A method of compressing a plurality of digital signals, computer program product, nontransitory computer-readable storage medium, processing unit, wireless device, and chips therefor

Abstract

A method (100) of compressing, by a receiving device, a first plurality of digital signals obtained from a first plurality (NRX) of frequency-division multiplexed, FDM, signals, received by the receiving device, the method comprising: obtaining (110) two or more channel estimate matrices (H1, H2, …, HK) associated with propagation channels for the FDM signals; applying (120) a function (F) to the two or more channel estimate matrices to obtain a resulting matrix (RM), the resulting matrix (RM) resulting from the applying (120); decomposing (130) the resulting matrix (RM) into a first decomposition matrix (U), comprising first vectors of coefficients (U1, U2, …, UN), and a second decomposition matrix (Λ), different from the first decomposition matrix (U), comprising second vectors of coefficients (Λ1, Λ2, …, ΛN), wherein the first decomposition matrix (U) is a unitary eigenvector matrix comprising one or more eigenvectors; from the first vectors of coefficients (U1, U2, …, UN), determining (140) vectors of spatial filter coefficients (T1, T2, …, TN); selecting (145) a subset (T1, …, TM) of the vectors of spatial filter coefficients (T1, T2, …, TN), wherein the subset (T1, …, TM) is associated with the most dominant eigenvectors of the first decomposition matrix (U); and compressing (150) the first plurality (NRX) of digital signals utilizing (151) only the subset (T1, …, TM). Corresponding computer program product, processing unit, wireless device, and chips are also disclosed.

Description

[0001] A method of compressing a plurality of digital signals, computer program product, non- transitory computer-readable storage medium, processing unit, wireless device, and chips therefor

[0002] Technical field

[0003] The present disclosure relates to a method of compressing a (first) plurality of digital signals, a computer program product, a processing unit, a wireless device, and a chip therefor.

[0004] More specifically, the disclosure relates to a method of compressing a (first) plurality of digital signals, a computer program product, a processing unit, a wireless device, and a chip as defined in the introductory parts of the independent claims.

[0005] Background art

[0006] US 10833751 B2 discloses that precoding coefficients can be compressed based on user equipment signal interference to noise ratio or path loss in front haul cloud radio access network systems.

[0007] Furthermore, for a full digital beamforming receiver architecture the signal received on a subcarrier k across the receiving (Rx) antennas can be modelled as yk= Hkxk+ wkwhere is a vector containing the input on subcarrier k from each of Nrxspatially distributed Rx antennas, is a channel matrix representing the radio channel between each of Ntxtransmit (Tx) (antenna) ports and Nrxreceiving (Rx) antennas on subcarrier k, Ntxx 1) is a vector containing the modulation symbols transmitted on subcarrier k from each of the Ntx

[0008] Tx (antenna) ports ("MIMO layers"), and is a vector containing interference on subcarrier k in terms of additive white gaussian noise as received across the NrxRx antennas.

[0009] A Minimum Mean Squared Error (MMSE) equalizer for a full digital beamforming receiver is derived as resulting in estimates of transmitted modulation symbols on subcarrier kxkGYk Ntxx 1)-

[0010] However, with this approach, signals received on all RX antennas are passed to baseband, which requires a high data rate and hence may result in high power consumption. Furthermore, the maximum rank of Hkis min(Nrx, Ntx), but depending on the radio channel, the rank may actually be lower. Moreover, single value decomposition (SVD) may be used for deriving inverses or pseudo-inverses in the calculation of Gkto mitigate effects of ill- conditioned matrices, particularly when using a least square (LS) approach rather than a MMSE approach (by which the regularizing identity matrix I in Gkis not present). When implementing digital beamforming in a receiver, such as a wireless device (WD), the fully digital beamformer (receiver architecture) has a high complexity. Thus, there is a need for a method and / or an apparatus / unit with reduced complexity (e.g., having nearly the same performance as a full digital beamforming receiver). Furthermore, there is also a need for a method and / or an apparatus / unit providing more flexibility in representing the radio channel with its variations spatially, spectrally, and / or temporally.

[0011] US 2012 / 0062421 Al discloses a mechanism for mitigating inter-user interference in a multi-user wireless communication environment.

[0012] An object of the present disclosure is to mitigate, alleviate or eliminate one or more of the above-identified deficiencies and disadvantages in the prior art and / or solve at least the above-mentioned problem or other problems.

[0013] According to a first aspect there is provided a method of converting / compressing, by a receiving device comprising spatial filters, a (first) plurality of digital signals obtained from a (first) plurality of frequency-division multiplexed, FDM, signals transmitted from a (second) plurality of transmit antenna ports of a transmitting device and received by the receiving device, the method comprising: obtaining two or more channel estimate matrices associated with propagation channels for the FDM signals; applying a function to the two or more channel estimate matrices to obtain a resulting matrix, the resulting matrix resulting from the applying; decomposing the resulting matrix into a first decomposition matrix, comprising first vectors of coefficients, and a second decomposition matrix, different from the first decomposition matrix, comprising second vectors of coefficients, wherein the first decomposition matrix is a unitary eigenvector matrix comprising one or more eigenvectors; determining vectors of spatial filter coefficients from the first vectors of coefficients; selecting a subset of the vectors of spatial filter coefficients, wherein the subset is associated with one or more of the 4 most dominant eigenvalues of the second decomposition matrix; and converting, by the spatial filters, the (first) plurality of digital signals into a (third) plurality of virtual antenna streams utilizing only the subset.

[0014] According to some embodiments, the number of virtual antenna streams is larger than the number of transmit antenna ports.

[0015] According to some embodiments, the FDM signals are orthogonal frequency-division multiplexed, OFDM, signals. According to some embodiments, the function is a quadratic function.

[0016] According to some embodiments, the subset is associated with the coefficients associated with the 1, 2 or 4 most dominant eigenvectors of the first decomposition matrix (only).

[0017] According to some embodiments, the (first) plurality of FDM signals are transmitted from a (second) plurality of transmit antenna ports of a transmitting device, such as a wireless device or a base station.

[0018] According to some embodiments, the method further comprises sending the (third) plurality of virtual antenna streams to a baseband (BB) processor for further processing / beamforming.

[0019] According to some embodiments, the method further comprises splitting the (third) plurality of virtual antenna streams into a first subset and a second subset, and obtaining (120), performing (130), determining (140), and compressing (150) is performed for each of the first and second subsets.

[0020] According to some embodiments, decomposing comprises performing singular value decomposition, SVD.

[0021] According to some embodiments, decomposing comprises performing eigenvalue decomposition, spectral decomposition, or eigendecomposition of the resulting matrix.

[0022] According to some embodiments, determining vectors of spatial filter coefficients comprises selecting the vectors of spatial filter coefficients as one or more column vectors of the first decomposition matrix.

[0023] According to some embodiments, the vectors of spatial filter coefficients are selected to be orthonormal vectors of spatial filter coefficients.

[0024] According to some embodiments, obtaining one or more channel estimate matrices comprises estimating one or more channel estimate matrices per subcarrier, estimating one or more channel estimate matrices per resource block, or estimating one or more channel estimate matrices per frequency range. According to some embodiments, the function comprises weights, and obtaining comprises configuring the weights per subcarrier, configuring the weights per resource block, configuring the weights per frequency range, and / or configuring the weights per TX (antenna) port (Q).

[0025] According to some embodiments, obtaining comprises determining at a first time instant the two or more channel estimate matrices [Hfe(t)] and determining at a second time instant the two or more channel estimate matrices [Hk(t — T)] and the function is a function of the two or more channel estimate matrices at the first time instant squared and the two or more channel estimate matrices at the second time instant squared.

[0026] According to some embodiments, the receiving device receives information (from the transmitting device) about the number of transmit antenna ports of the transmitting device.

[0027] According to a second aspect there is provided a program product comprising instructions, which, when executed on at least one processor of a processing device, cause the processing device to carry out the method according to the first aspect or any of the embodiments mentioned herein.

[0028] According to a third aspect there is provided a non-transitory computer-readable storage medium storing one or more programs configured to be executed by one or more processors of a processing device, the one or more programs comprising instructions which, when executed by the processing device, causes the processing device to carry out the method according to the first aspect or any of the embodiments mentioned herein.

[0029] According to a fourth aspect there is provided a processing unit configured to cause: obtainment of two or more channel estimate matrices associated with one or two or more propagation channels for FDM signals; application of a function, such as a quadratic function (QF), to the two or more channel estimate matrices to obtain a resulting matrix, the resulting matrix resulting from the application; matrix decomposition of the resulting matrix into a first decomposition matrix, comprising first vectors of coefficients, and a second decomposition matrix, different from the first decomposition matrix, comprising second vectors of coefficients; determination of a vector of spatial filter coefficients from the first vectors of coefficients; selection of a subset of the vectors of spatial filter coefficients, wherein the subset is associated with one or more of the 4 most dominant eigenvalues of the second decomposition matrix, such as the coefficients associated with the 1, 2 or 4 most dominant eigenvalues of the second decomposition matrix; and conversion of the (first) plurality of digital signals into a plurality of virtual antenna streams utilizing only the subset.

[0030] According to some embodiments, the number of virtual antenna streams is larger than a number of transmit antenna ports of a transmitting device utilized for transmitting the plurality of FDM signals.

[0031] According to a fifth aspect there is provided a wireless device (WD) comprising the processing unit of the fourth aspect.

