Cyclic prefix in beam domain DPD for multi-antenna architecture

Abstract

Various embodiments provide for a method and multi-antenna architecture for implementing beam-domain Digital Pre-Distortion (DPD) that utilizes Cyclic Prefix (CP) Orthogonal Frequency Division Multiplexing (OFDM) in a wireless communication system. In an embodiment, the transmitter can receive an input baseband signal matrix that is in a frequency domain and then convert the input baseband signal matrix to a time domain signal matrix in a time domain. The transmitter can then add a CP to the time domain signal matrix and perform beam-domain DPD to the time domain signal matrix with the CP and then perform either non-frequency selective precoding or frequency selective precoding to an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain before transmitting a transmission based on the precoded signal matrix.

Description

CYCLIC PREFIX IN BEAM DOMAIN DPD FOR MULTI-ANTENNA ARCHITECTURE Technical Field

[0001] The present disclosure relates to a method and multi-antenna architecture for implementing beam-domain Digital Pre-Distortion (DPD) that utilizes Cyclic Prefix (CP) Orthogonal Frequency Division Multiplexing (OFDM) in a wireless communication system. Background

[0002] With the recent trend towards large antenna arrays, the number of radio frequency (RF) branches present in a radio is increasing at a rapid rate. Each RF branch comes with a cost and one of the main resource intensive components is the digital pre- distortion (DPD) unit used to linearize the power amplifier / s connected to each RF branch.

[0003] Figure 1 illustrates an Orthogonal Frequency Division Multiplexing (OFDM) based RF branch in a state-of-art antenna array systems where precoding at block 102 is done in frequency domain on L layers (or beams) and is transformed into signal for Bbranches (e.g., branch 104) where DPD (denoted by operation ^^(. ) for the bth branch)at block 110 is performed after conversion to time domain by Inverse Fast Fourier Transform (IFFT) at block 106 (denoted by matrix multiplication by ^^), and cyclic- prefix addition (denoted by matrix multiplication by ^^^) at block 108. Note that crest- factor reduction (CFR) in the system has been omitted which would come before DPD for sake of simplicity. Although Figure 1 is the usual way of illustrating similar systems with independent branches, going forward it can be illustrated in a more compact way as shown in Figure 2. Most operations will be denoted by left -side matrix multiplication except precoding which will be right-sided matrix multiplication, and DPD which being non-linear operation cannot be compacted conveniently in a matrix multiplication.

[0004] In general, the number of layers (^) is lower than the number of branches(^). Typically, in most commercial deployments ^ = 8, whereas the number ofbranches can be 32 and is increasing to achieve more range and capacity. Hence the logical way of reducing complexity is to move the digital processing (like CFR and DPD) to the layer domain thereby reducing the cost by^^ times. Doing so also requires thetime domain components to be moved to frequency domain as shown in A. Brihuega, L. Anttila and M. Valkama, "Beam-Level Frequency-Domain Digital Predistortion for OFDM Massive MIMO Transmitters," in IEEE Transactions on Microwave Theory and Techniques, vol. 71, no. 4, pp. 1412-1427, April 2023, doi: 10.1109 / TMTT.2022.3222320 and Y. Wu, U. Gustavsson, M. Valkama, A. G. i Amat and H. Wymeersch, "Frequency-domain digital predistortion for Massive MU-MIMO-OFDM Downlink," GLOBECOM 2022 - 2022 IEEE Global Communications Conference, Rio de Janeiro, Brazil, 2022, pp. 579-584, doi: 10.1109 / GLOBECOM48099.2022.10001382.

[0005] The symbols shown in Figures 1 and 2, and throughout the application areshown in Table 1 below:^⃗ Input baseband signal matrix of size ^ × ^, where ^ is the size ofsystem FFT and ^ is number of layers to be transmitted over ^ RF- branches. Usually, layers are associated with directions in antenna-arrays^ Precoding matrix of size ^ × ^. Note that precoding is done withright side matrix multiplication, i.e., ^⃗^ is the output of precoding.^^^ Extended precoding matrix of size ^^^ × ^ and includes extra rowsto handle the IM beams. As with^ precoding is done with right side matrix multiplication, i.e.,^^⃗^^ is the output of precoding. Also^^^ = ^ + ^^^.^^ Hermitian transpose of DFT-matrix ^ of size ^ × ^. Operation isleft sided matrix multiplication. Typically, the operation is carried out with FFT.^^^ Addition of cyclic prefix of size ^ denoted by multiplication bymatrix of size ^^^ × ^ and is of the form:^^^ = ^^^×( !^) ^^^ " where ^^×( !^) is zero matrix of size ^ × (^ −^) and square identity matrices are denoted by ^ with is sizedenoted in subscript. Note that ^^^ = ^ + ^ and operation is leftsided matrix multiplication^(. ) DPD operation of its input signal.$⃗ System output of size ^^^ × ^, denoting ^^^ samples per branch.TABLE 1 Summary

