Downlink Beamforming in MU-MIMO Systems

Abstract

According to an embodiment of the present disclosure, downlink beamforming in a MU-MIMO system is proposed. The embodiment of the present disclosure combines a practical receiver to further mitigate the interference sensed by the terminal device. The beamforming scheme according to the embodiment of the present disclosure has many advantages. For example, the embodiment of the present disclosure solves interference from the perspective of the terminal device to mitigate the interference that actually affects the received signal of the terminal device. The embodiment of the present disclosure does not assume that the terminal device uses an Eigen receiver, but considers that the terminal device uses a more practical receiver to mitigate interference. The embodiment of the present disclosure is highly flexible, and different interference cancellation methods can be used instead of the ZF method.

Description

Technical Field

[0001] Embodiments of the present disclosure relate generally to communication technology, and more particularly to methods, devices, and computer-readable media for downlink beamforming in a Multi-User Multiple-Input Multiple-Output (MU-MIMO) system. Background Art

[0002] With the development of communication technology, MIMO has been proposed. MIMO technology can achieve the multiplication of radio link capacity by using multiple transmit and receive antennas to exploit multipath propagation. There are some factors that affect the performance of MIMO systems, such as interference. Therefore, further research on MIMO technology is needed to improve system performance. Summary of the invention

[0003] In general, embodiments of the present disclosure relate to a method for downlink beamforming in a MU-MIMO system and a corresponding communication device.

[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, together with the at least one processor, cause the first device to determine one or more beamforming vectors for one or more devices in a plurality of co-scheduled devices at the first device. The first device is also caused to iteratively update the one or more beamforming vectors by: determining the transmission properties of the channel based on the one or more beamforming vectors and a channel estimate of the channel between the first device and the plurality of co-scheduled devices; and performing precoding on the transmission properties of the channel to mitigate interference between the plurality of co-scheduled devices. The first device is also caused to perform beamforming on the plurality of co-scheduled devices based on the iteratively updated one or more beamforming vectors.

[0005] In a second aspect, a method is provided. The method includes determining, at a first device, one or more beamforming vectors for one or more devices in a plurality of co-scheduled devices. The method also includes iteratively updating the one or more beamforming vectors by: determining a transmission property of a channel based on the one or more beamforming vectors and a channel estimate of a channel between the first device and the plurality of co-scheduled devices; and performing precoding on the transmission property of the channel to mitigate interference between the plurality of co-scheduled devices. The method also includes performing beamforming on the plurality of co-scheduled devices based on the iteratively updated one or more beamforming vectors.

[0006] In a third aspect, an apparatus is provided. The apparatus includes a component for determining one or more beamforming vectors for at least one device among a plurality of co-scheduled devices at a first device; a component for iteratively updating the one or more beamforming vectors by: determining a transmission property of a channel based on the one or more beamforming vectors and a channel estimate of the channel between the first device and the plurality of co-scheduled devices; and performing precoding on the transmission property of the channel to mitigate interference between the plurality of co-scheduled devices; and a component for performing beamforming on the plurality of co-scheduled devices based on the iteratively updated one or more beamforming vectors.

[0007] In a fourth aspect, a computer-readable medium is provided, the medium comprising program instructions for causing an apparatus to at least execute the method according to the above second aspect.

[0008] 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 easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Some example embodiments will now be described with reference to the accompanying drawings, in which:

[0010] Figure 1 A schematic diagram of a communication system according to an embodiment of the present disclosure is illustrated;

[0011] Figure 2 A schematic diagram of a MIMO system according to an embodiment of the present disclosure is illustrated;

[0012] Figure 3 A flowchart of a method according to an embodiment of the present disclosure is illustrated;

[0013] Figure 4 A schematic diagram illustrating a transceiver layout according to an embodiment of the present disclosure is illustrated;

[0014] Figure 5 A schematic diagram of a transceiver array according to an embodiment of the present disclosure is illustrated;

[0015] Fig. 6A and Figure 6B A simplified block diagram of an implementation structure according to an embodiment of the present disclosure is illustrated;

[0016] Fig. 7A and Figure 7B illustrates a performance graph according to an embodiment of the present disclosure;

[0017] Fig. 8A and Figure 8B illustrates a performance graph according to an embodiment of the present disclosure;

[0018] Fig. 9A and Fig. 9B illustrates a performance graph according to an embodiment of the present disclosure;

[0019] Fig. 10A and Fig. 10B illustrates a performance graph according to an embodiment of the present disclosure;

[0020] Fig.11A and Fig. 11B illustrates a performance graph according to an embodiment of the present disclosure;

[0021] Fig. 12A and Fig. 12B illustrates a performance graph according to an embodiment of the present disclosure;

[0022] Fig.13 illustrates a simplified block diagram of an apparatus suitable for implementing embodiments of the present disclosure; and

[0023] Fig.14 A block diagram of an example computer-readable medium is illustrated according to some example embodiments of the present disclosure.

[0024] Throughout the drawings, the same or similar reference numerals refer to the same or similar elements. DETAILED DESCRIPTION

[0025] The principles of the present disclosure will now be described with reference to some example embodiments. It should be understood that the description of the example embodiments is only for the purpose of illustrating and helping those skilled in the art to understand and implement the present disclosure, and does not represent any limitation on the scope of the present disclosure. The disclosure described herein can be implemented in various other ways except for the way described below.

[0026] 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.

[0027] References in this disclosure to "one embodiment," "an embodiment," "an example embodiment," etc. indicate that the described embodiment may include a particular feature, structure, or characteristic, but not every embodiment necessarily includes 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, it is considered to be 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).

[0028] It should be understood that although the terms "first" and "second" etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. 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.

[0029] The terms used herein are for the purpose of describing specific embodiments only and are not intended to limit the exemplary embodiments. As used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "include", "including", "having", "having" and / or "containing" when used herein specify the presence of the features, elements and / or components, etc., but do not exclude the presence or addition of one or more other features, elements, components and / or combinations thereof.

[0030] As used in this application, the term "circuitry" may refer to one, more, or all of the following:

[0031] (a) a pure hardware circuit implementation (such as an implementation using only analog and / or digital circuitry), and

[0032] (b) a combination of hardware circuitry and software such as (where applicable):

[0033] (i) a combination of analog and / or digital hardware circuits and software / firmware, and

[0034] (ii) any portion of hardware processor(s) (including digital signal processor(s)), software and memory(s) with software that work together to cause a device (such as a mobile phone or server) to perform various functions, and

[0035] (c) Hardware circuits and / or processor(s), such as microprocessor(s) or portions of microprocessor(s), that require software (e.g., firmware) to operate, but which may not be present when the software is not required for operation.

[0036] 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 only 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 a 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.

[0037] As used herein, the term "communication network" refers to a network that follows any suitable communication standard, such as 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), new radio (NR), etc. In addition, the communication between the terminal equipment and the network equipment in the communication network can be performed according to the communication protocol of any suitable generation, including but not limited to the first generation (1G), second generation (2G), 2.5G, 2.65G, third generation (3G), fourth generation (4G), 4.5G, future fifth generation (5G) communication protocol, and / or and / or any other protocol currently known or developed in the future. The embodiments of the present disclosure can be applied to various communication systems. In view of the rapid development of communication, there will certainly be communication technologies and systems that can embody future types of the present disclosure. It should not be considered that the scope of the present disclosure is limited to the above-mentioned system.

