Distributed and scalable dowlink precoding and transmission in multi-user extreme MIMO systems

Abstract

An apparatus may be configured to estimate uplink channel based on reference signals from one or more user equipment; compute a precoding matrix based on the estimated uplink channel for the one or more user equipment; obtain, for the one or more user equipment, a set of pairs of complex vectors and values indicative of strength of effective channel between the first apparatus and the respective user equipment based on the computed precoding matrix; transmit, to a central processor, the obtained one or more sets of the values indicative of strength of effective channel for the one or more user equipment; receive, from the central processor, an indication of which values from the transmitted one or more sets of values are determined to be used for transmit beamforming for the one or more user equipment; and determine, based on the received indication, one or more associated complex vectors of the precoding matrix to be used for transmit beamforming for the one or more user equipment.

Description

DISTRIBUTED AND SCALABLE DOWLINK PRECODING AND TRANSMISSION IN MULTI-USER EXTREME MIMO SYSTEMS TECHNICAL FIELD

[0001] Various example embodiments generally relate to the field of wirelesscommunication. Some example embodiments relate to distributed and scalable downlink precoding and transmission methods in multi-user extreme MIMO systems. BACKGROUND

[0002] Some communication systems, such as 6G or 5G+, may enable a few times fasterdata rates than 5G. Therefore, a base station, such as a gNB, may be equipped with a higher number of antenna elements. For example, in extreme MIMO (multiple input, multiple output) system, the number of antenna elements may be expected to increase to 512-1024 compared to around 200 used for 5G. This may also necessitate a larger number of transceivers (TRX), wherein the number of TRX may increase to 256-512 compared to 32- 64 in 5G. In addition, the frequency band of interest is expected to be 7-15 GHz. Further, multiple user equipment (UE), for example in the range of 8-16, are expected to be co- scheduled under peak traffic conditions.

[0003] The larger number of TRXs may be accommodated within a same-sized antennaarray as in 5G by shifting to a larger carrier frequency. However, as the number of TRXs and co-scheduled UEs grows, so does the complexity of the signal processing at the gNB receiver. SUMMARY

[0004] This summary is provided to introduce a selection of concepts in a simplified formthat are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

[0005] Example embodiments of the present disclosure enable to provide a lowercomputational complexity of signal processing at a gNB receiver with distributed design. This and other benefits may be achieved by the features of the independent claims. Further example embodiments are provided in the dependent claims, the description, and the drawings.

[0006] According to a first aspect, an apparatus is disclosed. The apparatus may comprise:at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to: estimate uplink channel based on reference signals from one or more user equipment; compute a precoding matrix based on the estimated uplink channel for the one or more user equipment; obtain, for the one or more user equipment, a set of pairs of complex vectors and values indicative of strength of effective channel between the first apparatus and the respective user equipment based on the computed precoding matrix; transmit, to a central processor, the obtained one or more sets of the values indicative of strength of effective channel for the one or more user equipment; receive, from the central processor, an indication of which values from the transmitted one or more sets of values are determined to be used for transmit beamforming for the one or more user equipment; and determine, based on the received indication, one or more associated complex vectors of the precoding matrix to be used for transmit beamforming for the one or more user equipment.

[0007] According to an example embodiment of the first aspect, the apparatus is furthercaused to receive, from the central processor, data symbols associated with the indication; and perform transmit beamforming using at least part of the precoding matrix based on the determined complex vectors and the received data symbols.

[0008] According to an example embodiment of the first aspect, the number of computedcomplex vectors and values indicative of strength of the effective channel within a set corresponds to a number of data layers for the respective user equipment.

[0009] According to an example embodiment of the first aspect, the uplink channelestimates are obtained for a certain physical resource block granularity.

[0010] According to an example embodiment of the first aspect, the one or more sets ofpairs of complex vectors and values are determined for a certain physical resource block granularity.

[0011] According to an example embodiment of the first aspect, the reference signalscomprise sounding reference signals.

[0012] According to an example embodiment of the first aspect, the values indicative ofstrength of the effective channel comprise singular values generated by the precoding matrix.

[0013] According to an example embodiment of the first aspect, the precoding matrix is asingular value decomposition based zero forcing precoder.

[0014] According to an example embodiment of the first aspect, the precoding matrixcomprises an eigen zero forcing precoder.