[0006] Various embodiments provide for a method and multi-antenna architecture for implementing beam-domain Digital Pre-Distortion (DPD) that utilizes Cyclic Prefix (CP) Orthogonal Frequency Division Multiplexing (OFDM) in a wireless communication system. In an embodiment, the transmitter can receive an input baseband signal matrix that is in a frequency domain and then convert the input baseband signal matrix to a time domain signal matrix in a time domain. The transmitter can then add a CP to the time domain signal matrix and perform beam-domain DPD to the time domain signal matrix with the CP and then perform either non-frequency selective precoding or frequency selective precoding to an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain before transmitting a transmission based on the precoded signal matrix.

[0007] In an embodiment, a method for implementing a beam-domain DPD multi- antenna architecture that utilizes CP OFDM in a multi-antenna transmitter device is proposed. The method includes receiving an input baseband signal matrix that is in a frequency domain. The method also includes converting the input baseband signal matrix to a time domain signal matrix in a time domain, and adding a CP to the time domain signal matrix. The method also includes performing beam-domain DPD to the time domain signal matrix with the CP. The method includes non-frequency selective precoding or frequency selective precoding an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain and transmitting a transmission based on the precoded signal matrix.

[0008] In an embodiment, there is a respective beam-domain DPD block associated with each intermodulation beam.

[0009] In an embodiment, the precoding is non-frequency selective precoding performed by a single precoding block.

[0010] In an embodiment, the transmission is in a single frequency sub-band.

[0011] In an embodiment, the precoding is frequency selective precoding performed by a plurality of precoding blocks.

[0012] In an embodiment, the transmission is in a plurality of frequency sub-bands.

[0013] In an embodiment, the input baseband signal matrix comprises a number of input streams based on a product of a number of layers of the transmission and a number of frequency sub-bands wherein for each input stream, there is a respective precoded stream that is stacked with other precoded streams.

[0014] In an embodiment, there is a respective beam-domain DPD block associated with each input stream of the input baseband signal matrix.

[0015] In an embodiment, the method also includes performing additional signal processing to the precoded signal matrix before transmitting the transmission.

[0016] In an embodiment, a multi-antenna transmitter device is configured to implement a beam-domain DPD architecture that utilizes CP-OFDM where the transmitter device includes a radio interface and processing circuitry configured to receive an input baseband signal matrix that is in a frequency domain. The processing circuitry is also configured to convert the input baseband signal matrix to a time domain signal matrix in a time domain, and adding a CP to the time domain signal matrix. The processing circuitry is also configured to perform beam-domain DPD to the time domain signal matrix with the CP. The processing circuitry is also configured to non-frequency selective precode or frequency selective precode an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain and transmit a transmission based on the precoded signal matrix.

[0017] In an embodiment, a computer program is provided that includes instructions, which, when executed on at least one processor, cause the processor to carry out the method of the above embodiments. A carrier can also be provided that contains the computer program, wherein the carrier is one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium.

[0018] Some of the advantages of the proposed techniques disclosed herein are that the CP will be added correctly to the output signal and hence will improve the end-to- end network performance under dispersive channel conditions while maintaining low enough adjacent channel leakage ratio (ACLR) emissions. Also, the performance of the system has the same performance as branch-domain DPD which is the current standard.Brief Description of the Drawings

[0019] The accompanying drawing figures incorporated in and forming a part of this specification illustrateof the disclosure, and together with the description serve to explain the principles of the disclosure.