[0038] As used herein, the term "communication device" refers to a network device and / or a terminal device. The term "network device" refers to a node in a communication network, via which a terminal device accesses the network and receives services from it. A network device may refer to a base station (BS) or an access point (AP), such as a NodeB (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, a low-power node (such as a femto, a pico), etc., depending on the terminology and technology applied.

[0039] The term "terminal device" refers to any terminal device that can perform wireless communication. As an example and not limitation, the terminal device may also be referred to as a communication device, a user equipment (UE), a user station (SS), a portable user station, a mobile station (MS) or an access terminal (AT). The terminal device may include, but is not limited to, a mobile phone, a cellular phone, a smart phone, a voice over IP (VoIP) phone, a wireless local loop phone, a tablet computer, a wearable terminal device, a personal digital assistant (PDA), a portable computer, a desktop computer, an image acquisition terminal device such as a digital camera, a game terminal device, a music storage and playback device, a vehicle-mounted wireless terminal device, a wireless endpoint, a mobile station, a notebook computer embedded device (LEE), a notebook computer mounted device (LME), a USB dongle, a smart device, a wireless client device (CPE), an Internet of Things (IoT) device, a watch or other wearable device, a head-mounted display (HMD), a vehicle, a drone, medical equipment and applications (e.g., remote surgery), industrial equipment and applications (e.g., robots and / or other wireless devices operating in industrial and / or automated processing chain environments), consumer electronic devices, equipment 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.

[0040] As mentioned above, further research on MIMO technology is needed to improve system performance. In MIMO systems, beamforming is a crucial part that utilizes channel state information (CSI) to maximize channel capacity. The term "beamforming" used in this article refers to a technique that focuses a wireless signal to a specific receiving device, rather than letting the signal propagate in all directions from the broadcast antenna. Through the optimal beamforming algorithm, the best beam can be found and used to send the required data to the user equipment (UE).

[0041] If the base stations transmit signals in beams that interfere with each other, the UE will not receive the correct signal or suffer a degraded experience. According to conventional technology, for a single-user MIMO (SU-MIMO) system, Eigen beamforming (EBF) is an optimal beamforming method for transmitting data to a specific UE using (multiple) optimal beams.

[0042] For example, the channel matrix of a specific UE can be represented as H, which has a shape of M × N. It can then be decomposed using singular value decomposition (SVD) as follows.

[0043] H=UDV H (1)

[0044] The parameters U and V are unitary matrices, that is, U H U=I and V HV=I, parameter I is the identity matrix. Parameter D is a diagonal matrix whose diagonal elements are the singular values ​​of the channel matrix H.

[0045] Since it is customary to express Eigen-based beamforming in terms of Eigen values ​​rather than singular values, the parameter D can be written as The parameter λ i Indicates H H The eigenvalues ​​of H, parameter r is the rank of H. Without loss of generality, assume λ1>λ2>…>λ r .

[0046] Typically, the right singular vectors corresponding to the largest one or more eigenvalues ​​(hereinafter referred to as (multiple) eigenvectors) are used as beamforming vectors. Taking a common configuration as an example, it is assumed that only the eigenvector corresponding to the largest eigenvalue λ1 is used as the beamforming vector, that is, the first column of V (denoted as v1) is used as the beamforming vector.

[0047] In order to receive data correctly, the UE needs to combine the signals received at different antennas. Usually, the combining weights are calculated at the UE side based on channel estimation. Assuming that the pilot symbols are sent using the same beamforming as the data symbols, the effective channel estimated by the UE can be Assuming the simplest maximum ratio combining (MRC) receiver is used, for the current UE, the combining weights can be

[0048] The required data can be represented as s, and the received data can be represented as The noise can be expressed as n. The received data of the UE can be expressed as follows:

[0049]

[0050] However, for multi-user MIMO (MU-MIMO), EBF is no longer optimal. For the MU-MIMO case, there is often correlation between co-scheduled UEs in real-world systems, which is the cause of co-channel interference (CCI). EBF does not take these correlations into account, so it often suffers from performance degradation caused by CCI. Interference between beams of different UEs can significantly degrade the UE experience. According to conventional techniques, zero-forcing (ZF) beamforming is a classic method for MU-MIMO downlink systems. It aims to eliminate multi-user interference (MUI) between scheduled UEs.

[0051] When the UE has multiple antennas, block diagonal zero forcing (BD-ZF) beamforming should be used to obtain better performance. Assume that there are K UEs served by the same BS, is the packet channel matrix, where H k is the M of the kth UEk ×N channel matrix, M k is the number of antennas at the kth UE.

[0052] For a single-antenna UE, M k =1, (k=1, 2, ...K), ZF beamforming has the following simple form.

[0053] W=H H (HH H ) -1 (3)

[0054] The kth column of W (i.e., w k ) is the beamforming vector of the kth UE.

[0055] For multi-antenna UE, M k ≥2, (k=1, 2, ...K), the original ZF becomes suboptimal, so a new solution called BD-ZF is proposed. For BD-ZF, the beamforming vector w of the kth UE is k is chosen to satisfy the following formula.

[0056] in

[0057] Among them, H k is the M of the kth UE k ×N channel matrix.

[0058] There are many solutions to this optimization problem. According to the prior art, a solution is given as follows.

[0059]

[0060] in Only H is excluded k The extended channel matrix, H k is the M of the kth UE k ×N channel matrix, yes The pseudo-inverse of q k is with The right singular vector corresponding to the largest singular value of .

[0061] Generally, ZF algorithms, including BD-ZF, have dimensionality restrictions, i.e., the total number of all antennas on the UE side cannot be greater than the total number of antennas on the network device side. Otherwise, the ZF algorithm may not be able to alleviate CCI. In addition, the BD-ZF solution in equation (5) usually has high computational complexity, especially when many multi-antenna UEs are co-scheduled together. These two fatal shortcomings severely limit its application in actual MU-MIMO systems. BD-ZF is very sensitive to channel estimation errors and has a large amount of computation, so it is difficult to apply in actual systems.

[0062] According to other conventional techniques, it proposes a leakage-based beamforming method by maximizing the signal-to-noise ratio (SLNR) of all co-scheduled UEs. The SLNR of the kth UE is defined as follows:

[0063]

[0064] in Only H is excluded k The extended channel matrix, H k is the M of the kth UE k ×N channel matrix, is the noise n k The variance of , where the entries are assumed to be complex Gaussian variables with zero mean. Therefore, for the kth UE, the new matrix is ​​calculated as follows:

[0065]

[0066] The beamforming vector w of the kth UE is k will be the same as the matrix The eigenvector corresponding to the largest eigenvector of Only H is excluded k The extended channel matrix, H k It is the kth M k ×N channel matrix, is the noise n k The variance of , where each term is assumed to be a complex Gaussian variable with mean zero.

[0067] The max-SLNR algorithm is more flexible than the ZF algorithm and can work well even if the total number of all antennas on the UE side is greater than the total number of antennas on the network device side. However, it has a higher computational complexity due to the inverse matrix calculation of the large matrix. This fatal disadvantage limits its application in practical systems.

[0068] According to another conventional technique, it proposes a novel beamforming method called Eigen Zero Forcing (EZF).

[0069] The channel matrix of the kth UE can be expressed as Hk , which can be decomposed using SVD as follows.

[0070]

[0071] The right singular vector corresponding to the largest singular value (called the eigenvector v k1 ) is used as the initial beamforming vector, and for the kth UE, the Eigen receiver (i.e., U k The conjugate transpose of the first column of can be expressed as Therefore, based on the Eigen receiver, the effective channel between the base station antenna and the received signal after the receiver is That is, the Eigen channel of the kth UE.