[0015] According to an example embodiment of the first aspect, the apparatus is furthercaused to receive, from the central processor, a power scaling factor, wherein the power scaling factor is a positive real number; and scale at least the complex vectors determined to be used for transmit beamforming based on the power scaling vector.

[0016] According to an example embodiment of the first aspect, the apparatus is furthercaused to calculate a power scaling factor based on an antenna power constraint; and scale at least the complex vectors determined to be used for transmit beamforming based on the power scaling vector.

[0017] According to an example embodiment of the first aspect, a length of the complexvector corresponds to a number of antennas of the first apparatus.

[0018] According to an example embodiment of the first aspect, the received indicationcomprises one or more indices of values comprised in the transmitted one or more sets of values.

[0019] According to an example embodiment of the first aspect, the apparatus is furthercaused to detect that the transmitted set of values for at least one of the one or more user equipment is being excluded based on the received indication; and recompute the precoding matrix after removing the at least one user equipment with the excluded set of values from consideration.

[0020] According to an example embodiment of the first aspect, the excluded set of valuesis detected based on the received indication comprising a certain index indicative of the excluded set of values for the at least one user equipment.

[0021] According to a second aspect, a method is disclosed. The method may comprise:estimating uplink channel based on reference signals from one or more user equipment; computing a precoding matrix based on the estimated uplink channel for the one or more user equipment; obtaining, for the one or more user equipment, a set of pairs of complex vectors and values indicative of strength of effective channel between the first apparatus and the respective user equipment based on the computed precoding matrix; transmitting, to a central processor, the obtained one or more sets of the values indicative of strength of effective channel for the one or more user equipment; receiving, from the central processor, an indication of which values from the transmitted one or more sets of values are determined to be used for transmit beamforming for the one or more user equipment; and determining, based on the received indication, one or more associated complex vectors of the precoding matrix to be used for transmit beamforming for the one or more user equipment.

[0022] According to an example embodiment of the second aspect, the method comprisesreceiving, from the central processor, data symbols associated with the indication; and performing transmit beamforming using at least part of the precoding matrix based on the determined complex vectors and the received data symbols.

[0023] According to an example embodiment of the second aspect, the number ofcomputed complex vectors and values indicative of strength of the effective channel within a set corresponds to a number of data layers for the respective user equipment.

[0024] According to an example embodiment of the second aspect, the uplink channelestimates are obtained for a certain physical resource block granularity.

[0025] According to an example embodiment of the second aspect, the one or more sets ofpairs of complex vectors and values are determined for a certain physical resource block granularity.

[0026] According to an example embodiment of the second aspect, the reference signalscomprise sounding reference signals.

[0027] According to an example embodiment of the second aspect, the values indicative ofstrength of the effective channel comprise singular values generated by the precoding matrix.

[0028] According to an example embodiment of the second aspect, the precoding matrixis a singular value decomposition based zero forcing precoder.

[0029] According to an example embodiment of the second aspect, the precoding matrixcomprises an eigen zero forcing precoder.

[0030] According to an example embodiment of the second aspect, the method comprisesreceiving, from the central processor, a power scaling factor, wherein the power scaling factor is a positive real number; and scaling at least the complex vectors determined to be used for transmit beamforming based on the power scaling vector.

[0031] According to an example embodiment of the second aspect, method comprisescalculating a power scaling factor based on an antenna power constraint; and scaling at least the complex vectors determined to be used for transmit beamforming based on the power scaling vector.

[0032] According to an example embodiment of the second aspect, a length of the complexvector corresponds to a number of antennas of the first apparatus.

[0033] According to an example embodiment of the second aspect, the received indicationcomprises one or more indices of values comprised in the transmitted one or more sets of values.

[0034] According to an example embodiment of the second aspect, the method comprisesdetecting that the transmitted set of values for at least one of the one or more user equipment is being excluded based on the received indication; and recomputing the precoding matrix after removing the at least one user equipment with the excluded set of values from consideration.

[0035] According to an example embodiment of the second aspect, the excluded set ofvalues is detected based on the received indication comprising a certain index indicative of the excluded set of values for the at least one user equipment.