[0032] According to a sixth aspect there is provided a chip comprising the processing unit of the fourth aspect.

[0033] Effects and features of the second, third, fourth, fifth, and sixth aspects are fully or to a substantial extent analogous to those described above in connection with the first aspect and vice versa.

[0034] Embodiments mentioned in relation to the first aspect are fully or largely compatible with the second, third, fourth, fifth, and sixth aspects and vice versa.

[0035] An advantage of some embodiments is that a higher degree-of-freedom and / or more flexibility in representing the radio channel with its variations spatially, spectrally, and temporally, is achieved, e.g., compared to an ordinary MRC-based receiver.

[0036] Another advantage of some embodiments is that power consumption is reduced or optimized (for a wireless device).

[0037] A further advantage of some embodiments is that a lower complexity system / receiver is provided, e.g., compared to a full digital beamforming receiver (with nearly the same performance or with comparable performance).

[0038] Yet a further advantage of some embodiments is that low complexity is achieved.

[0039] Yet another advantage of some embodiments is that implementation is simplified.

[0040] Yet another further advantage of some embodiments is that complexity is reduced. Other advantages are that an improved, more robust and / or more accurate beamforming may be provided and / or that the signal quality is increased.

[0041] The present disclosure will become apparent from the detailed description given below. The detailed description and specific examples disclose preferred embodiments of the disclosure by way of illustration only. Those skilled in the art understand from guidance in the detailed description that changes, and modifications may be made within the scope of the disclosure.

[0042] Hence, it is to be understood that the herein disclosed disclosure is not limited to the particular component parts of the device described or steps of the methods described since such apparatus and method may vary. It is also to be understood that the terminology used herein is for purpose of describing particular embodiments only and is not intended to be limiting. It should be noted that, as used in the specification and the appended claims, the articles "a", "an", "the", and "said" are intended to mean that there are one or more of the elements unless the context explicitly dictates otherwise. Thus, for example, reference to "a unit" or "the unit" may include several devices, and the like. Furthermore, the words "comprising", "including", "containing" and similar wordings does not exclude other elements or steps. Moreover, the term "configured" or "adapted" is intended to mean that a unit or similar is shaped, sized, connected, connectable or otherwise adjusted for a purpose.

[0043] Brief of the

[0044] The above objects, as well as additional objects, features, and advantages of the present disclosure, will be more fully appreciated by reference to the following illustrative and non-limiting detailed description of example embodiments of the present disclosure, when taken in conjunction with the accompanying drawings.

[0045] Figure 1A is a schematic drawing illustrating a wireless device according to some embodiments;

[0046] Figure IB is a flowchart illustrating some method steps according to some embodiments;

[0047] Figure 2 is a schematic drawing illustrating a computer readable (storage) medium according to some embodiments; Figure 3 is a flowchart illustrating actions / method steps implemented in a wireless device or in a processing unit thereof according to some embodiments;

[0048] Figure 4 is a flowchart illustrating some method steps according to some embodiments;

[0049] Figure 5 is a flowchart illustrating some method steps according to some embodiments;

[0050] Figure 6 is a flowchart illustrating some method steps according to some embodiments;

[0051] Figure 7 is a flowchart illustrating some method steps according to some embodiments;

[0052] Figure 8 is a schematic drawing illustrating a chip according to some embodiments;

[0053] Figure 9 is a schematic drawing illustrating a multi-antenna receiver arrangement according to some embodiments; and

[0054] Figure 10 is a schematic drawing illustrating a system comprising wireless devices and transceiver nodes according to some embodiments.

[0055] Detailed description

[0056] The present disclosure will now be described with reference to the accompanying drawings, in which preferred example embodiments of the disclosure are shown. The disclosure may, however, be embodied in other forms and should not be construed as limited to the herein disclosed embodiments. The disclosed embodiments are provided to fully convey the scope of the disclosure to the skilled person.

[0057] Terminology

[0058] Herein is referred to a processor / processing unit. The processor may be a digital processor. Alternatively, the processor may be a microprocessor, a microcontroller, a central processing unit, a co-processor, a graphics processing unit (GPU), a digital signal processor (DSP), an image signal processor, a quantum processing unit, or an analog signal processor. The processing unit may comprise one or more processors and optionally other units, such as a control unit. Thus, the processor may be implemented as a single-processor, a dualprocessor system, or a multiprocessor system. Furthermore, the invention can also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network, e.g., 5G, to one or more local processors. In a distributed computing environment, program modules can be located in both local and remote memory storage devices. Moreover, some processing (e.g., for the data plane) may be moved to a centralized node, such as a centralized transceiver node (TNode). For example, baseband processing and / or higher layer processing, such as processing at layers above the physical layer, may be moved to a cloud, such as an mmW RAN cloud (wherein processing is performed by cloud processors). Such a (mmW) cloud deployment may bring significant cost savings to the operator due to centralized processing, collaborative radio processing, and availability of cheap commodity hardware.

[0059] Herein is referred to a baseband (BB) processor / processing unit. A BB processor is a processor specifically adapted for processing baseband signals / data.

[0060] Herein is referred to millimetre Wave (mmW) utilization, mmW communication, mmW communication capability and mmW frequency range. The mmW frequency range is from 24.25 Gigahertz (GHz) to 71 GHz or more generally from 24 to 300 GHz. The mmW frequency range may also be referred to as Frequency Range 2 (FR2).

[0061] Herein is referred to centimetre Wave (emW) utilization, emW communication, emW communication capability and emW frequency range. The emW frequency range is from 10 Gigahertz (GHz) to 30 GHz.

[0062] Herein is referred to a chip. A chip is an integrated circuit (chip) or a monolithic integrated circuit (chip) and may also be referred to as an IC, or a microchip.

[0063] Herein is referred to a wireless device (WD). A wireless device is any device capable of transmitting or receiving signals wirelessly. Some examples of wireless devices are user equipment (UE), mobile phones, cell phones, smart phones, Internet of Things (loT) devices, vehicle-to-everything (V2X) devices, vehicle-to-infrastructure (V2I) devices, vehicle-to-network (V2N) devices, vehicle-to-vehicle (V2V) devices, vehicle-to-pedestrian (V2P) devices, vehicle- to-device (V2D) devices, vehicle-to-grid (V2G) devices, fixed wireless access (FWA) points, and tablets. Herein is referred to a Transmission Configuration Indicator (TCI) State. A TCI state contains parameters for configuring a quasi-co-location relationship between one or two downlink reference signals and the Demodulation reference signal (DM-RS) ports of the physical downlink shared channel (PDSCH), the DM-RS port of physical downlink control channel (PDCCH) and / or the channel state information reference signal (CSI-RS) port(s) of a CSI-RS resource.

[0064] Herein is referred to a "transceiver node" (TNode). A TNode may be a radio unit (RU), a remote radio unit (RRU), a repeater, a wireless node, or a base station (BS), such as a radio base station (RBS), a Node B, an Evolved Node B (eNB) or a gNodeB (gNB). Thus, a TNode may be a network (NW) node. Furthermore, a TNode may be a BS for a neighbouring cell, a BS for a handover (HO) candidate cell, a radio unit (RRU), a distributed unit (DU), another WD (e.g., a remote WD) or a base station (BS) for a (active / deactivated) secondary cell (SCell) or for a serving / primary cell (PCell, e.g., associated with an active TCI state), a laptop, a wireless station, a relay, a repeater device, a reconfigurable intelligent surface, or a large intelligent surface.

[0065] Herein is referred to an antenna unit. An antenna unit may be one single antenna. However, an antenna unit may also be a dual antenna, such as a dual patch antenna with a first (e.g., horizontal) and a second (e.g., vertical) polarization, thus functioning as two separate antennas or an antenna unit having two ports. Moreover, an antenna unit may be an antenna array, e.g., if analog beamforming is performed.

[0066] The polarization of an antenna refers to the orientation of the electric field of the radio wave transmitted by it and is determined by the physical structure of the antenna and its orientation. E.g., an antenna composed of a linear conductor (such as a dipole or whip antenna) oriented vertically will result in vertical polarization; if turned on its side the same antenna's polarization will be horizontal.

[0067] Herein is referred to vectors. A vector is a mathematical vector or a tuple (and not a physical vector having a direction).

[0068] Other approaches

[0069] Except for the MMSE approach for a fully digital beamforming receiver / architecture described above, it is possible to adapt other approaches. An approach is to introduce spatial filters for a two-stage beamformer. Using spatial filters, the Nrxantenna streams are converted into Ns< Nrxvirtual antenna streams. The spatial filters can be expressed as the columns of a matrix T of dimensions NrxX Ns. The relation between antenna streams and virtual antenna streams can be described as follows: sk= THyk= THHkxk+ THwk(NsX 1).

[0070] An MMSE equalizer is derived as follows: resulting in estimates of transmitted modulation symbols on subcarrier k xk— Gksk(Ntxx 1).

[0071] Furthermore, the maximum rank of THHkis min(As, Ntx), but depending on the radio channel, the rank may be lower. Moreover, SVD may be used for deriving inverses or pseudoinverses in the calculation of Gkto mitigate effects of ill conditioned matrices, particularly when using an LS approach rather than a MMSE approach (by which the regularizing identity matrix I in Gkis not present).