[0072] Based on the Eigen channel of each UE, the extended Eigen channel matrix H eig It can be constructed as follows:

[0073]

[0074] Since the beamforming vectors are normalized to have unit norm, they can be omitted to simplify the calculation. That is, H eig =(v 11 , v 21 , …, v K1 ) H is also acceptable.

[0075] The beamforming matrix can be calculated as follows:

[0076]

[0077] The kth column of W (i.e., w k ) is the EZF beamforming vector of the kth UE.

[0078] When each UE utilizes more than one beam for data transmission, the Eigen channel matrix can be constructed from all active beams of the co-scheduled UEs. In this case, each beam can be regarded as a virtual UE, and for each virtual UE, the updated beamforming vector will be calculated by the EZF algorithm, which is then used for data transmission.

[0079] However, due to the correlation between the channel matrices of the co-scheduled UEs, the UE cannot obtain the best receiver based on Eigen. Generally, for UEs with low correlation, the UE can get an approximate Eigen receiver based on the channel matrix estimated at the UE side. But for UEs with high correlation, the receiver based on the channel matrix estimated by the UE may deviate greatly from the best receiver, and the residual interference will degrade the UE experience.

[0080] In addition, EZF has a dimensionality restriction, i.e., the total number of active beams of a UE should not be greater than the total number of antennas of the BS. This restriction is not as strict as the dimensionality restriction of ZF, but it still limits the application of EZF when more UEs are co-scheduled together.

[0081] Based on other conventional techniques, it proposes an iterative zero forcing precoding algorithm for MU-MIMO downlink system, called Classical Iterative Zero Forcing (CIZF), where a single data stream is assigned to each user. In this case, each UE has only one beamforming vector, making the problem much simpler.

[0082] For the kth UE, its channel matrix can be expressed as H k , its beamforming vector can be expressed as w k , whose receiving vector can be expressed as r k , the data required can be expressed as s k , then the global received signal at all UEs can be:

[0083]

[0084] in is the equivalent global channel matrix, W = (w1, w2, ..., w K ) is the global precoding matrix, s = (s1, s2, ..., s K ) T is a global transmission signal, It is the receiving noise.

[0085] Assume that all received vectors are known, i.e., r1, r2, ..., r K If it is known at the BS, the precoding matrix can be calculated using the ZF algorithm Based on this assumption, conventional techniques propose a joint optimization algorithm for alternately updating the receiving vector and the beamforming vector in an iterative mode, as shown in Table 1 below.

[0086] Table 1

[0087]

[0088] By alternately updating the receive and beamforming vectors of each UE, CIZF can obtain the optimal beam for each UE at the base station and the optimal receive vector at the UE side.

[0089] However, the classic iterative zero-forcing algorithm can be specialized for MU-MIMO downlink systems where a single data stream is allocated to each user. Therefore, its use in practical systems where each UE typically has multiple data streams is very limited.

[0090] For MU-MIMO downlink systems, CCI can be divided into two parts: one part can be sensed by the UE receiver and the other part cannot be sensed by the UE receiver. In other words, some CCI has no effect on the received signal, but most beamforming schemes aim to mitigate all interference from the base station's perspective. The EZF beamforming algorithm only aims to suppress the interference sensed by the best receiver. Therefore, it has much lower complexity than the original ZF beamforming algorithm.

[0091] However, the UE cannot obtain the best receiver. If the correlation between scheduled UEs is small, the ordinary receiver on the UE side (such as maximum ratio combining (MRC)) can suppress the residual interference to a certain extent. But for UEs with high correlation, the residual interference cannot be ignored and will degrade the performance.

[0092] According to an embodiment of the present disclosure, an iterative zero forcing (IZF) scheme is proposed. The embodiment of the present disclosure is combined with an actual receiver to further reduce the interference sensed by the terminal device. The beamforming scheme according to the embodiment of the present disclosure has many advantages. For example, the embodiment of the present disclosure solves interference from the perspective of the terminal device to reduce the interference that really affects the received signal of the terminal device. The embodiment of the present disclosure does not assume that the terminal device uses an Eigen receiver, but considers that the terminal device uses a more practical receiver to reduce interference. The embodiment of the present disclosure has high flexibility, and different interference elimination methods can be used to replace the ZF method.

[0093] Figure 1 A schematic diagram of a communication system 100 in which embodiments of the present disclosure may be implemented is illustrated. The communication system 100 as part of a communication network includes a device 110 (referred to as a "first device 110"). The communication system 100 also includes a device 120-1, a device 120-2, ..., a device 120-M (which may be collectively referred to as "(multiple) second devices 120"). M may be any suitable integer. It should be understood that Figure 1The number of devices shown is given for illustrative purposes and does not imply any limitation. It should also be noted that the first device 110 and the second device 120 can be interchanged. The device 120 can be a co-scheduled terminal device. The term "co-scheduled terminal device" as used herein refers to a terminal device that uses the same resources in the frequency domain and / or time domain.

[0094] The communication in the communication system 100 can be implemented according to any (multiple) appropriate communication protocols, including but not limited to cellular communication protocols of the first generation (1G), second generation (2G), third generation (3G), fourth generation (4G) and fifth generation (5G), 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. In addition, the communication can utilize any appropriate 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 duplexer (FDD), time division duplexer (TDD), multiple input multiple output (MIMO), orthogonal frequency division multiple access (OFDMA), and / or any other technology currently known or developed in the future.

[0095] Figure 2 FIG. 1 is a schematic diagram of a MIMO system according to an embodiment of the present disclosure. Figure 2 As shown, a first device 110 having N antennas can serve more than one second device 120. The kth device 120-k can have M k Antenna. Figure 2 As shown, the first device 110 may include input signals s1, s2, . . . , and s k The second device 120-1 can receive a signal from the first device 110 and obtain the signal The second device 120-2 may receive a signal from the first device 110 and obtain the signal Device 120-k may receive a signal from first device 110 and obtain the signal It should be noted that the parameter k can be any suitable number.

[0096] For purposes of illustration, the channel matrix of device 120-k may be represented as H k , and the beamforming weight of device 120-k may be expressed as w k The transmission property of the channel that can be estimated at device 120-k can be expressed as H k w k , H k w k is the signal sent kThe device 120-k may use the estimated effective channel H k w k To calculate the received signal

[0097] Consider the simplest receiver MRC, which only uses the conjugate transpose of the estimated effective channel, that is, Receiving Signals It can be expressed as:

[0098]

[0099] in represents a beamforming signal on the antenna of the first device 110, s represents a signal at the first device 110, Denotes the conjugate transpose of the estimated effective channel of device 120-k, H k represents the channel matrix of device 120-k, n k represents the noise on the channel between the device 120 - k and the first device 110 .

[0100] Receive signal n k It can be further written as the sum of three terms: desired signal, interference, and noise, as shown below.

[0101]

[0102] in Denotes the conjugate transpose of the estimated effective channel of device 120-k, H k represents the channel matrix of device 120-k, w k represents the beamforming weight of device 120 - k , s k Indicates the transmission signal. represents the interference between devices 120, n k represents the noise on the channel between the device 120 - k and the first device 110 .

[0103] Based on equation (13), the goal of maximizing channel gain and eliminating interference can be transformed as follows:

[0104] in

[0105] in Denotes the conjugate transpose of the estimated effective channel of device 120-k, H k represents the channel matrix of device 120-k, w k represents the beamforming weight of device 120-k, H k w i Indicates the effective interference channel.