[0036] According to a third aspect, an apparatus is disclosed. The apparatus may comprise:at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to receive, from a plurality of signal processing devices of a base station, one or more sets of values computed independently by the plurality of signal processing devices for one or more user equipment, wherein the values are indicative of strength of effective channel between the respective signal processing device and the one or more user equipment; determine, for each of the one or more user equipment, one or more largest values based on the one or more sets of values received from the plurality of signal processing devices; transmit, to the plurality of signal processing devices, an indication of which values received from the respective signal processing device for the one or more user equipment are to be used for transmit beamforming based on the determined one or more largest values for the one or more user equipment.

[0037] According to an example embodiment of the third aspect, the number of the one ormore largest values correspond to a number of data layers of the respective user equipment.

[0038] According to an example embodiment of the third aspect, the apparatus is furthercaused to determine a composite set of values for each of the one or more user equipment based on the one or more sets of values received from the plurality of signal processing devices; sort the composite set of values in an order of magnitude; and determine the one or more largest values based on the sorted composite set of values.

[0039] According to an example embodiment of the third aspect, the apparatus is furthercaused to transmit, to the plurality of signal processing devices, data symbols associated with the indication.

[0040] According to an example embodiment of the third aspect, the transmitted datasymbols are determined based on a number of the indicated values and an order of magnitude of the indicated values within the determined one or more largest values for the respective user equipment.

[0041] According to an example embodiment of the third aspect, the apparatus comprisesone of a radio unit of the base station, a distributed unit of the base station, one of the plurality of signal processing devices, or the plurality of signal processing devices wherein each of the signal processing devices is configured to serve as a central processor for non- overlapping frequency bands.

[0042] According to an example embodiment of the third aspect, the apparatus is furthercaused to determine a power scaling factor for each of the plurality of signal processing devices, wherein the power scaling factor is a positive real number; and transmit, to the plurality of signal processing devices, the determined power scaling factors.

[0043] According to an example embodiment of the third aspect, the indication comprisesat least one of: -one or more indices of values comprised in the determined one or more largestvalues that were received from the signal processing device; or -a certain index indicative that none of the values received from the signalprocessing device for at least one of the one or more user equipment were included in the determined one or more largest values for the at least one user equipment.

[0044] According to a fourth aspect, a method is disclosed. The method comprisesreceiving, from a plurality of signal processing devices of a base station, one or more sets of values computed independently by the plurality of signal processing devices for one or more user equipment, wherein the values are indicative of strength of effective channel between the respective signal processing device and the one or more user equipment; determining, for each of the one or more user equipment, one or more largest values based on the one or more sets of values received from the plurality of signal processing devices; transmitting, to the plurality of signal processing devices, an indication of which values received from the respective signal processing device for the one or more user equipment are to be used for transmit beamforming based on the determined one or more largest values for the one or more user equipment.

[0045] According to an example embodiment of the fourth aspect, the number of the oneor more largest values correspond to a number of data layers of the respective user equipment.

[0046] According to an example embodiment of the fourth aspect, the method comprisesdetermining a composite set of values for each of the one or more user equipment based on the one or more sets of values received from the plurality of signal processing devices;sorting the composite set of values in an order of magnitude; and determining the one or more largest values based on the sorted composite set of values.

[0047] According to an example embodiment of the fourth aspect, the method comprisestransmitting, to the plurality of signal processing devices, data symbols associated with the indication.

[0048] According to an example embodiment of the fourth aspect, the transmitted datasymbols are determined based on a number of the indicated values and an order of magnitude of the indicated values within the determined one or more largest values for the respective user equipment.

[0049] According to an example embodiment of the fourth aspect, the method is performedby one of the following: a radio unit of the base station, a distributed unit of the base station, one of the plurality of signal processing devices, or the plurality of signal processing devices wherein each of the signal processing devices is configured to serve as a central processor for non-overlapping frequency bands.

[0050] According to an example embodiment of the fourth aspect, the method comprisesdetermining a power scaling factor for each of the plurality of signal processing devices, wherein the power scaling factor is a positive real number; and transmitting, to the plurality of signal processing devices, the determined power scaling factors.

[0051] According to an example embodiment of the fourth aspect, the indication comprisesat least one of: -one or more indices of values comprised in the determined one or more largestvalues that were received from the signal processing device; or -a certain index indicative that none of the values received from the signalprocessing device for at least one of the one or more user equipment were included in the determined one or more largest values for the at least one user equipment.