[0072] The equivalence of a full digital beamforming receiver may be derived by using Ns= Nrxfilters, i.e., a matrix T of the dimension NrxX Nrx, and choosing the filters such that the matrix T is unitary, e.g., such that the matrix T is an identity matrix. All information received by the antennas is then passed on to the baseband for second stage / further processing / beamforming.

[0073] However, a drawback with the use of a full digital beamforming receiver / architecture is that it requires a very high data rate between radio units and baseband units. Another drawback is that it requires significant baseband processing in terms of Fast Fourier Transforms (FFTs), channel estimation, and / or channel equalization. One benefit is that a full digital beamforming receiver / architecture can handle any variation in spatio-spectral characteristics.

[0074] Furthermore, the equivalence of antenna selection, by which beamforming in its entirety is left to the baseband, may be defined by using Ns< Nrxorthogonal filters, where for each filter all coefficients but one are zero. A drawback with antenna selection is that energy can only be harvested from Ns< Nrxantennas, thus the full energy received by all the antennas can not be utilized. One benefit is that for the selected subset of antennas, any variation in spatio-spectral characteristics can be handled.

[0075] Moreover, the equivalence of a maximum ratio combining (MRC) digital beamformer may be derived by using Ns= Ntxfilters, each filter representing the conjugate of the average spatial channel characteristics across the spectrum for each of the TX (antenna) ports. Spatially combined information in Nsstreams is passed on to the baseband for further, e.g., a second stage of, beamforming.

[0076] A drawback with MRC digital beamformers is that the number of streams Nsis limited by the number Ntxof observable Tx (antenna) ports. This limits the degrees-of-freedom when handling variations in spatio-spectral characteristics. One benefit with MRC digital beamformers is that it can harvest energy from all Nrxantennas into Ns= Ntxstreams (as opposed to antenna selection), and processing of Ns< Nrxstreams results in lower complexity than required by a full digital beamforming receiver.

[0077] When implementing digital beamforming in a receiver, such as a wireless device (WD), the fully digital beamformer is disadvantageous due to its high complexity and the antenna selection is disadvantageous due to the fact that this approach does not support harvesting of energy from all antennas, thus signal quality may be low or lower than a desired threshold. Therefore, in many cases, the only feasible alternative is utilization of an MRC beamformer. However, the MRC beamformer is limited regarding degrees-of-freedom and / or flexibility, since the filters can only be derived by observing the radio channels from each TX (antenna) port to a respective RX (antenna) port. The number of filters to use when implementing an MRC beamformer is thus limited by (e.g., fixed to) the number of TX (antenna) ports.

[0078] Furthermore, since spatial filters are derived sparsely, e.g., every 20 millisecond (ms), and a radio channel may vary somewhat during that time, and / or since the spatial characteristics may vary across the subcarriers, it is desirable to define filters that allow for a higher degree-of-freedom, i.e., filters that allow more flexibility in representing the radio channel with its variations spatially, spectrally, and / or temporally, than achievable when using e.g., an MRC-based approach in a two-stage digital beamforming architecture. Hence, there is a need for a method and / or an apparatus / unit providing more flexibility in representing the radio channel with its variations spatially, spectrally, and / or temporally.

[0079] Furthermore, there is a need for a method and / or an apparatus / unit having nearly the same performance, but (e.g., significantly) lower complexity than a full digital beamforming receiver.

[0080] Basic concept

[0081] A basic concept of this invention is that spatial filters are derived so that they are capturing the Ns> Ntxprincipal spatio-spectral components of the radio channel. In some embodiments, spatial filters for a two-stage beamformer (that allow higher degree of freedom, e.g., wherein the degree of freedom is larger than Ntx) are derived. When Nsis larger than Ntx, it results in increased degrees-of-freedom compared to what is achievable by purely maximum ratio combining-based digital beamformer. The increase in degrees-of- freedom achieved by the invention described herein results in an increased likelihood that spatio-spectral variations in the radio channel can be accommodated, i.e., results in that weighted combinations of the spatial filters more accurately can describe the radio channel as perceived across the different Rx antennas and at the different subcarriers, including any variations over time between occasions when the spatial filters are updated. Thus, an improved, more robust and / or more accurate beamforming may be provided. Thereby, the power consumption is reduced and / or the signal quality is increased. In some embodiments, the spatial component(s) of the spatio-spectral principal components are determined via eigenvalue decomposition:

[0082] F = Sfc HkHk= UMJH(NrxX Nrx).

[0083] The spatial filters are then extracted from the columns UC1, UC2, ..., UCM of the resulting unitary eigenvector matrix U that are associated with the Nseigenvalues of largest magnitude, where if it is assumed that the eigenvalues on the diagonal of A are sorted in descending order,

[0084] Since the columns of T are orthonormal when designed in this manner, the enhanced MMSE equalizer becomes as follows: i.e., the white property of the noise is preserved (the remaining noise is still white), which is similar to the aforementioned fully digital beamformer receiver. In fact, if keeping all columns of U, i.e., T = U, the two-stage beamformer becomes equivalent to a full digital beamforming receiver. Thus, when designing spatial filters as described above, one can choose by the number of filters Nshow close to the performance by a full digital beamforming receiver to operate. Hence, the approach disclosed in the invention disclosed herein is a consistent complexity reduction method of a fully digital beamformer. The invention described herein can be utilized for frequencies of 10 MHz and higher frequencies, including emW and mmW frequencies.

[0085] Embodiments

[0086] In the following, embodiments will be described where figure 1A illustrates a wireless device according to some embodiments and figure IB illustrates some method steps according to some embodiments. Figure 1A depicts a wireless device (WD) 302. The WD 302 comprises a processing unit 300. Furthermore, the wireless device comprises, in some embodiments, a multi-antenna receiver arrangement 400 (shown in figure 9) and / or one, more or all components thereof. The method 100 is for compressing, by a receiving device (e.g., the WD 302), a (first) plurality (i.e., NRX) of digital signals (or antenna streams). The digital signals are obtained from a (first) plurality (i.e., NRX) of frequency-division multiplexed (FDM) signals received by the receiving device. In some embodiments, the FDM signals are orthogonal frequency-division multiplexed (OFDM) signals. Alternatively, the FDM signals are non- orthogonal frequency-division multiplexed (N-OFDM) signals. As another alternative, the FDM signals comprises a mix of OFDM signals and N-OFDM signals, i.e., the FDM signals comprises one or more OFDM signals and one or more N-OFDM signals. In some embodiments, the method 100 comprises receiving, by the receiving device, a plurality of FDM signals (transmitted from / by a plurality, e.g., NTX, of transmit antenna ports of a transmitting device). In some embodiments, the receiving device receives information (from the transmitting device) about the number (NTX) of transmit antenna ports of the transmitting device (utilized for transmitting the FDM signals). Furthermore, in some embodiments, the method 100 comprises obtaining, by the receiving device (or by analog to digital converters thereof), a plurality of digital signals from the received plurality of FDM signals (e.g., one digital signal for each received FDM signal). Furthermore, in some embodiments, the receiving device is the WD 302. The method 100 comprises obtaining 110, by the processing unit 300, one or two or more channel estimate matrices Hl, H2, ... HK associated with (one or two or more) propagation channels for / associated with the FDM signals. The propagation channels are, in some embodiments, the channels over which the FDM signals are transmitted from the NTX (or a second plurality of) transmit (antenna) ports of a transmitting / sending device (e.g., a transceiver node, such as a base station, BS) to the NRX (first plurality of) antennas / antenna ports of the receiving device / WD 302. In some embodiments, obtaining 110 one or more channel estimate matrices comprises estimating 112 one or more channel estimate matrices per subcarrier, e.g., per one or more (i.e., a subgroup) of 12 subcarriers. Alternatively, obtaining 110 one or more channel estimate matrices comprises estimating 114 one or more channel estimate matrices per resource block. As another alternative, obtaining 110 one or more channel estimate matrices comprises estimating 116 one or more channel estimate matrices per frequency range. The frequency ranges may vary over the system bandwidth (i.e., the span of a frequency range may be different from a span of another frequency range). This may be useful for highly dispersive channels, i.e., an improved performance may be achieved for highly dispersive channels. Some (e.g., a first and / or second) frequency ranges may cover only one or a few subcarriers, while other (e.g., a third and or fourth) frequency ranges may cover one or more resource blocks.