[0106] The core idea of ​​equation (14) is that the effective channel H of device 120-k is k w k The “effective interference channel” H of the second device that should be co-scheduled k w i Orthogonal.

[0107] For simplicity, the covariance matrix of device 120-k can be expressed as

[0108] Therefore, formula (14) can be written as:

[0109] in

[0110] Where R k represents the covariance matrix of device 120-k, H k represents the channel matrix of device 120-k, w k represents the beamforming weight of device 120-k, H k w i Denotes the effective interference channel. In order to solve the optimization problem described in equation (15), the embodiments of the present disclosure aim to suppress the interference between effective channels and achieve better performance.

[0111] Figure 3 1 is a flow chart of a method 300 according to an embodiment of the present disclosure. The method 300 may be implemented at any suitable device. For example, the method may be implemented at the first device 110. It should be noted that the method 300 may also be implemented at the second device 120. For the purpose of illustration, reference is made to Figure 2 and Figure 3 Embodiments of the present disclosure are described in detail.

[0112] At block 310, the first device 110 is at least one device (e.g., device 120-k) of the plurality of co-scheduled devices 120 and determines one or more beamforming vectors. It should be noted that device 120-k is just an example. The one or more beamforming vectors of device 120-k may be represented as w k , such as the main eigenvector v k If the device 120-k has 2 or more beams, the beamforming vector may be represented as w k =(w k1 , w k2 , …), where w ki is the beamforming vector of the i-th beam of device 120-k. In some embodiments, the first device 110 may randomly determine one or more beamforming vectors.

[0113] Alternatively or additionally, the first device 110 may determine a covariance matrix R of the plurality of co-scheduled devices 120 k One or more eigenvectors of , and the one or more eigenvectors may be considered as one or more beamforming vectors. The one or more beamforming vectors may be optimal SU-MIMO beamforming vectors, which may be obtained by decomposing the covariance matrix R using an eigenvalue decomposition (EVD) algorithm k to calculate.

[0114] Assume that the covariance matrix R k has been decomposed by any EVD algorithm as shown below.

[0115]

[0116] Where V k is a unitary matrix, that is, Σ k =diag{λ k1 ,λ k2 , …, λ kr} is a diagonal matrix whose diagonal elements are R k eigenvalues ​​of (in descending order).

[0117] In addition, the eigenvector corresponding to the largest one or more eigenvalues ​​for data transmission of device 120-k may be represented as v k =(v k1 , v k2 , ...), then the optimal beamforming vector of device 120-k may be:

[0118] w k =v k (k=1,2,…,K) (17)

[0119] With such an Eigen initialization, the optional orthogonalization step in the generalized IZF algorithm can be omitted, and the IZF can generally converge faster. In general, 1 iteration may be sufficient for a full-rank terminal device, and 3 iterations may be sufficient for a non-full-rank terminal device. The term "full-rank terminal device" as used herein refers to a terminal device that employs the same number of beams as its receive antennas, and the term "non-full-rank terminal device" as used herein refers to a terminal device that employs fewer beams than its receive antennas.

[0120] The first device 110 iteratively updates one or more beamforming vectors by performing the following operations. At block 320, the first device 110 determines a transmission property of a channel between the first device 110 and the device 120. For example, an effective channel matrix of the device 120 may be represented as follows.

[0121] H ef =(R1w1, R2w2, ..., R K w K ) H (18)

[0122] Among them, H ef represents the effective channel matrix, R k represents the covariance matrix of device 120-k, w k represents the beamforming vector(s) of device 120 - k .

[0123] At block 330, the first device 110 performs precoding on the transmission properties of the channel to mitigate interference between the plurality of co-scheduled devices 120. In some embodiments, the first device 110 may perform an interference cancellation operation on the one or more beamforming vectors to update the one or more beamforming vectors. For example, the first device 110 may perform a zero forcing operation. It should be noted that the first device may perform any suitable interference cancellation operation.

[0124] In some embodiments, the updated beamforming matrix may be expressed as follows:

[0125]

[0126] The kth column of W (i.e., w k ) is the updated beamforming vector(s) of device 120-k, H ef represents the effective channel matrix.

[0127] In some embodiments, the first device 110 may perform IZF on the transmission properties of the channel. The IZF process may be viewed as a power iteration (PI) with zero interference constraints for the co-scheduled device 120. If equation (19) is modified to Then it becomes the classic power iteration algorithm. The right-hand side of equation (19) The separated beamforming vectors are projected to the nulling space of the co-scheduled devices 120. Therefore, IZF can be viewed as a generalization of PI for the MU-MIMO case, while classic PI is only dedicated to the SU-MIMO case.

[0128] In other words, the classic PI algorithm tries to find the best beam for each UE without considering the co-scheduled terminal devices, and it is an implementation of the EBF method, which calculates the best selfish beam for each terminal device. The embodiments of the present disclosure take into account the co-scheduled terminal devices and determine the best selfless beam for each terminal device. More importantly, since the channel conditions of some cell edge terminal devices may be poor, ZF operation may not be the right approach. Since the generalized IZF is very flexible, it can be used to The corresponding columns in are set to (0, ..., 1, ..., 0) T , so that other co-scheduled end devices do not interfere with this special end device. But this special end device may still use its Eigen beam to maximize its performance.

[0129] As described above, device 120-k may have more than one beamforming vector. In this case, device 120-k may make two beams orthogonal to each other, i.e., for device 120-k, Additionally, as discussed with reference to block 310, one or more beamforming vectors may be a covariance matrix R k In this case, the orthogonalization step may be omitted even if device 120-k has more than one beamforming vector.

[0130] In other embodiments, the first device 110 may determine a regularization parameter based on a signal-to-noise ratio of the plurality of co-scheduled devices 120. In this way, in some cases, the IZF may be more robust and achieve better performance. For example, the first device 110 may determine the number of beams of the plurality of co-scheduled devices 120. If the number of beams of the plurality of co-scheduled devices 120 is greater than the number of antennas of the first device 110, the EZF may not eliminate interference and may suffer performance degradation.

[0131] The regularization parameter can be added to the interference cancellation operation during the update of the beamforming matrix. For example, equation (19) can be rewritten as follows:

[0132]

[0133] Where I represents the identity matrix and β represents the regularization parameter. Generally, β is determined by the signal-to-noise ratio (SNR) of the UE. Assuming that all co-scheduled devices 120 have the same SNRρ, β can be set to the inverse of ρ, that is, β=1 / ρ.

[0134] Alternatively, different terminal devices may often have different SNRs. Assume that the noise n received at device 120-k is k The entries have zero mean and variance complex Gaussian variable, then the SNR of device 120-k can be defined as Therefore, equation (19) can be replaced by the following regularized version:

[0135]

[0136] in is the regularization term, M k represents the number of antennas of device 120-k, I kis the identity matrix whose size is determined by the number of active beams of device 120-k. For example, if 2 beams are used for data transmission, then I k will be the 2×2 identity matrix.

[0137] In this way, this regularized version of IZF, called iterative regularized zero forcing (IRZF) beamforming, has no size limit. It has a wider range of applications than general ZF and EZF algorithms. For example, when the total number of all antennas on the terminal device side is greater than the total number of antennas on the network device side, the general ZF algorithm cannot work, but EZF, IZF, and IRZF can work well. When the total number of all beams on the terminal device side is greater than the total number of antennas on the network device side, the general ZF, EZF, and IZF can no longer work, but IRZF can still work normally.

[0138] Similar to IZF, using eigenvectors as initial beamforming vectors is still the best choice for IRZF. Similarly, for convenience, IRZF with eigenvectors as initial beamforming vectors can be called iterative Eigen regularized zero forcing (IERZF).