[0020] Figure 1 illustrates an example of branch-domain Digital Pre-Distortion (DPD) in a multi-antenna system according to some embodiments of the present disclosure;

[0021] Figure 2 illustrates another way to illustrate the example of branch-domain DPD of Figure 1 according to some embodiments of the present disclosure;

[0022] Figure 3 illustrates an example of beam-domain DPD with Cyclic Prefix (CP) addition in a multi-antenna system according to some embodiments of the present disclosure;

[0023] Figure 4 illustrates another way to illustrate the example of beam-domain DPD of Figure 3 according to some embodiments of the present disclosure;

[0024] Figure 5 illustrates a simplified beam-domain DPD structure with CP addition according to some embodiments of the present disclosure;

[0025] Figure 6 illustrates a branch-domain DPD structure with frequency selective precoding according to some embodiments of the present disclosure;

[0026] Figure 7 illustrates a beam-domain DPD structure with frequency selective precoding according to some embodiments of the present disclosure;

[0027] Figure 8 illustrates a sample basic architecture of Multi-Dimension Reduction (MDR) according to some embodiments of the present disclosure;

[0028] Figure 9 illustrates a flow chart of a method for implementing a beam-domain DPD multi-antenna architecture that utilizes CP Orthogonal Frequency Division Multiplexing (OFDM) in a transmitter device according to some embodiments of the present disclosure;

[0029] Figure 10 illustrates one example of a cellular communications system according to some embodiments of the present disclosure;

[0030] Figure 11 is a schematic block diagram of a radio access node according to some embodiments of the present disclosure; and

[0031] Figure 12 is a schematic block diagram of the radio access node of Figure 11 according to some other embodiments of the present disclosure.Detailed Description

[0032] The embodiments set forth below represent information to enable those skilled in the art to practice the embodiments and illustrate the best mode of practicing the embodiments. Upon reading the following description in light of the accompanying drawing figures, those skilled in the art will understand the concepts of the disclosure and will recognize applications of these concepts not particularly addressed herein. It should be understood that these concepts and applications fall within the scope of the disclosure.

[0033] Transmitter device: As used herein, a “transmitter device” could be either a radio access node or a wireless communication device, or any other multi-antenna transmitter in a wireless communication system that can employ Cyclic Prefix (CP addition) Orthogonal Frequency Division Multiplexing (OFDM).

[0034] Note that the description given herein focuses on a Third Generation Partnership Project (3GPP) cellular communications system and, as such, 3GPP terminology or terminology similar to 3GPP terminology is oftentimes used. However, the concepts disclosed herein are not limited to a 3GPP system.

[0035] Note that, in the description herein, reference may be made to the term “layer”; however, in this document, beams may be used instead of layers and, as such, it is important to note that the concepts described herein are equally applicable to both layers and beams.

[0036] Some of the challenges faced by the existing technology is that no current system deal with the cyclic prefix (CP) which is integral part of any OFDM system. As an example, in some embodiments, the CP is added after final inverse discrete Fourier transform (IDFT), but the CP signal is not correctly pre-distorted since the Digital Pre- Distortion (DPD) operation before the precoding has no knowledge of the CP and its temporal position for which it gets distorted differently than the main block of data due to DPD memory taps. This will only work if the DPD (hence the Power Amplifier (PA)) is memoryless which is unlikely. Hence the system may only be approximate to what is desired. Addition of CP after DPD will not only cause out-of-band emission but also degrade network capacity under dispersive channel conditions where any OFDM system uses CP to mitigate inter-symbol interference as in W. Henkel, G. Taubock, P. Odling, P. O. Borjesson and N. Petersson, "The cyclic prefix of OFDM / DMT - an analysis," 2002 International Zurich Seminar on Broadband Communications Access - Transmission -Networking (Cat. No.02TH8599), Zurich, Switzerland, 2002, pp. 22-22, doi: 10.1109 / IZSBC.2002.991762.

[0037] To overcome these and other challenges, proposed is a method to correctly add the CP for beam-domain DPD architectures and DPD structure for frequency selective precoding using sub-banding techniques. The proposed architecture incorporates CP that can handle DPD memory taps while satisfying 3GPP CP related requirements. The proposed architecture can support both single precoding (non- frequency selective precoding) as well as frequency selective precoding. Some of the advantages of the proposed techniques disclosed herein are that the CP will be added correctly to the output signal and hence will improve the end-to-end network performance under dispersive channel conditions while maintaining low enough adjacent channel leakage ratio (ACLR) emissions. Also, the performance of the system has the same performance as branch-domain DPD which is the current standard.

[0038] Various embodiments provide for a method and multi-antenna architecture for implementing beam-domain DPD that utilizes CP OFDM in a wireless communication system. In an embodiment, the transmitter can receive an input baseband signal matrix that is in a frequency domain and then convert the input baseband signal matrix to a time domain signal matrix in a time domain. The transmitter can then add a CP to the time domain signal matrix and perform beam-domain DPD to the time domain signal matrix with the CP and then perform either non-frequency selective precoding or frequency selective precoding to an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain before transmitting a transmission based on the precoded signal matrix.