[0052] According to a fifth aspect, a base station is disclosed. The base station maycomprise a plurality of apparatuses according to the first aspect, and the apparatus according to the third aspect. The base station may be further configured to perform any example embodiments of the first aspect and the second aspect, as provided in the description and / or the claims.

[0053] According to a sixth aspect, a method is disclosed. The method may compriseperforming a method according to the second aspect and a method according to the fourthaspect, or any example embodiment(s) thereof, as provided in the description and / or the claims.

[0054] According to a seventh aspect, an apparatus is disclosed. The apparatus maycomprise means for performing the method according to the second, fourth, or sixth aspect, or any example embodiment(s) thereof, as provided in the description and / or the claims.

[0055] According to an eight aspect, a computer program, a computer program product, ora (non-transitory) computer-readable medium is disclosed. The computer program, computer program product, or (non-transitory) computer-readable medium may comprise instructions, which when executed by an apparatus, cause the apparatus at least to perform the method according to the second, fourth, or sixth aspect, or any example embodiment(s) thereof, as provided in the description and / or the claims.

[0056] Example embodiments of the present disclosure can thus provide apparatuses,methods, computer programs, computer program products, or computer readable media for improving various aspects of signal processing. Any example embodiment may be combined with one or more other example embodiments. These and other aspects of the present disclosure will be apparent from the example embodiment(s) described below. According to some aspects, there is provided the subject matter of the independent claims. Some further aspects are defined in the dependent claims. DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which are included to provide a further understandingof the example embodiments and constitute a part of this specification, illustrate example embodiments and, together with the description, help to explain the example embodiments. In the drawings:

[0058] FIG. 1 illustrates an example of a communication network;

[0059] FIG. 2 illustrates an example of an apparatus configured to practice one or moreexample embodiments;

[0060] FIG. 3 illustrates an example of a block diagram of distributed signal processingbetween a central processor and a plurality of chiplets;

[0061] FIG. 4 illustrates an example of signalling and operations for distributed signalprocessing;

[0062] FIG. 5 illustrates an example of throughput performance comparison betweendifferent types of signal processing techniques;

[0063] FIG. 6 illustrates an example of computational complexity comparison betweendifferent types of signal processing techniques;

[0064] FIG. 7 illustrates an example of comparison of link rate requirements betweendifferent types of signal processing techniques;

[0065] FIG. 8 illustrates an example of a method for distributed signal processing; and

[0066] FIG. 9 illustrates an example of another method for distributed signal processing.

[0067] Like references are used to designate like parts in the accompanying drawings.DETAILED DESCRIPTION

[0068] Reference will now be made in detail to example embodiments, examples of whichare illustrated in the accompanying drawings. The detailed description provided below in connection with the appended drawings is intended as a description of the present examples and is not intended to represent the only forms in which the present example may be constructed or utilized. The description sets forth the functions of the example and the sequence of steps for constructing and operating the example. However, the same or equivalent functions and sequences may be accomplished by different examples.

[0069] FIG. 1 illustrates an example of a communication network. Communicationnetwork 100 may comprise one or more base stations 104, 106, 108. Base station(s) 104, 106, 108 may be part of a radio access network (RAN) configured to enable a device, represented throughout the description by UE 102, to access communication services provided by core network 110. In connection with communication network 100, base station(s) 104, 106, 108 and core network 110 may be collectively referred to as the ‘network’. UE 102 may comprise a user device, a user node, a mobile device, or the like. UE 102 may be configured to communicate with base station(s) 104, 106, 108 over a radio interface, which may be also referred to as an air interface. Base stations 104, 106, 108 may be also referred to as network devices, access nodes, network nodes, network entities, or the like.

[0070] The radio interface may be configured for example based on the 5G NR (NewRadio) standard defined by the 3rdGeneration Partnership Project (3GPP), or any future standard or technology (e.g., 6G). Base stations 104, 106, 108 may comprise, for example, 5thgeneration access nodes (gNB). Transmission by a base station to UE 102 may be called downlink (DL) transmission. Transmission by UE 102 to a base station may be called uplink (UL) transmission. UE 102 may be therefore configured to operate as a transmitter for uplink transmissions and as a receiver for downlink transmissions. Base station(s) 104, 106, 108may be configured to operate as a receiver for uplink transmissions and as a transmitter for downlink transmissions. Communication network 100 may comprise a wireless communication network or a mobile communication network, such as for example a cellular communication network.