[0139] One or more beamforming vectors are iteratively updated until the end condition is met. Figure 3 At block 340, the first device 110 determines whether an end condition is satisfied. The end condition may be a threshold number of iterations. For example, one or more beamforming vectors are iteratively updated until the number of iterations reaches a threshold number.

[0140] Alternatively or additionally, the termination condition may be an interference threshold. For example, one or more beamforming vectors are iteratively updated until the interference is below a threshold. The generalized IZF may have an explicit termination condition because It only represents the interference between co-scheduled UEs. The matrix The item in the i-th row and j-th column is represented by a ij , indicating that the function off(A)=∑ i≠j |a ij | 2 , which summarizes all off-diagonal entries in A. Therefore, when off(A) is zero, IZF converges to the optimal solution. In general, the iteration terminates when off(A) is not greater than the interference threshold ε.

[0141] If the end condition is satisfied, then the updated one or more beamforming vectors may be output. For example, the first device 110 may output a column of W, i.e., w k (k=1, 2, . . . , K), as the final beamforming vector(s) of the co-scheduled device 120.

[0142] In block 350, the first device 110 may perform beamforming on the multiple co-scheduled devices 120 based on the iteratively updated one or more beamforming vectors. Compared with conventional techniques, the embodiments of the present disclosure have many advantages: low complexity: the computational complexity of IZF is much lower than that of BD-ZF and max-SLNR algorithms; wide application range: the application range of IRZF is the same as that of max-SLNR, and IRZF can be applied even if the total number of antennas / beams of the terminal device is greater than the number of antennas at the base station, while BD-ZF cannot be used in this case; better performance; and high flexibility.

[0143] In addition, the embodiments of the present disclosure are less complex than conventional techniques. For simplicity, it is assumed that K terminal devices are paired and each terminal device has antennas, and the network device has N antennas for one polarization. For the BD-ZF, max-SLNR, EZF, and IEZF algorithms, they all need to calculate the covariance matrix of the same size and the (multiple) eigenvectors of these covariance matrices. Apart from this, the main difference between these algorithms lies in the calculation of the inverse matrix.

[0144] For the BD-ZF algorithm, it is necessary to calculate K times of size Therefore, in addition to the EVD calculation, the main computational complexity of BD-ZF is For the max-SLNR algorithm, it is necessary to calculate K inverse matrices of size N×N. Therefore, in addition to the EVD calculation, the main computational complexity of max-SLNR is O(KN 3 ). Since BD-ZF can only be used So the complexity of BD-ZF is slightly lower than that of max-SLNR.

[0145] For each iteration of EZF and IEZF, only K b ×K b The inverse matrix of K b is the total number of beams used for data transmission, generally Therefore, except for the EVD calculation, the main computational complexity of EZF is O(K b 3 ), and the main computational complexity of IEZF is O(iterNum×K b 3 ).

[0146] Taking a common situation in an actual system as an example, the network device has 64 antennas (assuming 32 antennas are used for one polarization, that is, N = 32), except for the EVD calculation, the main complexity of these beamforming algorithms can be calculated as shown in Table 2. It can be seen that the complexity of EZF and IEZF is much lower than that of BD-ZF and max-SLNR algorithms.

[0147] Table 2

[0148]

[0149]

[0150] For the iterative zero-forcing scheme, the main complexity comes from two parts: one is the eigenvector calculation or EVD part, and the other is the pseudo-inverse or ZF part. The dimension of the covariance matrix determines the computational complexity of the first part, while the total number of beams of the co-scheduled terminal devices determines the computational complexity of the second part.

[0151] In general, a network device may have up to 64 transceivers (TRX) for downlink transmission, and the EVD of a 64×64 covariance matrix should be calculated to obtain the best Eigen beam. Assuming that power iteration is used for EVD calculation, the complexity of each eigenvector will be O(64 2 ) level. Assuming K co-scheduled rank 2 terminal devices are served by such a network device, there are 2K beams in total and the complexity of the pseudo-inverse will be O((2K) 3 )level.

[0152] In order to reduce the overall complexity and facilitate the application of IZF algorithms including the special case EZF to actual products, some conventional techniques can be used. For example, polarization averaging can be used to reduce the dimension of the covariance matrix. Usually, 64 TRXs are laid out in a 4×8×2 pattern, such as Figure 4 As shown, there are 16 TRXs in the horizontal direction (8 TRXs per polarization) and 8 TRXs in the vertical direction (4 TRXs per polarization). For the ordering of the N TRXs (i.e., the columns of H in equation (1)), the array column index changes fastest, followed by the array row index, and finally the polarization index. Therefore, all elements of the same polarization are indexed consecutively, and within each polarization, all elements in the same array row are indexed consecutively. Since there are a total of 32 TRXs per polarization and 2 polarizations can use the same beam weights for downlink transmission, this feature can be used to reduce the dimensionality of the covariance matrix and the total number of beams for co-scheduled terminal devices.

[0153] Specifically, when polarization averaging is used, the EVD of the 32×32 covariance matrix is ​​calculated instead, which can reduce the computational complexity of each eigenvector to O(322 ) level. Under the same assumption that the network device serves K co-scheduled rank-2 terminal devices, there are a total of K beams, and the complexity of the pseudo-inverse will be reduced to O(K 3 )level.

[0154] In addition, the computational complexity of the EVD of the covariance matrix can be further reduced by adopting the Kronecker structure, which decomposes the full-dimensional EVD of the 32×32 covariance matrix into two independent EVDs of the 4×4 vertical covariance matrix and the 8×8 horizontal covariance matrix. And by calculating the Kronecker product of the eigenvector of the 4×4 covariance matrix and the eigenvector of the 8×8 covariance matrix, the eigenvector of the 32×32 covariance matrix can be approximately recovered. Since the complexity of calculating the eigenvectors of the 4×4 and 8×8 covariance matrices is much lower than that of calculating the eigenvector of the 32×32 covariance matrix, adopting this approximation can greatly reduce the computational complexity and promote the application of EBF, EZF and IEZF algorithms when computing resources are quite limited.

[0155] Considering the actual TDD 5G system, the network equipment uses N TRXs (B TRXs per polarization) to serve K co-scheduled terminal devices in a cell with the same time / frequency resources. row ×N col ×N pol Pattern laying, for example, 64 TRX can be laid out in 4×8×2 pattern, such as Figure 4 As shown. On the terminal device side, since the actual terminal device may have fewer transmit chains than receive chains, it is assumed that each terminal device has R receive chains and T transmit chains (T≤R), and is served by L (L≤min(R, N)) layers.

[0156] In a TDD system, the downlink channel matrix can be estimated from the sounding reference signal (SRS) on the uplink channel by exploiting channel reciprocity. The estimated channel of device 120-k in time slot t and PRB / RBG f can be expressed as H k (f, t), whose size is T×N. The matrix H k Each row of (f, t) corresponds to a UE transmit chain. For a 4×8×2 array, the matrix H k The correspondence between the column index of (f, t) and the physical TRX array position in terms of array row, array column and array polarization is as follows: Figure 5 shown.

[0157] In general, generalized IZF beamforming can be applied to broadband or subband modes. Assume that P subbands span the entire system bandwidth and denote the set of physical resource blocks (PRBs) (or RBGs) in the sth subband as the set in Indicates the total number of PRBs (or RBGs) in a subband.

