in a code-dependent domain, a first component and a second component derived from the reference signal, the first component and the second component being orthogonal to each other; as well as Channel state information of the channel is determined based on the parameter set.

9. The second device of claim 8, wherein the second device is caused to determine the channel state information by: Acquiring an association between the channel state information and a reference parameter characterizing the channel; and The channel state information is determined based on the parameter set and the association.

10. The second device of claim 9, wherein the second device is further caused to: generating a further reference signal for analog transmission on an analog channel between the first device and the second device; determining an analog receive signal based on the further reference signal and a predetermined set of parameters indicative of channel state information of the analog channel; determining a set of simulation parameters characterizing the simulated channel based on the simulated received signal; and The association is determined based on the simulation parameter set and the predetermined parameter set.

11. The second device of claim 10 , wherein the second device is caused to determine the simulation parameter set by: Obtaining a first analog component and a second analog component from the analog received signal, wherein the first analog component and the second analog component are orthogonal to each other; determining a first transformed analog component and a second transformed analog component for characterizing the analog channel by performing a transform on the first analog component and the second analog component; as well as A set of analog parameters associated with characteristics of the analog channel is determined by quantizing the first transformed analog component and the second transformed analog component.

12. The second device of claim 10, wherein the second device is caused to determine the association by: Training a machine learning model indicating the association with the simulation parameter set and the predetermined parameter set.

13. The second device according to claim 8, wherein the first device comprises a terminal device, and the second device comprises a network device.

14. A method comprising: In response to receiving a reference signal from the second device on a channel between the first device and the second device, obtaining a first component and a second component from the reference signal, the first component and the second component being orthogonal to each other; determining a first transformed component and a second transformed component for characterizing the channel by performing a transform on the first component and the second component in a code dependent domain; generating a set of parameters associated with characteristics of the channel by quantizing the first transformed component and the second transformed component; and The parameter set is transmitted to the second device.

15. The method according to claim 14, further comprising: A trigger for initiating the quantization and transmitting the parameter set is received from the second device, the trigger indicating at least a predetermined quantization scheme for the reference signal.

16. The method of claim 15, wherein the trigger is included in downlink control information or in a control element for medium access control.

17. The method of claim 14, wherein the first component is an in-phase component of the reference signal and the second component is a quadrature component of the reference signal.

18. The method of claim 14, wherein determining the first transformed component and the second transformed component comprises: performing the transform on the first component and the second component in the code dependent domain; determining a transformation result associated with the first component representing the channel in the code-related domain as the first transformed component; as well as A transformation result associated with the second component, which represents the channel in the code-dependent domain, is determined as the second transformed component.

19. The method of claim 14, wherein generating the parameter set comprises: binarizing values ​​of a first group of elements in the first transformed component into a first group of bits; binarizing values ​​of a second group of elements in the second transformed component into a second group of bits; as well as The parameter set is generated based on the first set of bits and the second set of bits.

20. The method of claim 14, wherein the first device comprises a terminal device, and the second device comprises a network device.

21. A method comprising: transmitting a reference signal to the first device over a channel between the first device and the second device; receiving, from the first device, a parameter set associated with characteristics of the channel, the parameter set being generated at the first device by transforming and quantizing, in a code-dependent domain, a first component and a second component derived from the reference signal, the first component and the second component being orthogonal to each other; as well as Channel state information of the channel is determined based on the parameter set.

22. The method of claim 21 , wherein determining the channel state information comprises: Acquire an association between the channel state information and a reference parameter characterizing the channel; as well as The channel state information is determined based on the parameter set and the association.

23. The method of claim 21, further comprising: generating a further reference signal for analog transmission on an analog channel between the first device and the second device; determining an analog receive signal based on the further reference signal and a predetermined set of parameters indicative of channel state information of the analog channel; determining a set of simulation parameters characterizing the simulated channel based on the simulated received signal; and The association is determined based on the simulation parameter set and the predetermined parameter set.

24. The method of claim 23, wherein determining the simulation parameter set comprises: Obtaining a first analog component and a second analog component from the analog received signal, wherein the first analog component and the second analog component are orthogonal to each other; determining a first transformed analog component and a second transformed analog component for characterizing the analog channel by performing a transform on the first analog component and the second analog component; as well as A set of analog parameters associated with characteristics of the analog channel is determined by quantizing the first transformed analog component and the second transformed analog component.

25. The method of claim 23, wherein determining the association comprises: Training a machine learning model indicating the association with the simulation parameter set and the predetermined parameter set.

26. The method of claim 21, wherein the first device comprises a terminal device, and the second device comprises a network device.

27. An apparatus comprising: means for deriving a first component and a second component from a reference signal in response to receiving the reference signal from the second device over a channel between the first device and the second device, the first component and the second component being orthogonal to each other; means for determining a first transformed component and a second transformed component for characterizing the channel by performing a transform on the first component and the second component in a code dependent domain; means for generating a set of parameters associated with characteristics of the channel by quantizing the first transformed component and the second transformed component; as well as Means for transmitting the parameter set to the second device.

28. An apparatus comprising: means for transmitting a reference signal to a first device over a channel between the first device and the second device; means for receiving, from the first device, a parameter set associated with characteristics of the channel, the parameter set being generated at the first device by transforming and quantizing in a code-dependent domain a first component and a second component derived from the reference signal, the first component and the second component being orthogonal to each other; as well as means for determining channel state information for the channel based on the set of parameters.

29. A non-transitory computer-readable medium comprising program instructions for causing an apparatus to at least perform the method according to any one of claims 14 to 20. 30 . A non-transitory computer-readable medium comprising program instructions for causing an apparatus to at least perform the method according to claim 21 .

Images