[0039] The present disclosure includes a simple structure to handle the CP. The disclosure first illustrates a narrowband system where it is assumed that there is only one precoder for the entire band of operation. For systems with frequency-selective precoding a separate proposal is handled next. A Beam-Domain DPD architecture is proposed for frequency-selective precoding which gives insights when Beam-Domain DPD is advantageous over Branch-Domain DPD. The focus of this disclosure is only handling CP and performing DPD in beam-domain for non-frequency and frequency- selective beamforming and hence other usual operations in radio line-up like crest-factor reduction (CFR) or channel-filtering is not discussed.

[0040] In a first embodiment, a beam-domain DPD is proposed using a single precoder. This can be illustrated via a block diagram schematic in Figure 3 that illustrates an example of beam-domain DPD with CP addition in a multi-antenna system according to some embodiments of the present disclosure.

[0041] Here beam domain DPD receives frequency domain input samples, ^⃗ (302) before the precoding. Although this is called frequency domain DPD 304, the actual non-linear distortion is modelled in the time domain. Hence ^⃗ 302 is passed through the first ^-point Inverse Fast Fourier Transform (IFFT) 306 to convert into time domain samples for the DPD. In addition, we add the CP at 308 so that DPD is done at block 310 on actual signal which will be on air and the CP is correctly distorted in presence of memory taps in the power amplifier (PA). In Figure 3, the CP add at 308 can be performed before the DPD at 310. The output of the DPD 310 is divided into Main-path and CP-path at 312 and treated separately, i.e., frequency domain precoded at 314 and 316 with ^^^followed the IFFT. Here, precoding is done on the frequency domain signals and CP is treated separately (e.g., at 316) to preserve their correct position in time domain signals. Since the CP is treated separately, it is stacked at the end at 318 to produce the output $⃗.

[0042] One notable difference from branch DPD structure is that the precoding matrix is now extended to deal with inter-modulated (IM) beams and hence its size isincrease to ^^^ × ^, where ^^^ = ^ + ^^^.The number of IM-beams is design dependentand usually the 3rdorder beams are most strong. It can be shown that number of 3rdorder IM beams = ^% (^ − 1)⁄ 2. The Beam-Domain DPD gets input data with CP-addedfrom ^ independent data streams but combines them to produce single output corresponding to either main or IM beams. Hence there are ^^^DPDs and each is ^-input-1-output. An example for ^ = 2 is given below.

[0043] In an embodiment, let )*and )%represent two input streams. Here input streams can be considered as two different users, or a single user receives two streamsof data using two different precoder. Let +⃗ and - ,⃗ to represent correspondingbeamforming (precoding) vectors where +⃗,- ,⃗ ∈ ℂ*×^ and there are ^ branches. Forprecoder - ,⃗ , the array-factor 23(-,⃗ , 4⃗) in the direction of radiation (or arrival) 4⃗ can bewritten as: 235-,⃗ , 4⃗6 = - ,⃗ .789:⃗.;⃗, where < = %=> , and ? being the wavelength of the radiofrequency (RF) signal. Here, vector @⃗ comprises positions of radiating elements and ′. ′indicates a scalar product.

[0044] One can extend above mentioned equations for other precoders as well. The received (or transmitted) signal at point 4⃗, i.e., ^(4⃗), will be superposition of both precoders and hence can be written as in Eqn.1.

[0045] ^(4⃗) = +⃗) .789:⃗.;⃗ -⃗ 89:⃗.;⃗* + , )%.7 Eqn.1polynomial andcan as in Eqn.2 and which canbe expanded to Eqn.3 by using CD = +D)* + ,D)% where +Dand ,D are B-th elements ofvectors +⃗ and - ,⃗ respectively:

[0047] ^2D(CD) = E*CD + EFCD|CD|% Eqn.2

[0048] ^2(C) = +) (E + E |+|%|) |% + 2E | |%| |%) ( | |%| |%D D D * * F D * F ,D )% + ,D)% E* + EF ,D )% +∗ ∗ ∗ ∗above-mentionedcreated four directions where +Dand ,Drepresenting the main desired beam directions. +D∗,D%and ,D∗+D%are IM3 beams. These two IM3 beams are created due to the non-linearity at the PA and they can have different directions from the main beams. To compensate, PA with IM3 terms, one can construct DPDs as follows.