[0158] Fig. 6A illustrates the implementation of EZF, and Figure 6B The figure shows the implementation of HVEZF. Fig. 6A As shown, the covariance matrix (CoMa) module 610 can obtain the covariance matrix of the device 120-k on the subband s in the time slot t, which is calculated as EVD module 620 may obtain the conjugate transpose of the effective eigenchannel matrix of device 120-k, which is denoted as E k (s, t) = (v k1 , …, v kL ), where v kl YesR k The precoding weight module 630 can obtain the beamforming vectors for all co-scheduled devices 120, which can be obtained by W(s, t) = E(s, t) (E H (s, t)E(s, t)+βI) -1 To calculate, where E(s, t) = (E1(s, t), E2(s, t), ..., E K (s, t)) is the extended eigenchannel matrix of size N×KL for K co-scheduled UEs (compare with equation (9)), and β is the (SNR-dependent) regularization parameter.

[0159] When polarization averaging and horizontal / vertical Kronecker decomposition are utilized, the process is very different from the EZF process. Figure 6B As shown, the so-called HVEZF (Horizontal and Vertical EZF) is a simplified version of the EZF that employs polarization averaging and a Kronecker structure.

[0160] In the HVEZF process, the two polarizations use the same beamforming vector, and the CoMa module 640 can obtain the characteristic channel matrix E of the device 120-k. k (s, t), the matrix becomes an N / 2×L / 2 matrix and is only one quarter of the Eigen channel matrix in the EZF process. Similar to the EZF process, the EVD module 650 can obtain E(s, t)=(E1(s, t), E2(s, t), ..., E K (s, t)), which is an extended characteristic channel matrix of size N / 2×KL / 2 of the co-scheduled device 120. The precoding weight module 660 can obtain the final full-dimensional beamforming vector by:

[0161]

[0162] in is the tensor product (Kronecker) of matrices. Note that the 2×2 identity matrix is not the only option. If some cross-polarization information is available (e.g., via information reported from a terminal device), then can be replaced by another matrix containing the cross-polarization information.

[0163] In the HVEZF process, a tensor product (Kronecker) structure can be introduced to calculate the Eigen channel matrix E of the device 120-k k (s, t) to further reduce the computational complexity. When the tensor product (Kronecker) structure is used, the horizontal and vertical covariance matrices ( and ), instead of calculating the full-dimensional covariance matrix R k (s, t). Based on the above definition, the horizontal covariance matrix can be calculated as follows:

[0164]

[0165] in (·) T and(·) * denote the transpose and conjugate of a matrix, respectively.

[0166] Similarly, the vertical covariance matrix can be calculated as follows:

[0167]

[0168] in It should be noted that both polarizations utilize the same beamforming vector in the HVEZF, which is why the horizontal and vertical covariance matrices are averaged between polarizations.

[0169] After obtaining the horizontal and vertical covariance matrices, the Eigen pair of the two covariance matrices can be calculated by the EVD algorithm. For the horizontal covariance matrix Its eigenvalues ​​and eigenvectors are expressed as and For the vertical covariance matrix Its eigenvalues ​​and eigenvectors are expressed as and therefore, The first eigenvector of Calculation, the second eigenvector can be calculated by the following formula:

[0170]

[0171] Using these eigenvectors, the effective Eigen channel matrix of device 120-k can be obtained by E k (s, t) = (v k1 , …, v kL / 2 ) is obtained. It should be noted that due to the use of polarization averaging, E k The size of (s, t) is N / 2×L / 2.

[0172] 7A to 12B The simulation performance of the embodiment of the present disclosure and the conventional technology is illustrated.

[0173] like 7A to 12B As shown, the embodiments of the present disclosure are compared with EBF, BD-ZF, max-SLNR and EZF schemes. The simulation parameters are as follows: modulation order: 16-QAM constellation; number of antennas at the network device: N = 32 (32 transmit antennas); number of co-scheduled terminal devices: K = 8 or K = 12 (each UE uses 1 to 2 beams for transmission); number of antennas at the terminal device: or (2 or 4 receive antennas); minimum mean square error with interference rejection combining (MMSE-IRC) receiver is adopted.

[0174] In the simulation, it is assumed that H k The elements of (i.e., the channel matrix of the k-th terminal device) are complex Gaussian variables with zero mean and unit variance. In addition, it is assumed that the additive noise vector n k has mean zero and variance The received SNRρ k Defined as

[0175] The maximum orthogonality principle is used for MU pairing. Eight or 12 terminal devices with the smallest correlation are selected from the 16 candidate terminal devices as the co-scheduled terminal devices to maximize the orthogonality between the co-scheduled terminal devices. The results are averaged over 3,000 terminal device group samples.

[0176] Fig. 7A , Figure 7B , Fig. 8A , Figure 8B , Fig. 9A and Fig. 9B The figure shows the simulation results for the low correlation case. In this case, the correlation between the co-scheduled terminal devices is low. During transmission, each terminal device uses 1 to 2 main beams for transmission, and the other beams are not used. The symbol error rate (SER) performance of different beamformers is shown in Figure 1. Fig. 7A and Figure 7B shown.

[0177] exist Fig. 7A and Figure 7B In the example, there are 8 or 12 co-scheduled terminal devices, each of which has 2 antennas and uses 1 main beam for data transmission. Fig. 7A As shown in Figure 2, in the case of 8 co-scheduled terminal devices, IEZF and EZF have the best performance, while BD-ZF and max-SLNR have the second best performance. Selfish EBF has the worst performance.

[0178] Figure 7B 12 co-scheduled terminal devices are shown in FIG. 1 . IEZF achieves the best performance, slightly better than EZF. The performance of max-SLNR is worse than that of EZF, but slightly better than that of BD-ZF. EBF still achieves the worst performance.

[0179] exist Fig. 8A and Figure 8B In the example, each terminal device has 4 antennas and uses 2 main beams for data transmission. Fig. 8A The figure shows the situation of 8 co-scheduled terminal devices, among which IEZF still obtains the best performance, EZF obtains the second best performance. Max-SLNR obtains the third best performance, slightly better than BD-ZF. Selfish EBF still has the worst performance.

[0180] Figure 8B The figure shows the situation of 12 co-scheduled terminal devices, where IEZF still obtains the best performance. EZF obtains the second best performance, and max-SLNR obtains the third best performance. Due to size limitations, BD-ZF has the worst performance, even worse than SU-MIMO EBF.

[0181] exist Fig. 9A and Fig. 9B In , the throughput performance of different beamforming methods based on MMSE-IRC receiver is plotted. Fig. 9A As shown in Figure 1, each terminal device has 2 antennas and uses 1 main beam for data transmission. IEZF and EZF achieve the highest throughput, while max-SLNR and BD-ZF achieve the second highest throughput. EBF is the best beamformer in the SU-MIMO case, and its throughput is the lowest.

[0182] like Fig. 9BAs shown in the figure, each terminal device has 4 antennas and uses 2 main beams for data transmission. IEZF obtains the highest throughput, EZF obtains the second highest throughput, and max-SLNR obtains the third highest throughput. The throughput of BD-ZF is lower than that of EBF at low SNR, but higher than that of EBF at high SNR. These results further verify the SER performance of the above different beamformers.

[0183] Fig. 10A , Fig. 10B , Fig.11A , Fig. 11B , Fig. 12A and Fig. 12B The figure shows the simulation results for the high correlation case. In this case, the correlation between the co-scheduled terminal devices is high, which is an approximation of the scenario where the terminal devices are very close. During transmission, each terminal device uses 1 to 2 main beams for transmission, and the other beams are not used. The SER performance of different beamformers is shown in Figure 1. Figures 10A-11B shown.