[0071] In a split access node architecture, the base station may be divided into two physicalentities called CU (central unit or centralized unit) and DU (distributed unit). A part of protocol layers may be implemented at the CU of a base station, e.g., a gNB-CU, which may be configured to handle upper layers of the protocol stack, for example SDAP (service data adaptation protocol) and PDCP (packet data convergence protocol) layers. Furthermore, the central unit may be configured to handle radio resource control (RRC) operations. A central unit of a base station may be associated with, e.g., configured to control, one or more DUs of the base station, e.g., gNB-DU, which may be configured to handle lower layers of the protocol stack, for example RLC (radio link control), MAC (medium access control), and physical layer. Radio unit(s) of the distributed unit(s) may be configured to transmit / receive data to / from UE(s) over the radio interface. A central unit may be referred to as a central node and a distributed unit may be referred to as a distributed node.

[0072] In the multi-user extreme MIMO systems, scalable signal processing at the gNB forboth uplink and downlink transmission may be needed. When progressing is shifted from 64 TRX to 256 or more TRX at a gNB, the following challenge immediately presents itself. At present, there is a limit on the number of TRX that can fit in a single monolithic system-on- chip (SoC) without dangerously overheating the chip. This number is closer to 64 TRXs for safe operations using a currently available CMOS process (e.g., a 5 nm or a 7nm CSMOS technology platform). This means that in order to accommodate 256 or more TRXs in a SoC, it may be needed to use multiple SoCs (or chiplets) in a package (i.e., System-in-Package or SiP). Therefore, to have 256 or more TRX, a modular base-station with several chiplets may be designed.

[0073] A chiplet may refer to an integrated circuit that comprises a well-defined subset offunctionality. A chiplet may be modular such that it is designed to be combined, or to function e.g., communicate, with other chiplets on the same package. A chiplet may be also referred to as a sub processing unit, a sub processing device or a processing module.

[0074] In one approach, a gNB may include multiple chiplets, each chiplet comprising asingle panel of antennas of size 32 TRX each, and each chiplet is assigned a set of user equipment (UE) to be served on the downlink. A UE may be assigned one chiplet, and eachchiplet may individually process the signals meant for the set of UEs that are assigned to it. This type of MIMO gNB may be called a “structural MIMO” or S-MIMO.

[0075] One alternative to individually designing a precoder in each chiplet is to jointlydesign a DL precoding matrix across all chiplets for all the UEs. This may involve exchanging the channel state information (CSI) between all the chiplets. For example, each chiplet can transmit the CSI for all the UEs that was estimated by it to a central processor. This central processor can then jointly design a zero-forcing (ZF) precoding matrix and transmit back the relevant part of this precoding matrix to each chiplet which then transmit the data after transmit beamforming. One example of an applicable ZF precoding matrix is a singular-value decomposition (SVD)-based ZF precoder, which can be referred to as SVD- ZF.

[0076] Hence, in the example approaches, signal processing can be performed based onindependent chiplet processing or centralized processing. In independent chiplet processing, each chiplet may be configured to independently design a DL precoder for the set of UEs assigned to it. However, unless there is a complete nulling of the channel from all the other unassigned chiplets to the UE, there may be a degradation of the signal-to-interference-and- noise ratio (SINR) at each UE resulting in throughput loss. In centralized processing, each chiplet may transmit the CSI for all UEs that was estimated by the chiplet to a central processor. However, in this case, a lot of information may need to be exchanged through the links between the chiplets and the central processor. This may not be feasible, for example, if the central processor is in a distributed unit (DU) of a base station with the links being evolved common public radio interfaces (eCPRI). Further, the central processor may need to perform the bulk of the precoding computation, which may be a significant amount of workload on the central processor.

[0077] An example embodiment enables a distributed processing technique for jointlydesigning the DL precoder for a multiple SoC-based base station (e.g., a gNB or S-MIMO). The distributed processing technique may enable to significantly reduce the amount of information exchange between the chiplets and the central processor. In addition, the precoder design computation may be distributed among the chiplets such that the central processor may not be over-burdened. Compared to independent chiplet processing, the obtainable UE throughputs may be improved, and the amount of throughput degradation may be reduced compared to the joint DL precoder design using a centralized processing approach.