[0087] Furthermore, the method comprises applying 120, by the processing unit 300 or by a baseband (BB) processor, a function F, to the two or more channel estimate matrices Hl, H2, ... HK to obtain a resulting matrix RM. The resulting matrix RM is the matrix resulting from the applying 120. In some embodiments, the function F is a quadratic function QF. An example of such a QF is HkHk, where H is Hermitian transpose, a.k.a. conjugate transpose. Furthermore, in some embodiments, the function F is a polynomial, such as a linear polynomial, a quadratic polynomial, a cubic polynomial, a quartic polynomial (of degree four) or a quintic polynomial (of degree five). In some embodiments, the function F is a positive- definite function. Moreover, in some embodiments (shown in figure 4), obtaining 110 comprises determining 117 at a first time instant t the two or more channel estimate matrices Hk(t) and determining 118 at a second time instant t-r the two or more channel estimate matrices Hk(t — T). In these embodiments, the function F is a function of the two or more channel estimate matrices at the first time instant squared Hk(t)Hk(t) and the two or more channel estimate matrices at the second time instant squared Hk(t — )Hk(t — T). Furthermore, in some embodiments (shown in figure 5), the function F comprises weights. In these embodiments, applying 120 comprises configuring 122 the weights per subcarrier (SC). Alternatively, or additionally, applying 120 comprises configuring 124 the weights per resource block. As another alternative, or additionally, applying 120 comprises configuring 126 the weights per frequency range ak. As yet another alternative, or additionally, applying 120 comprises configuring 128 the weights perTX (antenna) port Q. Moreover, the method comprises decomposing 130 (e.g., utilizing matrix decomposition), e.g., by the processing unit 300, the resulting matrix RM into a first decomposition matrix U and a second decomposition matrix A. The first decomposition matrix U comprises (one or two or more) first vectors of coefficients Ul, U2, ..., UN (or a first set of vectors of coefficients Ul, U2, ..., UN). In some embodiments, the first vectors of coefficients Ul, U2, ..., UN are the column vectors UC1, UC2, ..., UCM of the first decomposition matrix U. The second decomposition matrix A is different from the first decomposition matrix U. In some embodiments, the second decomposition matrix A is a diagonal matrix. Furthermore, the second decomposition matrix A comprises (one or two or more) second vectors of coefficients Al, A2, ..., AN (or a second set of vectors of coefficients Al, A2, ..., AN). In some embodiments, each second vector (or each vector of the second set of vectors) comprises zero or one non-zero coefficient (i.e., all coefficients or all but one coefficient of each second vector are zero-valued). Thus, in these embodiments, each second vector is / comprises either a single non-zero coefficient / value (e.g., an eigenvalue) or no non-zero coefficients / values. Moreover, the first decomposition matrix U is a unitary eigenvector matrix comprising one or two or more eigenvectors. In some embodiments, the second decomposition matrix A is a diagonal matrix, and the diagonal elements of the diagonal matrix are the eigenvalues corresponding to the one or two or more eigenvectors (i.e., the diagonal matrix comprises eigenvalues corresponding to the one or two or more eigenvectors). The method 100 comprises, from the first vectors of coefficients Ul, U2, ..., UN determining 140, e.g., by the processing unit 300, vectors of spatial filter coefficients Tl, T2, ..., TN. In some embodiments, the vectors of spatial filter coefficients Tl, T2, ..., TN are determined directly from the first vectors of coefficients Ul, U2, ..., UN and / or determined from the first vectors of coefficients Ul, U2, ..., UN only. Alternatively, the vectors of spatial filter coefficients Tl, T2, ..., TN are determined directly from the first vectors of coefficients Ul, U2, ..., UN without consulting any second vector of coefficients Al, A2, ..., AN. In some embodiments, determining 140 vectors of spatial filter coefficients Tl, T2, ..., TN comprises selecting 142, e.g., by the processing unit 300, the vectors of spatial filter coefficients Tl, T2, ..., TN as one or two or more column vectors of the first decomposition matrix U. Alternatively, determining 140 vectors of spatial filter coefficients Tl, T2, ..., TN comprises selecting 144, e.g., by the processing unit 300, the vectors of spatial filter coefficients Tl, T2, ..., TN as one or two or more row vectors of the first decomposition matrix U. Furthermore, the method 100 comprises selecting 145, e.g., by the processing unit 300, a subset Tl, ..., TM of the vectors of spatial filter coefficients Tl, T2, ..., TN. The (selected) subset Tl, ..., TM is associated with (e.g., one or more of) the (e.g., 4) most dominant eigenvectors of the first decomposition matrix U and therefore also associated with the most dominant eigenvalues of the second decomposition matrix (since each eigenvector of the first decomposition matrix has a corresponding eigenvalue of the second decomposition matrix A). In some embodiments, the (selected) subset Tl, ..., TM is associated with one or more of the 1, 2, 3 or 4 most dominant eigenvectors of the first decomposition matrix U (and with one or more of the 1, 2, 3 or 4 most dominant eigenvalues of the second decomposition matrix A). As an example, the subset Tl, ..., TM is associated with the most dominant eigenvector / eigenvalue, the second most dominant eigenvector / eigenvalue, the third most dominant eigenvector / eigenvalue and the fourth most dominant eigenvector / eigenvalue (and optionally one or more other eigenvectors / eigenvalues of the, e.g., 10 or 20, most dominant eigenvectors / eigenvalues) of the first / second decomposition matrix U / A. As another example, the subset Tl, ..., TM is associated with the second most dominant eigenvector / eigenvalue, the third most dominant eigenvector / eigenvalue, the fourth most dominant eigenvector / eigenvalue, and the fifth most dominant eigenvector / eigenvalue (and optionally one or more other eigenvectors / eigenvalues of the, e.g., 10 or 20, most dominant eigenvectors / eigenvalues) of the first / second decomposition matrix U / A. As yet another example, the subset Tl, ..., TM is associated with the most dominant eigenvector / eigenvalue, and the second most dominant eigenvector / eigenvalue (and optionally one or more other eigenvectors / eigenvalues of the, e.g., 5, 10 or 20, most dominant eigenvectors / eigenvalues) of the first / second decomposition matrix U / A. As a further example, the subset Tl, ..., TM is associated with the second most dominant eigenvector / eigenvalue, and the third most dominant eigenvector / eigenvalue (and optionally one or more other eigenvectors / eigenvalues of the, e.g., 5, 10, or 20, most dominant eigenvectors / eigenvalues) of the first / second decomposition matrix U / A. As another further example, the subset Tl, ..., TM is associated with the most dominant eigenvector / eigenvalue, and the third most dominant eigenvector / eigenvalue (and optionally one or more other eigenvectors / eigenvalues of the, e.g., 5, 10, or 20, most dominant eigenvectors / eigenvalues) of the first / second decomposition matrix U / A. As a further another example, the subset Tl, ..., TM is associated with the most dominant eigenvector / eigenvalue, the second most dominant eigenvector / eigenvalue, the third most dominant eigenvector / eigenvalue, and / or the fourth most dominant eigenvector / eigenvalue (and optionally one or more other eigenvectors / eigenvalues of the, e.g., 5, 10, or 20, most dominant eigenvectors / eigenvalues) of the first / second decomposition matrix U / A. Moreover, the method 100 comprises, e.g., by the processing unit 300, compressing 150 the (first) plurality of (i.e., NRX) digital signals utilizing only the subset Tl, ..., TM. Thus, in some embodiments, the (first) plurality (i.e., NRX) of digital signals is compressed by utilizing the subset Tl, ..., TM (only) as coefficients in / for / of a spatial filter. In some embodiments, the (first) plurality of FDM signals are transmitted from a (second) plurality (i.e., NTX) of transmit (antenna) ports (and / or transmitters) of a transmitting device (and received by NRX / first plurality of receiving antennas / antenna ports, which may be spatially distributed). In these embodiments, compressing 150 may comprise converting 152, e.g., by the processing unit 300, the (first) plurality of (i.e., NRX) digital signals into a (third) plurality (i.e., NS) of virtual antenna streams. The third plurality is larger than the second plurality, i.e., the number of virtual antenna streams is larger than the number of transmit antenna ports (NS>NTX). Moreover, in some embodiments, the receiving device receives information from the transmitting device (or from an intermediary device) about the number of transmit antenna ports of the transmitting device. In some embodiments (e.g., if the receiving device has received information about the number of transmit antenna ports of the transmitting device from the transmitting device), the number of virtual antenna streams is selected, by the receiving device (or by the processing unit 300), to be larger than the number of transmit antenna ports. In some embodiments, the conversion 152 is a lossy conversion (e.g., by utilizing only a subset of the vectors of spatial filter coefficients), and thus the plurality of digital signals is compressed into a plurality of virtual antenna streams. Alternatively, the conversion 152 is a non-lossy conversion.

[0088] Furthermore, in some embodiments, compressing 150 further comprises disregarding all other eigenvectors of the first decomposition matrix and / or disregarding all vectors of spatial filter coefficients Tl, T2, ..., TN other than the subset Tl, ..., TM, i.e., disregarding the vectors of spatial filter coefficients TM+1, TN not belonging to the subset Tl, TM. By compressing the (first) plurality of digital signals utilizing only the subset Tl, TM, it is guaranteed that the most dominant components (of the resulting matrix) are utilized. Furthermore, by compressing the (first) plurality of digital signals utilizing only the subset Tl, ..., TM complexity is reduced. Moreover, in some embodiments, the method 100 comprises sending 156 the (third) plurality (i.e., NS) of virtual antenna streams to a baseband, BB, processor for further processing / beamforming. In some embodiments, the third plurality (i.e., NS) is a known parameter, e.g., user-defined, standard-defined, or system-defined. In these embodiments, the method 100 may comprise splitting 154 the (third) plurality (i.e., NS) of virtual antenna streams into a first subset and a second subset (e.g., before obtaining 110). Furthermore, in some embodiments, one or more of obtaining 110, applying 120, decomposing 130, determining 140, selecting 145 and compressing 150 is performed for each of the first and second subsets. As an example, splitting 154 may be utilized if there are 2 TX ports (first and second TX ports). A user (or the system) may then define 2 spatial filters for receiving from the first TX port and another 2 (different) spatial filters for receiving from the second TX port. Alternatively, splitting 154 is not utilized and a user (or the system) defines 4 spatial filters to be utilized when receiving regardless of if receiving from the first or the second TX port. By utilizing splitting 154, receiving and / or transmitting is adapted to be performed with a specific TX port (only), e.g., when communicating in a multiple transmission and reception points (multi-TRP) case. Furthermore, in some embodiments, the vectors of spatial filter coefficients Tl, T2, ..., TN and / or the vectors of the subset Tl, ..., TM thereof are selected to be orthonormal vectors of spatial filter coefficients.