[0184] exist Fig. 10A and Fig. 10B In the example, there are 8 or 12 highly correlated co-scheduled terminal devices, each of which has 2 antennas and uses 1 main beam for data transmission. Fig. 10A The figure shows the case of 8 co-scheduled terminal devices, where IEZF achieves the best performance, far better than the performance of other beamformers. EZF achieves the second best performance, while max-SLNR and BD-ZF achieve the third best performance. EBF is the best beamformer in the SU-MIMO case, and its performance is the worst.

[0185] Fig. 10B The case of 12 co-scheduled terminal devices is shown in FIG, where IEZF obtains the best SER performance and EZF obtains the second best SER performance. The performance of Max-SLNR and BD-ZF is similar, which is worse than EZF but much better than EBF.

[0186] exist Fig.11A and Fig. 11B In the example, there are 8 or 12 highly correlated co-scheduled terminal devices, each with 4 antennas and using 2 main beams for data transmission. Fig.11A The case of 8 co-scheduled terminal devices is shown in Figure 1, where IEZF achieves the best performance. EZF achieves the second best performance, and max-SLNR achieves the third best performance. The performance of BD-ZF is slightly worse than max-SLNR, but much better than the SU-MIMO EBF beamformer.

[0187] Fig. 11BThe case of 12 co-scheduled terminal devices is shown in FIG, where IEZF still obtains the best SER performance. EZF obtains the second best performance, and max-SLNR obtains the third best performance. Due to size limitations, BD-ZF fails to obtain better performance than EBF, and both obtain the worst SER performance.

[0188] exist Fig. 12A and Fig. 12B In , the throughput performance of different beamforming methods based on MMSE-IRC receiver is plotted. Fig. 12A As shown in Figure 1, where each terminal device has 2 antennas and transmits using 1 main beam, IEZF achieves the highest throughput, EZF achieves the second highest throughput, Max-SLNR achieves the third highest throughput, slightly better than BD-ZF, and EBF achieves the lowest throughput.

[0189] like Fig. 12B As shown in Figure 1, where each terminal device has 4 antennas and transmits using 2 main beams, IEZF obtains the highest throughput, EZF obtains the second highest throughput, and max-SLNR obtains the third highest throughput. BD-ZF has a lower throughput than EBF at low SNR, but a higher throughput than EBF at high SNR.

[0190] In some embodiments, an apparatus for performing method 300 (e.g., first device 110) may include corresponding components for performing corresponding steps in method 300. These components may be implemented in any suitable manner. For example, it may be implemented by a circuit system or a software module.

[0191] In some embodiments, the apparatus includes a component for determining, at a first device, one or more beamforming vectors for one or more devices among a plurality of co-scheduled devices; a component for iteratively updating the one or more beamforming vectors by: determining transmission properties of a channel based on the one or more beamforming vectors and a channel estimate of the channel between the first device and the plurality of co-scheduled devices; performing precoding on the transmission properties of the channel to mitigate interference between the plurality of co-scheduled devices; and a component for performing beamforming on the plurality of co-scheduled devices based on the iteratively updated one or more beamforming vectors.

[0192] In some embodiments, means for determining one or more beamforming vectors comprises means for determining one or more eigenvectors of at least one covariance matrix of a plurality of co-scheduled devices.

[0193] In some embodiments, a component for determining one or more eigenvectors of a device among one or more devices of a plurality of co-scheduled devices includes: a component for determining a horizontal covariance matrix of the device; a component for determining a vertical covariance matrix of the device; a component for determining at least one eigenvector of a horizontal covariance matrix of at least one device; and a component for determining at least one eigenvector of a vertical covariance matrix.

[0194] In some embodiments, the apparatus includes means for recovering one or more beamforming vectors of the device based on at least one eigenvector of the horizontal covariance matrix and at least one eigenvector of the vertical covariance matrix.

[0195] In some embodiments, means for updating one or more beamforming vectors includes means for determining a regularization parameter based on a signal-to-noise ratio of the plurality of co-scheduled devices; and means for performing precoding of the transmission property and the regularization parameter.

[0196] In some embodiments, the means for performing precoding the transmission properties of the channel includes means for performing an interference cancellation operation on the one or more beamforming vectors to update the one or more beamforming vectors.

[0197] In some embodiments, one or more beamforming vectors are iteratively updated until the number of iterations reaches a threshold number.

[0198] In some embodiments, one or more beamforming vectors are iteratively updated until the interference is below a threshold.

[0199] In some embodiments, the apparatus further comprises means for orthogonalizing the updated one or more beamforming vectors of a device among the one or more devices in response to determining that the number of beamforming vectors of the device exceeds one.

[0200] In some embodiments, the first device is a network device and at least one device is a terminal device.

[0201] Fig.13 1 is a simplified block diagram of a device 1300 suitable for implementing embodiments of the present disclosure. The device 1300 may be provided to implement a communication device, such as Figure 1 The first device 110, the second device 120 or the vehicle is shown. As shown in the figure, the device 1300 includes one or more processors 1310, one or more memories 1320 coupled to the processor 1310, and one or more communication modules 1340 coupled to the processor 1310.

[0202] The communication module 1340 is used for two-way communication. The communication module 1340 has at least one antenna to facilitate communication. The communication interface may represent any interface necessary to communicate with other network elements.

[0203] Processor 1310 may be of any type suitable for the local technology network, and may include, as non-limiting examples, one or more of: 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 1300 may have multiple processors, such as application specific integrated circuit chips that are time slaved to a clock synchronized with a main processor.

[0204] The memory 1320 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) 1324, electrically programmable read-only memory (EPROM), flash memory, hard disk, compact disk (CD), digital video disk (DVD), and other magnetic storage and / or optical storage. Examples of volatile memories include, but are not limited to, random access memory (RAM) 1322 and other volatile memories that do not persist during power outages.

[0205] Computer program 1330 includes computer executable instructions executed by associated processor 1310. Program 1330 may be stored in ROM 1324. Processor 1310 may perform any suitable actions and processes by loading program 1330 into RAM 1322.

[0206] The embodiments of the present disclosure can be implemented with the aid of program 1320, so that device 1300 can execute the following steps: Figure 3 Any process of the present disclosure discussed. The embodiments of the present disclosure may also be implemented by hardware or a combination of software and hardware.

[0207] In some example embodiments, program 1330 may be tangibly embodied in a computer-readable medium that may be included in device 1300 (such as in memory 1320) or in other storage devices accessible to device 1300. Device 1300 may load program 1330 from the computer-readable medium to RAM 1322 for execution. Computer-readable media may include any type of tangible non-volatile memory, such as ROM, EPROM, flash memory, hard disk, CD, DVD, etc. Fig.14 An example of computer readable medium 1400 is shown in the form of a CD or DVD. The computer readable medium has program 1330 stored thereon.

[0208] It should be understood that future networks may utilize network function virtualization (NFV), which is a network architecture concept that proposes virtualizing network node functions into "building blocks" or entities that can be operationally connected or linked together to provide services. A virtualized network function (VNF) may include one or more virtual machines that use standard or general-purpose servers rather than customized hardware to run computer program code. Cloud computing or data storage may also be used. In radio communications, this may mean that node operations are at least partially performed in a central / centralized unit CU (e.g., a server, host, or node) that is operationally coupled to a distributed unit DU (e.g., a radio head / node). Node operations may also be distributed among multiple servers, nodes, or hosts. It should also be understood that the distribution of work between core network operations and base station operations may vary depending on the implementation.