[0078] FIG. 2 illustrates an example of an apparatus configured to practice one or moreexample embodiments. Apparatus 200 may comprise a device such as a base station (e.g., gNB 104, 106, 108), an access node, a radio network node, or a split portion thereof (e.g., a radio unit or distributed unit of a base station), a network device, or in general any apparatus configured to implement functionality described herein. In one example, apparatus 200 may comprise one or more signal processing devices of a modular base station. A signal processing device may comprise a system-on-a-chip, such as a central processor or one or more chiplets of a base station, or a combination thereof.

[0079] Apparatus 200 may comprise at least one processor 202. The at least one processor202 may comprise, for example, one or more of various processing devices, such as for example a co-processor, a microprocessor, a controller, a digital signal processor (DSP), a processing circuitry with or without an accompanying DSP, or various other processing devices including integrated circuits such as, for example, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a microcontroller unit (MCU), a hardware accelerator, a special-purpose computer chip, or the like.

[0080] Apparatus 200 may further comprise at least one memory 204. The memory 204may be configured to store, for example, computer program code or the like, for example operating system software and application software. Memory 204 may comprise one or more volatile memory devices, one or more non-volatile memory devices, and / or a combination thereof. For example, the memory may be embodied as magnetic storage devices (such as hard disk drives, magnetic tapes, etc.), optical magnetic storage devices, or semiconductor memories (such as mask ROM, PROM (programmable ROM), EPROM (erasable PROM), flash ROM, RAM (random access memory), etc.). Memory 204 is provided as an example of a (non-transitory) computer readable medium. The term “non-transitory,” as used herein, is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storage persistency (e.g., RAM vs. ROM).

[0081] Apparatus 200 may further comprise a communication interface 208 configured toenable apparatus 200 to transmit and / or receive information. Communication interface 208 may comprise an external communication interface, such as for example a radio interface between UE 102 and access node(s) 104, 106, 108, or a communication interface between a central unit and distributed unit(s) of an access node. Communication interface 208 may comprise a communication interface between a plurality of distributed signal processing devices of a base station. Communication interface 208 may comprise one or more radiotransmitters or receivers, which may be coupled to one or more antennas or apparatus 200, or be configured to be coupled to one or more antennas external to apparatus 200.

[0082] Apparatus 200 may further comprise other components and / or functions such as auser interface (not shown) comprising at least one input device and / or at least one output device. The input device may take various forms such a keyboard, a touch screen, or one or more embedded control buttons. The output device may for example comprise a display, a speaker, or the like.

[0083] When apparatus 200 is configured to implement some functionality, somecomponent and / or components of apparatus 200, such as for example the at least one processor 202 and / or the at least one memory 204, may be configured to implement this functionality. Furthermore, when the at least one processor 202 is configured to implement some functionality, this functionality may be implemented using program code 206 comprised, for example, in the at least one memory 204.

[0084] The functionality described herein may be performed, at least in part, by one ormore computer program product components such as software components. According to an example embodiment, apparatus 200 comprises a processor or processor circuitry, such as for example a microcontroller, configured by the program code 206, when executed, to execute the embodiments of the operations and functionality described herein. Program code 206 is provided as an example of instructions which, when executed by the at least one processor 202, cause performance of apparatus 200.

[0085] Alternatively, or in addition, the functionality described herein can be performed,at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include field- programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip systems (SOCs), complex programmable logic devices (CPLDs), graphics processing units (GPUs), or the like.

[0086] Apparatus 200 may be configured to perform, or cause performance of, method(s)described herein or comprise means for performing method(s) described herein. In one example, the means comprises the at least one processor 202, the at least one memory 204 including instructions (e.g., program code 206) configured to, when executed by the at least one processor 202, cause apparatus 200 to perform the method(s). In general, computer program instructions may be executed on means providing generic processing functions. Such means may be embedded for example in a personal computer, a smart phone, a network device, or the like. The method(s) may be thus computer-implemented, for example, basedon algorithm(s) executable by the generic processing functions, an example of which is the at least one processor 202. The means may comprise transmission or reception means, for example one or more radio transmitters or receivers, which may be coupled or be configured to be coupled to one or more antennas. Although apparatus 200 is illustrated as a single device, it is appreciated that, wherever applicable, functions of apparatus 200 may be distributed to a plurality of devices.