[0050] I^IJ-⃗ = )* 5K*,J-⃗ + KF,J-⃗ |)*|% + 2KF,J-⃗ |)%|%6 which is multiplied by +⃗ duringbeamforming. I^I %|) | which is multiplied by- ,⃗ duringbeamforming.DPD

[0051] Similarly, for IM3s,J-⃗ ∗- ^⃗ L = KF, J- ∗- ^⃗ L)∗*)%% which is multiplied by +∗⃗- ,⃗ %.Similarly, I^I ∗ %^^F,- ^⃗ ∗ J-⃗ L = KF,- ^⃗ ∗ J-⃗ L)%)*by - ,⃗ ∗ +⃗%

[0052] One can utilize different implementation architecture for these DPDs such as multi-dimension look up tables (LUT), such as multi dimension reduction (MDR) based LUTs as well as machine learning. In general, there will be ^^^single output DPDs, each having ^ inputs or ^-dimensional function e.g., polynomial, multi-dimensional LUT or MDR kind.

[0053] The embodiment depicted in Figure 3 is similar to that commonly used and discussed in literature, albeit without CP-addition, however, the representation can besimplified by using the standard properties of matrices. To do this let’s formulate the separation of main and CP signal after DPD at 402 as matrix multiplication and regrouping as shown in Figure 4 along with the stacking later after IFFT as another matrix operation and addition.

[0054] It can be easily shown that Figures 3 and 4 are equivalent if we use thefollowing values of the newly introduced simple matrices as in Eqn. 4.MN = O0 ×^ ^ QM ^^ 0^×^^ = ^ 0( !^)×( R^) "4

[0055] Moreover, one can simplify each branch by noting the DFT-matrices ^ and ^^are unitary matrices and hence ^^^ = ^. We also use T⃗NJDU = ^MNT⃗ in Eqn. 5 andT⃗^^ = ^M^^T⃗ in Eqn. 6:567S+ S^^M^^ = ^ ^^ 0^×^ "Eqn. 8

[0061] Hence using Eqn. 8, Eqn. 7 can be reduced to Eqn. 9.

[0062] $⃗ = T⃗^^^ Eqn. 9

[0063] Which means that the elaborate structure shown in Figure 3 and Figure 4 is not necessary and can be as simple as that depicted in Figure 5.

[0064] Figure 5 illustrates a simplified beam-domain DPD structure with CP addition for non-frequency selective precoding according to some embodiments of the present disclosure.

[0065] The input ^⃗ 502, which can be an input baseband signal matrix can be in the frequency domain and then converted to the time domain via IFFT block 504. The time domain signal matrix can then have the CP added at block 506, and then the time domain signal matrix with the CP can be beam-domain DPD at block 508, with the output 510 of the beam-domain DPD precoded at block 512. The precoded signal matrix 514 can then be further processed before being transmitted. The simplified structure is possible because of the single precoder used for the entire band.

[0066] For frequency selective precoding, when the precoders are different for different parts of the band, the different precoders can be either due to different user directions or different frequency values in the entire band but in general different parts of the band will be associated with different precoders. These parts can be called sub- bands and each sub-band has equal size V. To handle frequency-selectivity inprecoding we introduce the following symbols in table 2:^^⃗ Data matrix of size V × ^ pertaining to sub-band W. There are Vsamples from ^ layers for each sub-band. Since the FFT size was^ and hence there are ^ = ^ / V such sub-bands, W ∈ O0, ^ − 1Q.^⃗ Block diagonal matrix of size ^ × ^^, where diagonal entries aremade from ^^⃗. ^Z⃗ ⋯ 0^⃗ = @B+Y5^Z⃗, ^*⃗, … , ^\⃗!*6 = ] ⋮ ⋱ ⋮ a0 ⋯ ^\⃗!*^^ Precoding matrix of size ^ × ^ for the sub-band W. Note thatprecoding is done with right side matrix multiplication, i.e., ^^⃗^^is the output of precoding for sub-band W.^ Complete precoding matrix of size ^^ × ^ made by horizontallystacking the sub-band precoders ^^.^Z^= b^*⋮ c^\!*^^^ Extended precoding matrix of size ^^^ × ^ and includes extra rowsto handle the IM beams, hence^^^ = ^^ + ^^^.Table 2

[0067] The branch domain structure shown in Figure 2 still valid with the above symbols and is replicated in Figure 6 with the correct dimensions of ^⃗ and ^

[0068] Figure 6 illustrates a branch-domain DPD structure with frequency selective precoding according to some embodiments of the present disclosure.