[0089] According to some embodiments, a computer program product comprising a non- transitory computer readable medium 200, such as a punch card, a compact disc (CD) ROM, a read only memory (ROM), a digital versatile disc (DVD), an embedded drive, a plug-in card, or a universal serial bus (USB) memory, is provided. Figure 2 illustrates an example computer readable medium in the form of a compact disc (CD) ROM 200. The computer readable medium has stored thereon, a computer program comprising program instructions. The computer program is loadable into a data processor (PROC) 220, which may, for example, be comprised in a computer or a computing device or the control unit 210. When loaded into the data processor 220, the computer program may be stored in a memory (MEM) 230 associated with or comprised in the data processor 220. According to some embodiments, the computer program may, when loaded into and run by the data processor 220, cause execution of method steps according to, for example, the method illustrated in figure IB, which is described herein. Furthermore, in some embodiments, there is provided a computer program product comprising instructions, which, when executed on at least one processor of a processing device, cause the processing device to carry out the method illustrated in figure IB. Moreover, in some embodiments, there is provided a non-transitory computer-readable storage medium storing one or more programs configured to be executed by one or more processors of a processing device, the one or more programs comprising instructions which, when executed by the processing device, causes the processing device to carry out the method illustrated in figure IB.

[0090] Figure 3 illustrates actions / method steps caused by the processing unit 300 (described above in connection with figures 1A and IB) comprised or comprisable in the WD 302 (described above in connection with figures 1A and IB). In some embodiments, the processing unit 300 is configured to cause reception, by the receiving device, of a plurality of FDM signals (transmitted from / by a plurality, e.g., NTX, of transmit antenna ports of a transmitting device). To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first reception unit (e.g., first receiving circuitry, or a first receiver). Furthermore, in some embodiments, the processing unit 300 is configured to cause obtainment, by the receiving device (or by analog to digital converters thereof), of a plurality of digital signals from the received plurality of FDM signals (e.g., one digital signal for each received FDM signal). To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) the first or second obtainment unit (e.g., first or second obtaining circuitry, or first / second obtainer). The processing unit 300 is configured to cause obtainment 310 of two or more channel estimate matrices Hl, H2, ..., HK associated with the propagation channels for the FDM signals. To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first obtainment unit (e.g., first obtaining circuitry, or a first obtainer). Furthermore, the processing unit 300 is configured to cause application 320 of a function F, such as a quadratic function QF, to the two or more channel estimate matrices to obtain a resulting matrix RM. The resulting matrix RM results from the application 320. To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first application unit (e.g., first applying circuitry, a first applier, or a second processor, such as a BB processor). Moreover, the processing unit 300 is configured to cause matrix decomposition 330 of the resulting matrix RM into a first decomposition matrix U, comprising first vectors of coefficients Ul, U2, UN, and a second decomposition matrix A, different from the first decomposition matrix U, comprising second vectors of coefficients Al, A2, ..., AN. To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first decomposing unit (e.g., first decomposing circuitry, a first decomposer, or a second processor, such as a BB processor). The processing unit 300 is configured to cause, from the first vectors of coefficients Ul, U2, ..., UN, determination 340 of a vector of spatial filter coefficients Tl, T2, ..., TN. To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first determining unit (e.g., first determining circuitry, a first determiner, or a second processor, such as a BB processor). Furthermore, the processing unit 300 is configured to cause selection 345 of a subset Tl, ..., TM of the vectors of spatial filter coefficients Tl, T2, ..., TN, wherein the subset Tl, ..., TM is associated with the most dominant eigenvectors of the first decomposition matrix U (and therefore associated with the most dominant eigenvalues of the second decomposition matrix), such as the coefficients associated with the 1, 2 or 4 most dominant eigenvectors of the first decomposition matrix U (and thus the coefficients associated with the 1, 2 or 4 most dominant eigenvalues of the second decomposition matrix). To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first selecting unit (e.g., first selecting circuitry, a first selector, or a second processor, such as a BB processor). Moreover, the processing unit 300 is configured to cause compression 350 of the (first) plurality (i.e., NRX) of digital signals utilizing only the subset Tl, ..., TM. To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first compressing unit (e.g., first compressing circuitry, a first compressor, or a spatial filter, such as the spatial filter described below in connection with figure 9). Moreover, in some embodiments, the processing unit 300 is configured to cause sending 356 of the (third) plurality (i.e., NS) of virtual antenna streams to a baseband, BB, processor for further processing / beamforming. To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first sending unit (e.g., first sending circuitry, or a first sender. In some embodiments, the processing unit 300 is configured to cause splitting 354 of the (third) plurality (i.e., NS) of virtual antenna streams into a first subset and a second subset (e.g., before obtainment 310). To this end, the processing unit 300 may be associated with (e.g., operatively connectable, or connected, to) a first splitting unit (e.g., first splitting circuitry, or a first splitter. In some embodiments, obtainment 320, decomposition 330, determination 340, and compression 350 is performed for each of the first and second subsets. Additionally, the processing unit 300 is, in some embodiments, configured to cause one or more actions / method steps corresponding to the method steps described below in connection with figures 4-7.

[0091] Figures 4-7 illustrate some method steps according to some embodiments. Figure 4 depicts that, in some embodiments, obtaining 110 one or more channel estimate matrices comprises estimating 112 one or more channel estimate matrices per subcarrier. As seen in figure 4, alternatively, obtaining 110 one or more channel estimate matrices comprises estimating 114 one or more channel estimate matrices per resource block. As another alternative (shown in figure 4), obtaining 110 one or more channel estimate matrices comprises estimating 116 one or more channel estimate matrices per frequency range.

[0092] Moreover, in some embodiments, obtaining 110 comprises determining 117 at a first time instant t the two or more channel estimate matrices Hk(t) and determining 118 at a second time instant t-r the two or more channel estimate matrices Hk(t - T). In these embodiments, the function F (applied by applying 120) is a function of the two or more channel estimate matrices at the first time instant squared (Hk(t)Hk(t)) and the two or more channel estimate matrices at the second time instant squared (Hk(t — )Hk(t — T)). In some embodiments, the function F comprises weights. Figure 5 depicts that, in these embodiments, applying 120 comprises configuring 122 the weights per subcarrier (SC). Alternatively, or additionally (as shown in figure 5), applying 120 comprises configuring 124 the weights per resource block, RB. As another alternative (shown in figure 5), or additionally, applying 120 comprises configuring 126 the weights per frequency range ak. As yet another alternative (shown in figure 5), or additionally, applying 120 comprises configuring 128 the weights per TX (antenna) port Q.

[0093] Figure 6 illustrates that, in some embodiments, determining 140 vectors of spatial filter coefficients Tl, T2, ..., TN comprises selecting 142 the vectors of spatial filter coefficients Tl, T2, ..., TN as one or more column vectors of the first decomposition matrix U. Alternatively (as shown in figure 6), determining 140 vectors of spatial filter coefficients Tl, T2, ..., TN comprises selecting 144 the vectors of spatial filter coefficients Tl, T2, TN as one or more row vectors of the first decomposition matrix U.

[0094] In some embodiments, the (first) plurality of FDM signals are transmitted from a (second) plurality of transmit (antenna) ports (i.e., NTX transmit antenna ports and / or transmitters) of a transmitting device (and received by NRX receiving antennas / antenna ports, which may be spatially distributed). In these embodiments, as illustrated by figure 7, the method 100 or the method step compressing 150 comprises converting 152 the (first) plurality (i.e., NRX) of digital signals into a (third) plurality (i.e., NS) of virtual antenna streams, and the third plurality is larger than the second plurality (i.e., NS>NTX). Moreover, in some embodiments, as shown in figure 7, the method 100 comprises (e.g., after compressing 150 or as part of compressing 150) sending 156 the (third) plurality (i.e., NS) of virtual antenna streams to a baseband, BB, processor for further processing / beamforming.

[0095] As described above, weighting or weights may be applied when deriving spatial filters or coefficients thereof. Thus, the spatial characteristics of some subcarriers (SCs) are given lower weights than other subcarriers. In one non-limiting example, a weight 0<a_k<l may be applied per subcarrier when forming the matrix to be decomposed, i.e., each subcarrier SCI, SC2, ..., SCk may be associated with a respective weight al, a2, ..., ak and each weight al, a2, ..., ak has a value in the range from 0 to 1. The value of each weight al, a2, ..., ak may depend on the signal-to-noise-plus-interference ratio (SINR) of / associated with the corresponding subcarrier SCI, SC2, ..., SCk, the rank of / associated with the corresponding subcarrier SCI, SC2, ..., SCk, the importance of a subband manifested as a range of subcarriers, and / or the interference received in a subband. If it is desirable that a specific subcarrier SCn contributes more to the spatial filter(s), the value of the corresponding weight an for the specific subcarrier SCn is increased and if it is desirable that a specific subcarrier Scm contributes less to the spatial filter(s), the value of the corresponding weight am for the specific subcarrier is decreased.