[0087] FIG. 3 illustrates an example of a flow chart for distributed signal processingaccording to an example embodiment. The distributed signal processing may be performed by a base station, such as a gNB. The signal processing may be performed in a distributed fashion by the base station with a central processor (CP) communicatively coupled with a plurality of chiplets. The CP and the chiplets may be modules comprised in the base station. The plurality of chiplets may be physically separate entities. Each chiplet may be configured to handle the entire bandwidth of subcarriers. In one example, the CP may be a separate entity. For example, the CP may be one of a separate entity within a RU, a separate entity in a DU connected to the chiplets using a eCPRI link, or one of the chiplets. In one example, the CP may be comprised in the plurality of chiplets, wherein each chiplet serves as the CP for non-overlapping frequency subbands.

[0088] At operation 300, each of the plurality of chiplets may be configured to estimate anUL channel using reference signals from all the co-scheduled UEs ^^, wherein ^^ = 1, ⋯ , ^^^^,and ^^^^denotes the total number of users (UEs). The reference signals may be UL RS for DL, such as sounding reference signals (SRSs). The chiplets may be configured to estimate the UL channel at a known granularity of physical resource blocks (PRBs).

[0089] At operation 302, each of the plurality of chiplets may be configured to compute aprecoder based on the part of the UL channel that the chiplet has estimated. The precoder may be, for example, a ZF precoder. In other words, each chiplet may compute the ZF precoder based on its own measurements and independently from the other chiplets. In one example, the ZF precoder may be an SVD-ZF precoder. The SVD-ZF precoder may comprise a set of pairs of complex vectors and non-negative real-valued singular values. The set of pairs of complex vectors and the singular values may be computed for each co- scheduled UE and for the known granularity of PRBs. The number of the complex vectors and the singular values may equal the number of transmit antennas in the chiplet. For example, each of the plurality of chiplets may comprise, or be coupled to, a sub-antenna panel of the base station. Instead of the SVD-ZF precoder, the chiplets may be configured to use any arbitrary ZF precoder. In one example, the ZF precoder may be an eigen ZF precoder(EZF). The used ZF precoder may be configured to generate the singular values. A singular value may indicate a strength of link between the chiplet and a measured UE.

[0090] At operation 304, each of the plurality of chiplets may be configured to transmit theset of singular values for each co-scheduled UE at the known granularity of PRBs to the CP.

[0091] At operation 306, the CP may be configured to accumulate all the received singularvalues from all the chiplets for each UE to obtain a composite set of singular values. Based on the composite set of singular values, the CP may determine one or more singular values having largest values. For example, the CP may be configured to sort the composite set of received singular values in a descending order of magnitude, and select one or more first singular values of the sorted array of singular values. Alternatively, the CP may be configured to sort the composite set of values in an ascending order of magnitude, and select one or more last singular values of the sorted array of singular values. The number of selected singular values may be based on a number of data layers of the UE. The number of data layers of the UE may depend on a number of receivers at the UE. Each of the singular values may be indexed. For example, the index of a singular value may indicate an order of the singular value in a certain set of singular values received from a certain chiplet.

[0092] After determining the one or more largest singular values for the UE from thereceived sets of singular values for that UE, the CP may be configured to inform the plurality of chiplets about the determined values. For example, the CP may be configured to transmit to at least the chiplets from which chiplets the sets of singular values comprising the one or more largest singular values were received from, an indication of which values of the received sets of singular values were determined as the largest. In one example, at operation 308, the CP may be configured to transmit, to the plurality of chiplets, the indices of the first elements of the sorted array for each UE ^^. The elementsmay denote the number of data layers for UE i. The indices may be transmitted to the respective chiplets. That is, if a singular value transmitted by a certain chiplet is not selected by the CP, no index may be transmitted to the certain chiplet. Alternatively, the CP may transmit an index indicative that no singular value transmitted by the certain chiplet was selected, such as an index {0}. It is noted, that the indices are one example, and the CP may be also configured to indicate the selected singular values by other means.