[0209] In one embodiment, the server can generate a virtual network through which the server communicates with the distributed unit. In general, a virtual network can involve the process of combining hardware and software network resources and network functions into a single software-based management entity (virtual network). Such a virtual network can provide a flexible distribution of operations between the server and the radio head / node. In practice, any digital signal processing task can be performed in the CU or DU, and the boundary of the transfer of responsibilities between the CU and the DU can be selected according to the implementation.

[0210] Thus, in one embodiment, a CU-DU architecture is implemented. In this case, the device 1300 may be included in a central unit (e.g., a control unit, an edge cloud server, a server) that is operably coupled (e.g., via a wireless or wired network) to a distributed unit (e.g., a remote radio head / node). That is, the central unit (e.g., an edge cloud server) and the distributed units may be independent devices that communicate with each other via a radio path or via a wired connection. Alternatively, they may be in the same entity that communicates via a wired connection, etc. The edge cloud or edge cloud server may serve multiple distributed units or radio access networks. In one embodiment, at least some of the described processes may be performed by the central unit. In another embodiment, the device 1300 may instead be included in a distributed unit, and at least some of the described processes may be performed by the distributed unit.

[0211] In one embodiment, the execution of at least some functions of the device 1300 can be shared between two physically separate devices (DU and CU) that form an operational entity. Therefore, the device can be regarded as depicting an operational entity including one or more physically separate devices for performing at least some of the processes described. In one embodiment, such a CU-DU architecture can provide a flexible distribution of operations between the CU and the DU. In practice, any digital signal processing task can be performed in the CU or the DU, and the boundary of the transfer of responsibilities between the CU and the DU can be selected according to the implementation. In one embodiment, the device 1300 controls the execution of the process regardless of the location of the device and where the process / function is executed.

[0212] Generally, 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 controllers or other computing devices, or some combination thereof.

[0213] 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 real or virtual processor to perform the above referenced Figure 3 Method 300 described. Typically, 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 functions of program modules can be combined or split between program modules as needed. The machine executable instructions of program modules can be executed in local or distributed devices. In distributed devices, program modules can be located in both local and remote storage media.

[0214] The program code for executing the method of the present disclosure 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 completely on the machine, partially on the machine, as an independent software package, partially on the machine and partially on a remote machine, or completely on a remote machine or server.

[0215] 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.

[0216] The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. A more specific example of a computer readable storage medium will include an electrical connection with one or more wires, a portable computer floppy disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0217] In addition, although the 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 order or performing all the operations shown to obtain the desired result. In some cases, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the above discussion, these should not be interpreted as limitations on the scope of the present disclosure, but rather 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.

[0218] Although the present disclosure has been described in language specific to structural features and / or methodological acts, it should be understood that the present disclosure defined in the appended claims is not necessarily limited to the specific features or acts described above. Instead, 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: determining one or more beamforming vectors for one or more devices of a plurality of co-scheduled devices; The one or more beamforming vectors are iteratively updated by: determining a transmission property of a channel between the first device and the plurality of co-scheduled devices based on the one or more beamforming vectors and a channel estimate of the channel; as well as performing precoding on the transmission properties of the channel to mitigate interference between the plurality of co-scheduled devices; as well as Based on the iteratively updated one or more beamforming vectors, beamforming is performed for the plurality of co-scheduled devices.

2. The first device of claim 1 , wherein the first device is configured to determine the one or more beamforming vectors by: One or more eigenvectors of at least one covariance matrix of the plurality of co-scheduled devices are determined.

3. The first device of claim 2, wherein the first device is caused to determine the one or more feature vectors for one of the one or more devices by: Sure a horizontal covariance matrix of the device; determining a vertical covariance matrix for the device; determining at least one eigenvector of the horizontal covariance matrix; as well as At least one eigenvector of the vertical covariance matrix is ​​determined.

4. The first device according to claim 3, wherein the first device is further configured to: The one or more beamforming vectors for the device are recovered based on the at least one eigenvector of the horizontal covariance matrix and the at least one eigenvector of the vertical covariance matrix.

5. The first device of claim 1 , wherein the first device is caused to update the one or more beamforming vectors by: determining a regularization parameter based on a signal-to-noise ratio of the plurality of co-scheduled devices; and The precoding is performed on the transmission properties and the regularization parameters.

6. The first device of claim 1 , wherein the first device is caused to perform precoding on the transmission property of a channel by: An interference cancellation operation is performed on the one or more beamforming vectors to update the one or more beamforming vectors.

7. The first device of claim 1, wherein the one or more beamforming vectors are iteratively updated until a number of iterations reaches a threshold number.

8. The first device of claim 1, wherein the one or more beamforming vectors are iteratively updated until the interference is below a threshold.

9. The first device according to claim 1, wherein the first device is further caused to: Based on determining that the number of the beamforming vectors for a device among the one or more devices exceeds one, orthogonalizing the updated one or more beamforming vectors for the device.

10. The first device according to any one of claims 1 to 9, wherein the first device is a network device, and at least one of the one or more devices is a terminal device.

11. A method for downlink beamforming in a multi-user multiple-input multiple-output (MU-MIMO) system, comprising: determining, at a first device, one or more beamforming vectors for one or more devices of a plurality of co-scheduled devices; The one or more beamforming vectors are iteratively updated by: determining a transmission property of a channel between the first device and the plurality of co-scheduled devices based on the one or more beamforming vectors and a channel estimate of the channel; as well as performing precoding on the transmission properties of the channel to mitigate interference between the plurality of co-scheduled devices; as well as Based on the iteratively updated one or more beamforming vectors, beamforming is performed for the plurality of co-scheduled devices.

12. The method of claim 11, wherein determining the one or more beamforming vectors comprises: One or more eigenvectors of at least one covariance matrix of the plurality of co-scheduled devices are determined.

13. The method of claim 12, wherein determining the one or more feature vectors for a device of the one or more devices comprises: determining a horizontal covariance matrix for the device; determining a vertical covariance matrix for the device; determining at least one eigenvector of the horizontal covariance matrix; as well as At least one eigenvector of the vertical covariance matrix is ​​determined.

14. The method according to claim 13, further comprising: The one or more beamforming vectors for the device are recovered based on the at least one eigenvector of the horizontal covariance matrix and the at least one eigenvector of the vertical covariance matrix.

15. The method of claim 11, wherein updating the one or more beamforming vectors comprises: determining a regularization parameter based on a signal-to-noise ratio of the plurality of co-scheduled devices; as well as The precoding is performed on the transmission properties and the regularization parameters.

16. The method of claim 11, wherein performing precoding on the transmission property of a channel comprises: An interference cancellation operation is performed on the one or more beamforming vectors to update the one or more beamforming vectors.

17. The method of claim 11, wherein the one or more beamforming vectors are iteratively updated until the number of iterations reaches a threshold number.

18. The method of claim 11, wherein the one or more beamforming vectors are iteratively updated until the interference is below a threshold.

19. The method according to claim 11, further comprising: Based on determining that the number of the beamforming vectors for a device among the one or more devices exceeds one, orthogonalizing the updated one or more beamforming vectors for the device.

20. The method according to any one of claims 11 to 19, wherein the first device is a network device, and at least one of the one or more devices is a terminal device.

21. A computer-readable storage medium comprising program instructions stored thereon, which instructions, when executed by a device, cause the device to perform the method according to any one of claims 11 to 20.

22. An apparatus comprising means for performing the method according to any one of claims 11 to 20.

Images