[0096] Furthermore, weighting or weights may be applied when deriving the spatial filters, by giving / setting different weights to the spatial characteristics for different Tx ports, i.e., by configuring the weights per TX port (Q). In a non-limiting example, a diagonal weight matrix Q of size NTXxNTX is used. The diagonal elements of the diagonal weight matrix Q are all from 0 to 1. All other elements of the diagonal weight matrix Q are zero. Furthermore, the p:th diagonal element constitutes the weight imposed on the contribution by the p:th Tx port when forming the resulting matrix, i.e., the matrix to be decomposed. Thus, a lower value gives less influence, and a larger value gives more influence on the spatial filters. It is recognized that other forms of weighing can be carried out. For instance, different diagonal matrices can be used for different subcarriers to allow different weighting in different parts of the spectrum. This is merely one example of weighting. It is recognized that weighting may be carried out in many different ways. Regardless of approach, the desirable net effect is to deemphasize some subcarriers over others when deriving the spatial filters.

[0097] In some embodiments, the decomposition carried out (during decomposing 130) for deriving spatial filters at one time instant t is based on information from one or more previous time instants at which channel estimates across all NrxRx antennas have been computed and / or one or more previous time instants t-rl, t-r2, ..., t-rk at which spatial filters have been derived. As an example, the decomposition is based on a weighted average, 0 < p < 1,

[0098] As another example, the decomposition is based on a recursive equation / algorithm,

[0099] 0 < p < 1, M(t < 0) = 0

[0100] The above examples are merely two examples of how to include information / statistics from one or more previous time instants when deriving spatial filters. Regardless of approach, the desirable net effect is to have previous spatial characteristics to have some influence (through a weighting factor P) when deriving the spatial filters based on the current spatial characteristics. One reason to do so is to account for a higher variability of the radio channel than observed during a single time instant when deriving the spatial filters. Another reason is to reduce the noise in channel estimates by averaging across two or more time instants. Whether to include information from previous time instant(s) may depend on whether the Transmission Configuration Indicator (TCI) state at a previous time instant is the same as for the current time instant. If the TCI state at a previous time instant is not the same as for the current time instant, one may choose to derive the spatial filters solely on spatial characteristics from the current time instant, i.e., no inclusion of information from previous time instant(s).

[0101] Whether to include information from previous time instant(s) depends, in some embodiments, on the time between the current and the previous time instant T. In some embodiments, if T is larger than a time threshold, e.g., 200 ms, the spatial filter(s) are selected solely based on spatial characteristics from the current time instant (i.e., without inclusion of information from previous time instants) and if T is smaller than the time threshold, e.g., 200 ms, the spatial filter(s) are selected based on spatial characteristics from the current time instant as well as spatial characteristics from one or more previous time instants. Alternatively, instead of having a fixed time threshold (value), the time threshold (value) depends on acquired / measured / detected channel characteristics, e.g., the less variability the radio channel has (over time), the larger the time threshold is.

[0102] Alternatively, or additionally, whether to include information from previous time instant(s) depends on the (instantaneous) SINR of received reference signals, synchronization signals, or otherwise known signals used for channel estimation, e.g., when forming Hk. If SINR is lower than a SINR threshold, e.g., OdB, information from previous time instant(s) is included, whereas If SINR is higher than the SINR threshold, information from previous time instant(s) is not included.

[0103] The derived spatial filters may, in some embodiments, be only nearly orthogonal rather than completely orthogonal as a result of mapping the derived filter coefficients to spatial filter coefficients in a fixed-point representation with limited number of integer and fractional bits. The key essence of the invention is the way of deriving the spatial filters based on the principal spatio-spectral components. Then necessary engineering considerations such as mapping results to representations in fixed and potentially a smaller number of bits, with floating or fixed point, may lead to minor inaccuracies due to quantization and adaptation to fixed-point arithmetic. Therefore, in some embodiments, the (derived) spatial filters are scaled, e.g., mutually orthogonally and / or mutually orthonormally. In some embodiments, the (derived) spatial filters are scaled mutually orthogonally (but not necessarily orthonormally), i.e., so that the scaled spatial filters are mutually orthogonal (but not necessarily orthonormal). Furthermore, in some embodiments, due to finite precision and / or fixed-point representations, orthogonal (base) vectors derived / obtained are not absolute orthogonal but only nearly orthogonal. By scaling (e.g., mutually orthogonally) the spatial filters, an improved fixed-point representation of the filter coefficients may be achieved.

[0104] Although described above that a (third) plurality of (i.e., NS) filters is derived from the spatio-spectral characteristics of all Tx ports, in some (alternative) embodiments, the Tx ports are divided into subsets, and for each subset of Tx ports, a corresponding subset of filters is derived. Furthermore, in some (alternative) embodiments, subcarriers are divided into subsets, and for each subset of subcarriers, a corresponding subset of filters is derived.

[0105] Figure 8 illustrates a chip 990 according to some embodiments. The chip 990 comprises the processing unit 300. Furthermore, in some embodiments, the chip 900 comprises the multi-antenna receiver arrangement 400 (and / or one or more of the components thereof) described below in connection with figure 9.

[0106] Figure 9 illustrates a multi-antenna receiver arrangement 400. As seen in figure 9, the multi-antenna receiver arrangement 400 comprises a (first) plurality N (NRX) of receivers / transceivers 500, 501, ..., 515 configured to receive the (first) plurality (NRX) of analog radio signals via a (first) plurality (NRX) of antenna units / ports 700, 701, ..., 715 (N=NRX). In some embodiments, the multi-antenna receiver arrangement 400 comprises the (first) plurality (NRX) of antenna units / ports 700, 701, ..., 715 (for receiving the analog radio signals). Furthermore, the multi-antenna receiver arrangement 400 comprises a (fourth) plurality (I) of analog to digital converters (ADCs) 600, 601, ..., 615 configured to convert the (first) plurality (NRX) of analog radio signals into a (first) plurality (NRX) of digital (baseband) signals. The (fourth) plurality (I) may be equal to the (first) plurality (NRX), i.e., there is one ADC for each receiver / transceiver / analog signal (I =N RX). However, in other embodiments, the (fourth) plurality is twice as large as the (first) plurality (i.e., 2N), i.e., there are two ADCs for each analog signal, e.g., one for an in-phase (I) branch and one for a quadrature phase (Q) branch (I =2 N or l=2*NRX). Moreover, the multi-antenna receiver arrangement 400 comprises an extraction unit 900 configured to extract physical resources used for estimating channel characteristics, such as reference signals, from each of the (first) plurality (NRX) of digital signals. In some embodiments, the extraction unit 900 comprises a (first) plurality (NRX) of sub-extraction units 901, 902, ..., 916, i.e., one sub-extraction unit for each digital signal. The multi-antenna receiver arrangement 400 comprises a channel analyzer 920. The channel analyzer 920 is configured to determine characteristics for each of the (first) plurality (NRX) of digital signals based on the extracted reference signals. Furthermore, the multi-antenna receiver arrangement 400 comprises a (second) plurality (m or NTX) of spatio-temporal filters or spatial filters 800, ..., 807. The spatio-temporal / spatial filters 800, ..., 807 are configured to process or processes the (first) plurality (NRX) of digital signals to obtain a (third) plurality (NS) of combined signals. In some embodiments, the (first) plurality (NRX) is larger / greater than the (third) plurality (NS), i.e., NRX>NS). In some embodiments, the (third) plurality (NS) is 2 or larger, e.g., 3. In some embodiments, the (first) plurality is 3 or larger, e.g., 16. In some embodiments, the multi-antenna receiver arrangement 400 comprises a transform unit 940. The transform unit 940 is configured to transform or transforms each of the (third) plurality (NS) of combined signals into a frequency domain. In some embodiments, the transform unit 940 is or comprises a (third) plurality (NS) of transform sub-units. Each transform sub-unit is configured (connected and otherwise adapted) to process a respective signal of the (third) plurality (NS) of combined signals. In some embodiments, the transform unit transforms each of the combined signals in a serial manner. In some embodiments, a (third) plurality of transform sub-units process NRX / NS signals each. Furthermore, in some embodiments, the multi-antenna receiver arrangement 400 comprises a post-processing unit 960. The postprocessing unit 960 is configured to post-process or post-processes the transformed signals in the frequency domain to obtain a (fifth) plurality (k) of frequency domain processed signals. Moreover, in some embodiments, the (first) plurality (NRX) of analog radio signals is coded. Thus, in some embodiments, the multi-antenna receiver arrangement 400 comprises a decoder 980. The decoder 980 is configured to decode or decodes the (fifth) plurality (k) of frequency domain processed signals (in order to obtain information signals). The (third) plurality (NS) is larger / greater than the (fifth) plurality (k), i.e., NS>k. In some embodiments, the spatio-temporal filters (spatial filters) 800, ..., 807 of figure 9 are configured, e.g., by the processing unit 300, with the subset (Tl, ..., TM) of the vectors of spatial filter coefficients (Tl, T2, ..., TN) described above. Thus, in some embodiments, the coefficients of the spatiotemporal filters or spatial filters 800, ..., 807 are determined / selected according to the method 100 described above and the spatio-temporal filters or spatial filters 800, ..., 807 are utilized to compress the (first) plurality (i.e., NRX) of digital signals utilizing only the subset Tl, ..., TM. In some embodiments, the WD 302 comprises the multi-antenna receiver arrangement 400.