[0069] Using this sub-banding it can be shown in same way as in previous section that the architecture can be reduced to almost same as Figure 5 but with more complexity in DPD. This is shown in Figure 7 which illustrates a beam-domain DPD structure with frequency selective precoding with the correct dimensions.

[0070] The complexity of DPD now has increased ^ fold because there are ^^ dimensions for the DPD block at 508 instead of the ^ was in case of non-frequencyselective precoder in Figure 5. As an example, with ^ = 2 and ^ = 2, the number ofIM3s will be^L\L(^\!*)% or 24. Hence there will ^^ or 4 main beams and 24 IM3 beams, thereby total number of DPDs will be 28 and in the original precoder matrix ^, another 24 rows must be added to cater for the IM3 beams.

[0071] The concept of beam domain DPD for sub-band is same as explained earlier. The main difference is that now we have L*P different input streams instead of L. Since there are more input steams, there will be more IM terms. These IMs can be created by same L but with different P as well as with different L. Figure 8 shows basic architecture of MDR. As can be seen there are N LUTs for each of the DPD. There are L*P DPDs which we call main beam DPD. There are DPDs for IMs as well. Please note that number of N LUTs varies for different L, P as well as for different IM terms. Typically, IM3 towards non-user direction requires few LUTs and also can be totally removed depending on the requirements.

[0072] Finally, each DPD is multiplied by their corresponding input signals before sending to the precoding block. One can follow a similar approach to construct any IM terms. As mentioned earlier one needs to also extend precoding matrix for the IMs.

[0073] Figure 9 illustrates a flow chart of a method for implementing a beam-domain DPD multi-antenna architecture that utilizes CP OFDM in a transmitter device according to some embodiments of the present disclosure.

[0074] At 902, the method includes receiving an input baseband signal matrix that is in a frequency domain. In an embodiment, the input baseband signal matrix comprises a number of input streams based on a product of a number of layers of the transmission and a number of frequency sub-bands wherein for each input stream, there is a respective precoded stream that is stacked with other precoded streams.

[0075] At 904, the method includes converting the input baseband signal matrix to a time domain signal matrix in a time domain.

[0076] At 906, the method includes adding a CP to the time domain signal matrix.

[0077] At 908, the method includes performing beam-domain DPD to the time domain signal matrix with the CP. In an embodiment, there is a respective beam- domain DPD block (e.g., block 508 in Figure 5 or 7) associated with each input stream of the input baseband signal matrix.

[0078] At 910, the method includes non-frequency selective precoding or frequency selective precoding an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain.

[0079] In an embodiment, there is a respective beam-domain DPD block associated with each intermodulation beam.

[0080] In an embodiment, the precoding could be non-frequency selective precoding as shown in block 512 in Figure 5. In such an example, there could be a single precoding block. In other embodiments however, the precoding could be frequency selective precoding performed by a plurality of precoding blocks, as illustrated by precoding block 512 in Figure 7.

[0081] At 912, the method optionally includes performing additional signal processing to the precoded signal matrix before transmitting the transmission. The additional signal processing could include stacking the outputs of the precoders at the end (as seen in 318 in Figure 3) to produce the output $.⃗ Other processing could includenormal transmit chain processing that is well known in the art, including power amplification, etc.

[0082] At 914, the method includes transmitting a transmission based on the precoded signal matrix. In an embodiment where the precoding performed is frequency selective precoding as depicted in Figure 7, the transmission could be in a plurality of frequency sub-bands.