[0107] Figure 10 illustrates a system 999. The system 999 may be a wireless / cellular communication system, a cellular network, a mobile network, a telecommunications network, a cellular radio system, a digital cellular network, a mobile phone network, a mobile phone cellular network, such as 1G, 2G, 3G, 4G, 5G, 6G or similar. Furthermore, the system 999 comprises one or more wireless devices (WD) 302, 303, ..., 308. Moreover, the system 999 comprises one or more transceiver nodes (TNodes) 397, 398, 399. The one or more transceiver nodes (TNodes) 397, 398, 399 may be base stations (gNBs, eNBs, RBS), remote radio units (RRUs), or remote wireless nodes. The WD 302 (as well as the WDs 303, ..., 308) is, in some embodiments, configured to communicate with (e.g., send and / or receive signals, such as radio signals, e.g., comprising baseband / information signals, to / from) one or more of the remote transceiver nodes (TNodes) 397, 398, 399. In some embodiments, the system comprises one or more transmitting devices (e.g., 303, 304, ..., 308, 397, 398) and one or more receiving devices (e.g., 302, 303, 397, 399).

[0108] A (nonzero) vector v of dimension N is an eigenvector of a square N x N matrix A if it satisfies a linear equation of the form

[0109] Av = Av for some scalar X. Then the scalar (X) is called the eigenvalue corresponding to v (i.e., each eigenvector has a corresponding eigenvalue). Geometrically speaking, the eigenvectors of A are the vectors that A merely shrinks or elongates, and the amount that they shrink / elongate by is the eigenvalue. The above equation is called the eigenvalue equation or the eigenvalue problem. Furthermore, as mentioned above under "Basic Concept", the columns UC1, UC2, ..., UCM of the resulting unitary eigenvector matrix U (i.e., the eigenvectors) are associated with the Nseigenvalues. Thus, each column (or eigenvector) of the resulting unitary eigenvector matrix is associated with (or has) a corresponding eigenvalue.

[0110] List of examples:

[0111] Example 1. A method (100) of controlling a Multi-Antenna Transmitter and Receiver Arrangement, MATARA, (400), wherein the MATARA (400) is comprisable in a wireless device, WD, (402) and comprises a (first) plurality of transceivers (500, 501, ..., 515), the method comprising: configuring (110) a first PLL (520) to control centre frequency generation for a first set of transceivers (500, 501); configuring (120) a second PLL (522) to control centre frequency generation for a second set of transceivers (514, 515); receiving (130), by the MATARA (400), a first radio signal transmitted by a first remote transceiver node, TNode, (398), on a first centre frequency (CF1) during a first time period (Tl); configuring (140) a (second) plurality of spatio-temporal filters (800, ..., 807) with a first set of filter coefficients; determining (150) whether reception, by the MATARA (400), of a second radio signal transmitted by a second remote TNode (398, 399) at a second centre frequency (CF2), different from the first centre frequency (CF1), during a second time period (T2) following the first time period, is necessary; determining (160) whether reception, by the MATARA (400), of the first radio signal at a third centre frequency (CF3), different from the second centre frequency (CF2), during a third time period (T3) following the second time period (T2), is necessary; upon determining that reception of the second radio signal at the second centre frequency (CF2) during the second time period (T2) and reception of the first radio signal at the third centre frequency (CF3) during the third time period (T3) is necessary, configuring (170) the (second) plurality of spatio-temporal filters (800, ..., 807) with a second set of filter coefficients, wherein the second set of filter coefficients is selected so that signals from the first set of transceivers (500, 501) are only combined with signals from the first set of transceivers (500, 501) and signals from the second set of transceivers (514, 515) are only combined with signals from the second set of transceivers (514, 515).

[0112] Example 2. The method of example 1, wherein the first radio signal received during the first time period (Tl) and the first radio signal received during the third time period (T3) have the same Transmission Configuration Indicator, TCI, state.

[0113] Example 3. The method of example 1 or example 2, wherein the second time period (T2) is a measurement gap.

[0114] Example 4. The method of example 1 or example 2, wherein the MATARA (400) is configured to monitor only a first bandwidth part, BWP, during the first and third time periods (Tl, T3), and wherein the MATARA (400) is configured to monitor only a second BWP, different from the first BWP, during the second time period (T2).

[0115] Example 5. The method of any of examples 1-5, wherein the third centre frequency (CF3) is the same as the first centre frequency (CF1).

[0116] Example 6. A computer program product comprising a non-transitory computer readable medium (200), having stored thereon a computer program comprising program instructions, the computer program being loadable into a data processing unit (220) and configured to cause execution of the method of any of examples 1-5 when the computer program is run by the data processing unit (220).

[0117] Example 7. A control unit (995) for a Multi-Antenna Transmitter and Receiver Arrangement, MATARA, (400), wherein the MATARA (400) is comprisable in a wireless device, WD, (402), and wherein the MATARA (400) comprises: a (first) plurality of transceivers (500, 501, ..., 515) configured to receive a (first) plurality of analog radio signals via a (first) plurality of antenna units (700, 701, ..., 715); a plurality of analog to digital converters (600, 601, ..., 615) configured to convert the (first) plurality of analog radio signals into a (first) plurality of digital signals; an extraction unit (900) configured to extract reference signals from each of the (first) plurality of digital signals; a channel analyser (920) configured to determine characteristics for each of the (first) plurality of digital signals based on the extracted reference signals; a (second) plurality of spatio-temporal filters (800, ..., 807) configured to process the (first) plurality of digital signals to obtain a (second) plurality of combined signals; a first phase lock loop, PLL, (520); and a second PLL (522); the control unit (995) configured to cause: configuration (310) of the first PLL (520) to control the centre frequency generation for a first set of transceivers (500, 501); configuration (320) of the second PLL (522) to control the centre frequency generation for a second set of transceivers (514, 515); reception (330), by the MATARA (400), of the first radio signal transmitted by a first remote transceiver node, TNode, (398), on a first centre frequency (CF1) during a first time period (Tl); configuration (340) of the (second) plurality of spatio-temporal filters (800, 807) with a first set of filter coefficients; determination (350) of whether reception, by the MATARA (400), of a second radio signal, different from the first radio signal, transmitted by a second remote TNode (398, 399) at a second centre frequency (CF2), different from the first centre frequency (CF1), during a second time period (T2) following the first time period (Tl), is necessary; determination (360) of whether reception, by the MATARA (400), of the first radio signal, at a third centre frequency (CF3), different from the second centre frequency (CF2), during a third time period (T3) following the second time period (T2), is necessary; upon determining that reception of the second radio signal at the second centre frequency (CF2) during the second time period (T2) and reception of the first radio signal at the third centre frequency (CF3) during the third time period (T3) is necessary, configuration (370) of the (second) plurality of spatio-temporal filters (800, ..., 807) with a second set of filter coefficients, wherein the second set of filter coefficients are selected so that signals from the first set of transceivers (500, 501) are only combined with signals from the first set of transceivers (500, 501) and signals from the second set of transceivers (514, 515) are only combined with signals from the second set of transceivers (514, 515).

[0118] Example 8. A Multi-Antenna Transmitter and Receiver Arrangement, MATARA, (400), wherein the MATARA, (400) is comprisable in a wireless device, WD, (402), the MATARA, (400) comprising: a (first) plurality of transceivers (500, 501, ..., 515) configured to receive a (first) plurality of analog radio signals via a (first) plurality of antenna units (700, 701, ..., 715); a plurality of analog to digital converters (600, 601, ..., 615) configured to convert the (first) plurality of analog radio signals into a (first) plurality of digital signals; an extraction unit (900) configured to extract reference signals from each of the (first) plurality of digital signals; a channel analyzer (920) configured to determine characteristics for each of the (first) plurality of digital signals based on the extracted reference signals; a (second) plurality of spatio-temporal filters (800, 807) configured to process the

[0119] (first) plurality of digital signals to obtain a (second) plurality of combined signals; a first phase lock loop, PLL, (520); a second PLL (522); optionally the (first) plurality of antenna units (700, 701, ..., 715); and the control unit (995) of example 8.

[0120] Example 9. A wireless device, WD, (402) comprising the multi-antenna transmitter and receiver arrangement, MATARA of example 8, (300), and the (first) plurality of antenna units (700, 701, ..., 715).

[0121] Example 10. A chip (990) comprising the control unit (995) of example 7.

[0122] Generally, all terms used herein are to be interpreted according to their ordinary meaning in the relevant technical field, unless a different meaning is clearly given and / or is implied from the context in which it is used. Reference has been made herein to various embodiments. However, a person skilled in the art would recognize numerous variations to the described embodiments that would still fall within the scope of the claims. For example, the method embodiments described herein discloses example methods through steps being performed in a certain order. However, it is recognized that these sequences of events may take place in another order without departing from the scope of the claims. Furthermore, some actions / method steps may be performed in parallel even though they have been described as being performed in sequence. Thus, the steps of any methods disclosed herein do not have to be performed in the exact order disclosed, unless a step is explicitly described as following or preceding another step and / or where it is implicit that a step must follow or precede another step. In the same manner, it should be noted that in the description of embodiments, the partition of functional blocks into particular units is by no means intended as limiting. Contrarily, these partitions are merely examples. Functional blocks described herein as one unit may be split into two or more units. Furthermore, functional blocks described herein as being implemented as two or more units may be merged into fewer e.g., a single) unit. Any feature of any of the embodiments / aspects disclosed herein may be applied to any other embodiment / aspect, wherever suitable. Likewise, any advantage of any of the embodiments may apply to any other embodiments, and vice versa. Hence, it should be understood that the details of the described embodiments are merely examples brought forward for illustrative purposes, and that all variations that fall within the scope of the claims are intended to be embraced therein.

[0123] List of some acronyms and abbreviations that may appear in the description

[0124] 3GPP - 3rd Generation Partnership Project

[0125] 5G - fifth generation

[0126] 5G - NR (5G - New Radio) is a new RAT developed by 3GPP for the 5G mobile network

[0127] ADC - analog-to-digital converter

[0128] AGC - automatic gain controller

[0129] BB - baseband

[0130] BF - beamforming

[0131] BW - bandwidth

[0132] BWP - bandwidth part emW - centimetre Wave

[0133] CSI-RS - channel state information reference signal

[0134] CU - control unit

[0135] DAC - digital-to-analog converter

[0136] DCI - downlink control information

[0137] DIC - Digital Interface Chip

[0138] DL-PRS - downlink positioning reference signal

[0139] DM-RS - demodulation reference signal

[0140] DS - down-sampling

[0141] FDM - frequency-division multiplexing

[0142] FFT - Fast Fourier Transform

[0143] FR1 - Frequency Range 1 FR1.5 - Frequency Range 1.5

[0144] FR2 - Frequency Range 2

[0145] Fe - Front end

[0146] FWA - Fixed Wireless Access

[0147] GNSS - Global navigation satellite system

[0148] GPS - Global Positioning System

[0149] IF - intermediate frequency

[0150] I / O - input / output

[0151] LI - Layer 1

[0152] LNA - Low Noise Amplifier

[0153] LO - Local Oscillator

[0154] LoS - Line of Sight

[0155] LTE - Long-Term Evolution

[0156] MAC - Medium Access Control

[0157] MATARA - multi-antenna transmitter and receiver arrangement

[0158] MIMO - multiple input, multiple output

[0159] MMSE - Minimum Mean Squared Error mmW - millimetre wave

[0160] MRC - maximum ratio combining

[0161] NAS - Non-access Stratum nLoS - non-Line of Sight

[0162] NRX - Number of received signals NS - Number of Streams

[0163] NTX - Number of transmitting ports

[0164] OFDM - orthogonal frequency-division multiplexing

[0165] PA - power amplifier

[0166] PBCH - Physical Broadcast Channel

[0167] PCB - printed circuit board

[0168] PCell - primary cell

[0169] PDCCH - physical downlink control channel

[0170] PDP - Power delay profile

[0171] PDSCH - physical downlink shared channel

[0172] PHY - Physical Layer

[0173] PLL - phase locked loop

[0174] PSCell - primary secondary cell

[0175] PSS - primary synchronization signal

[0176] PT-RS - Phase Tracking Reference signal

[0177] PUCCH - physical uplink control channel

[0178] PUSCH - physical uplink shared channel

[0179] QCL - quasi co-located

[0180] QoS - quality of service

[0181] RAT - radio access technology

[0182] RRC - radio resource control

[0183] RSRP - Reference Signal Received Power RSRQ - Reference Signal Received Quality

[0184] RSSI - Received Signal Strength Indicator

[0185] SCell - Secondary Cell

[0186] SNR - Signal-to-noise ratio SSB - Synchronization Signal Block

[0187] SRS - sounding reference signal

[0188] SSS - secondary synchronization signal

[0189] STEF - spatio-temporal filter

[0190] STF - spatial transmission filter TCI - Transmission Configuration Indicator

[0191] TNode - transceiver node

[0192] VGA - variable gain amplifier

[0193] WD - wireless device

Claims

CLAIMS1. A method (100) of converting, by a receiving device comprising spatial filters (800, ..., 807), a plurality of digital signals obtained from a plurality (NRX) of frequency-division multiplexed, FDM, signals transmitted from a plurality (NTX) of transmit antenna ports of a transmitting device and received by the receiving device, the method comprising: obtaining (110) two or more channel estimate matrices (Hl, H2, ..., HK) associated with propagation channels for the FDM signals; applying (120) a function (F) to the two or more channel estimate matrices to obtain a resulting matrix (RM), the resulting matrix (RM) resulting from the applying (120); decomposing (130) the resulting matrix (RM) into a first decomposition matrix (U), comprising first vectors of coefficients (Ul, U2, ..., UN), and a second decomposition matrix (A), different from the first decomposition matrix (U), comprising second vectors of coefficients (Al, A2, ..., AN), wherein the first decomposition matrix (U) is a unitary eigenvector matrix comprising one or more eigenvectors; determining (140) vectors of spatial filter coefficients (Tl, T2, ..., TN) from the first vectors of coefficients (Ul, U2, ..., UN); selecting (145) a subset (Tl, ..., TM) of the vectors of spatial filter coefficients (Tl, T2, ..., TN), wherein the subset (Tl, ..., TM) is associated with one or more of the four most dominant eigenvalues of the second decomposition matrix (U); and converting (152), by the spatial filters 800, ..., 807, the plurality (NRX) of digital signals into a plurality (NS) of virtual antenna streams utilizing (151) only the subset (Tl, ..., TM), wherein the number of virtual antenna streams is larger than the number of transmit antenna ports.

2. The method of claim 1, wherein the plurality (NRX) of FDM signals is a plurality of orthogonal frequency-division multiplexed, OFDM, signals, and wherein the function (F) is a quadratic function.

3. The method of claim 2, further comprising: sending (156) the plurality (NS) of virtual antenna streams to a baseband, BB, processor for further processing / beamforming.

4. The method of claim 2 or claim 3, further comprising: splitting (154) the plurality (NS) of virtual antenna streams into a first subset and a second subset, and wherein obtaining (120), performing (130), determining (140), and compressing (150) is performed for each of the first and second subsets.

5. The method of any one of claims 1-4, wherein decomposing (130) comprises performing singular value decomposition, SVD, or performing eigenvalue decomposition.

6. The method of any one of claims 1-5, wherein determining (140) vectors of spatial filter coefficients (Tl, T2, ..., TN) comprises selecting (144) the vectors of spatial filter coefficients (Tl, T2, ..., TN) as one or more column vectors of the first decomposition matrix (U).

7. The method of any one of claims 1-6, wherein the vectors of spatial filter coefficients (Tl, T2, ..., TN) are selected to be orthonormal vectors of spatial filter coefficients.

8. The method of any one of claims 1-7, wherein obtaining (110) one or more channel estimate matrices comprises estimating (112) one or more channel estimate matrices per subcarrier, estimating (114) one or more channel estimate matrices per resource block, or estimating (116) one or more channel estimate matrices per frequency range.

9. The method of any one of claims 1-8, wherein the function (F) comprises weights, and wherein obtaining (120) comprises configuring (122) the weights per sub-carrier, configuring (124) the weights per resource block, configuring (126) the weights per frequency range (ak), and / or configuring (128) the weights perTX port (Q).

10. The method of any one of claims 1-9, wherein obtaining (120) comprises determining (117) at a first time instant (t) the two or more channel estimate matrices (Hfe(t)) and determining (118) at a second time instant (t-r) the two or more channel estimate matrices (Hk(t — T)) and wherein the function (F) is a function of the two or more channel estimate matrices at the first time instant squared (Hk(t)Hk(t)) and the two or more channel estimate matrices at the second time instant squared (Hk(t — )Hk(t — T)).

11. A computer program product comprising instructions, which, when executed on one or more processors of a processing device, cause the processing device to carry out the method according to any one of claims 1-10.

12. A non-transitory computer-readable storage medium storing one or more programs configured to be executed by one or more processors of a processing device, the one or more programs comprising instructions which, when executed by the processing device, causes the processing device to carry out the method according to any one of claims 1-10.

13. A processing unit (300) configured to cause: obtainment (310) of two or more channel estimate matrices (Hl, H2, ..., HK) associated with one or more propagation channels for FDM signals; application (320) of a function (F) to the two or more channel estimate matrices to obtain a resulting matrix (RM), the resulting matrix (RM) resulting from the application (320); matrix decomposition (330) of the resulting matrix (RM) into a first decomposition matrix (U), comprising first vectors of coefficients (Ul, U2, ..., UN), and a second decomposition matrix (A), different from the first decomposition matrix (U), comprising second vectors of coefficients (Al, A2, ..., AN); determination (340) of a vector of spatial filter coefficients (Tl, T2, ..., TN) from the first vectors of coefficients (Ul, U2, ..., UN); selection (345) of a subset (Tl, ..., TM) of the vectors of spatial filter coefficients (Tl, T2, ..., TN), wherein the subset (Tl, ..., TM) is associated with one or more of the four most dominant eigenvalues of the second decomposition matrix (U); and conversion (352) of a plurality (NRX) of digital signals into a plurality (NS) of virtual antenna streams utilizing (351) only the subset (Tl, ..., TM).

14. The processing unit of claim 13, wherein the function (F) is a quadratic function.

15. The processing unit of any one of claims 13-14, wherein the number of virtual antenna streams is larger than a number of transmit antenna ports of a transmitting device utilized for transmitting the plurality of FDM signals.

16. A wireless device, WD, (302) comprising the processing unit (300) of any one of claims13-15.

17. A chip (990) comprising the processing unit (300) of any one of claims 13-15.

Images