[0083] Figure 10 illustrates one example of a cellular communications system 1000 in which embodiments of the present disclosure may be implemented. In the embodiments described herein, the cellular communications system 1000 is a 5G system (5GS) including a Next Generation RAN (NG-RAN). In this example, the RAN includes base stations 1002-1 and 1002-2, which in the 5GS include NR base stations (gNBs), controlling corresponding (macro) cells 1004-1 and 1004-2. The base stations 1002-1 and 1002-2 are generally referred to herein collectively as base stations 1002 and individually as base station 1002. Likewise, the (macro) cells 1004-1 and 1004-2 are generally referred to herein collectively as (macro) cells 1004 and individually as (macro) cell 1004. The RAN may also include a number of low power nodes 1006-1 through 1006-4 controlling corresponding small cells 1008-1 through 1008-4. The low power nodes 1006-1 through 1006-4 can be small base stations (such as pico or femto base stations) or RRHs, or the like. Notably, while not illustrated, one or more of the small cells 1008-1 through 1008-4 may alternatively be provided by the base stations 1002. The low power nodes 1006-1 through 1006-4 are generally referred to herein collectively as low power nodes 1006 and individually as low power node 1006. Likewise, the small cells 1008-1 through 1008-4 are generally referred to herein collectively as small cells 1008 and individually as small cell 1008. The cellular communications system 1000 also includes a core network 1010, which in the 5GS is referred to as the 5GC. The base stations 1002 (and optionally the low power nodes 1006) are connected to the core network 1010.

[0084] The base stations 1002 and the low power nodes 1006 provide service to wireless communication devices 1012-1 through 1012-5 in the corresponding cells 1004 and 1008. The wireless communication devices 1012-1 through 1012-5 are generally referred to herein collectively as wireless communication devices 1012 and individually as wireless communication device 1012. In the following description, the wireless communication devices 1012 are oftentimes UEs, but the present disclosure is notlimited thereto. The multi-antenna transmitter device (multi-antenna transmitter device 1000) could be either a base station 1002 low power node 1006, or wireless communication device 1012.

[0085] Figure 11 is a schematic block diagram of a multi-antenna transmitter device 1100 that performs the method described in Figure 9 and depicted in Figures 5 and 7 according to some embodiments of the present disclosure. Optional features are represented by dashed boxes. The multi-antenna transmitter device 1100 may be, for example, a base station 1002 or 1006 or a network node that implements all or part of the functionality of the base station 1002 or gNB described herein. As illustrated, the multi-antenna transmitter device 1100 includes a control system 1102 that includes one or more processors 1104 (e.g., Central Processing Units (CPUs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs), and / or the like), memory 1106, and a network interface 1108. The one or more processors 1104 are also referred to herein as processing circuitry. In addition, the multi-antenna transmitter device 1100 may include one or more radio units 1110 that each includes one or more transmitters 1112 and one or more receivers 1114 coupled to one or more antennas 1116. The radio units 1110 may be referred to or be part of radio interface circuitry. In some embodiments, the radio unit(s) 1110 is external to the control system 1102 and connected to the control system 1102 via, e.g., a wired connection (e.g., an optical cable). However, in some other embodiments, the radio unit(s) 1110 and potentially the antenna(s) 1116 are integrated together with the control system 1102. The one or more processors 1104 operate to provide one or more functions of a multi- antenna transmitter device 1100 as described herein. In some embodiments, the function(s) are implemented in software that is stored, e.g., in the memory 1106 and executed by the one or more processors 1104.

[0086] In some embodiments, a computer program including instructions which, when executed by at least one processor, causes the at least one processor to carry out the functionality of multi-antenna transmitter device 1100 or a node implementing one or more of the functions of the multi-antenna transmitter device 1100 according to any of the embodiments described herein is provided. In some embodiments, a carrier comprising the aforementioned computer program product is provided. The carrier is one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium (e.g., a non-transitory computer readable medium such as memory).

[0087] Figure 12 is a schematic block diagram of the multi-antenna transmitter device 1100 according to some other embodiments of the present disclosure. The multi-antenna transmitter device 1100 includes one or more modules 1200, each of which is implemented in software. The module(s) 1200 provide the functionality of the multi-antenna transmitter device 1100 described herein.

[0088] Any appropriate steps, methods, features, functions, or benefits disclosed herein may be performed through one or more functional units or modules of one or more virtual apparatuses. Each virtual apparatus may comprise a number of these functional units. These functional units may be implemented via processing circuitry, which may include one or more microprocessor or microcontrollers, as well as other digital hardware, which may include Digital Signal Processors (DSPs), special-purpose digital logic, and the like. The processing circuitry may be configured to execute program code stored in memory, which may include one or several types of memory such as Read Only Memory (ROM), Random Access Memory (RAM), cache memory, flash memory devices, optical storage devices, etc. Program code stored in memory includes program instructions for executing one or more telecommunications and / or data communications protocols as well as instructions for carrying out one or more of the techniques described herein. In some implementations, the processing circuitry may be used to cause the respective functional unit to perform corresponding functions according to one or more embodiments of the present disclosure.

[0089] While processes in the figures may show a particular order of operations performed by certain embodiments of the present disclosure, it should be understood that such order is exemplary (e.g., alternative embodiments may perform the operations in a different order, combine certain operations, overlap certain operations, etc.).

[0090] Those skilled in the art will recognize improvements and modifications to the embodiments of the present disclosure. All such improvements and modifications are considered within the scope of the concepts disclosed herein.

Claims

Claims 1. A method for implementing a beam-domain Digital Pre-Distortion, DPD, multi- antenna architecture that utilizes Cyclic Prefix, CP, Orthogonal Frequency Division Multiplexing, OFDM, in a multi-antenna transmitter device (1100), the method comprising: receiving (902) an input baseband signal matrix that is in a frequency domain; converting (904) the input baseband signal matrix to a time domain signal matrix in a time domain; adding (906) a CP to the time domain signal matrix; performing (908) beam-domain DPD to the time domain signal matrix with the CP; non-frequency selective precoding or frequency selective precoding (910) an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain; and transmitting (914) a transmission based on the precoded signal matrix.

2. The method of claim 1, wherein there is a respective beam-domain DPD block associated with each intermodulation beam.

3. The method of any of claims 1 to 2, wherein the precoding is non-frequency selective precoding performed by a single precoding block.

4. The method of claim 3, wherein the transmission is in a single frequency sub- band.

5. The method of any of claims 1 to 2, wherein the precoding is frequency selective precoding performed by a plurality of precoding blocks.

6. The method of claim 5, wherein the transmission is in a plurality of frequency sub-bands.

7. The method of any of claims 5 to 6, wherein the input baseband signal matrix comprises a number of input streams based on a product of a number of layers of the transmission and a number of frequency sub-bands wherein for each input stream, there is a respective precoded stream that is stacked with other precoded streams.

8. The method of any of claims 2 to 7, wherein there is a respective beam-domain DPD block associated with each input stream of the input baseband signal matrix.

9. The method of any of claims 1 to 8, further comprising: performing (912) additional signal processing to the precoded signal matrix before transmitting the transmission.

10. A multi-antenna transmitter device (1100) configured to implement a beam- domain Digital Pre-Distortion, DPD, architecture that utilizes Cyclic Prefix, CP, Orthogonal Frequency Division Multiplexing, OFDM, the transmitter device (1000) comprising a radio interface and processing circuitry configured to: receive (902) an input baseband signal matrix that is in a frequency domain; convert (904) the input baseband signal matrix to a time domain signal matrix in a time domain; add (906) a CP to the time domain signal matrix; perform (908) beam-domain DPD, to the time domain signal matrix with the CP; non-frequency selective precode or frequency selective precode (910) an output of the beam-domain DPD resulting in a precoded signal matrix that is in the time domain; and transmit (914) a transmission based on the precoded signal matrix.

11. The multi-antenna transmitter device (1100) of claim 10, wherein there is a respective beam-domain DPD block associated with each intermodulation beam.

12. The multi-antenna transmitter device (1100) of any of claims 10 to 11, wherein the precoding is non-frequency selective precoding performed by a single precoding block.

13. The multi-antenna transmitter device (1100) of claim 12, wherein the transmission is in a single frequency sub-band.

14. The multi-antenna transmitter device (1100) of any of claims 10 to 11, wherein the precoding is frequency selective precoding performed by a plurality of precoding blocks.

15. The multi-antenna transmitter device (1100) of claim 14, wherein the transmission is in a plurality of frequency sub-bands.

16. The multi-antenna transmitter device (1100) of any of claims 14 to 15, wherein the input baseband signal matrix comprises a number of input streams based on a product of a number of layers of the transmission and a number of frequency sub-bands wherein for each input stream, there is a respective precoded stream that is stacked with other precoded streams.

17. The multi-antenna transmitter device (1100) of any of claims 11 to 16, wherein there is a respective beam-domain DPD block associated with each input stream of the input baseband signal matrix.

18. The multi-antenna transmitter device (1100) of any of claims 10 to 17, wherein the processing circuitry is further configured to: perform (912) additional signal processing to the precoded signal matrix before transmitting the transmission.

19. A computer program comprising instructions which, when executed on at least one processor, cause the processor to carry out the method according to any of claims 1 to 9.

20. A carrier containing the computer program of claim 19, wherein the carrier is one of an electronic signal, an optical signal, a radio signal, or a computer readable storage medium.

Images