[0093] For example, suppose that there are 4 chiplets and UE ^^ has 4 data layers (^^(^^) ^^= 4). Each chiplet transmits 4 singular values to the CP, and the CP receives a total of 16 singular values at the known PRB granularity. The CP sorts these 16 singular values andchooses the 4 largest singular values. Suppose that these singular values were the first two singular values of Chiplet 1, the first singular value of Chiplet 2, and the first singular value of Chiplet 3 in that order. The CP may then transmit the indices{1,2}to Chiplet 1,{1}to Chiplet 2,{1}to Chiplet 3 for that set of PRBs, and nothing (or{0}) to Chiplet 4.

[0094] At operation 308 the CP may be configured to transmit, to the plurality of chiplets,data symbols for each UE ^^. The data symbols may be transmitted to concerned chiplets according to the indices of the respective determined singular values. The transmitted data symbols may correspond to a data layer of the UE based on an order of the value received form the respective chiplet (index) within the selected set of largest values.

[0095] Taking the previous example, the CP may transmit the data symbols of the first andsecond layer for UE ^^ in that set of PRBs to Chiplet 1, the data symbols of the third layer forUE ^^ to Chiplet 2, and the data symbols of the fourth layer for UE ^^ to Chiplet 3, and nothingto Chiplet 4.

[0096] In one example, the CP may be further configured to transmit to each chiplet asingle power scaling factor. The power scaling factor may be a positive real number. The power scaling factor may be calculated by the CP for each chiplet, for example, based on a per antenna power constraint.

[0097] At operation 310 the concerned chiplets may be configured to choose the associatedcomplex vectors of the ZF precoder for each UE ^^ based on the received one or more indices (or other received indication of the singular values selected by the CP). Based on the chosen complex vector(s) and the received data symbols, the chiplet may be configured to perform transmit beamforming, and transmit the composite signal vectors from its antennas. Hence, the chiplet may perform signal transmission on relevant UEs based on the indication received from the CP.

[0098] Continuing with the previous example, Chiplet 1 may choose the first two complexvectors of the SVD-ZF precoder for the concerned set of PRBs and weight these vectors using the received data layers for the concerned UE. The Chiplet 1 may perform the same for all co-scheduled UEs and transmits (to all the UEs) the sum of all these weighted vectors for all the co-scheduled UEs in each resource element (RE) of the concerned set of PRBs.

[0099] In one example, the chiplet may also scale the precoder vectors by a power scalingfactor that it received from the CP. Further, in some embodiments, the chiplet can be configured to recalculate the ZF-precoder if it received no index for at least one of the UEs. Alternatively, the chiplet may be configured to recalculate the ZF-precoder if it received a certain index (such as{0}) indicative that no value from the set of singular values receivedfrom that chiplet was included in the determined one or more largest values. When recalculating the ZF precoder, the chiplet may remove CSI for the UE for which no index (or {0}) was received to remove the UE from consideration.

[0100] The distributed signal processing technique illustrated in FIG. 3 may enable toreduce the amount of information exchange between the chiplets and the CP compared to a centralized processing by the CP. Further, when the precoder design, which can be computationally intensive, is shared among the chiplets, the computational load on the CP can be decreased.

[0101] FIG. 4 illustrates an example of signaling and operations between a centralprocessor and a plurality of chiplets for distributed signal processing according to an example embodiment. The central processor and the plurality of chiplets may be comprised in a base station, such as gNB 104, 106 or 108.

[0102] In the following:- ^^^^ denotes the total number of co-scheduled users,- ^^^^ denotes the total number of gNB transmit antennas,- the number of rec ^^, ^^ = ^^(^^)eption antennas at UE ^^^^, -the number of data layers-^^^^,^^ ∈is the transmitted symbol vector of lengthon subcarrier ^^ where UE ^^ uses unit-energy constellation ℚ^^,- ^^^^,^^ ∈ ℂ^^(^^) ^^×1is received signal vector at UE ^^ on subcarrier ^^, -^^^^,^^ ∈ ℂ^^(^^) ^^×^^^^is channel matrix of UE ^^ on subcarrier ^^,--^^^^ ∈ ℂ^^^^×^^^^ precoding matrix on subcarrier ^^ , with ^^^^ ≝-^^(^^) (^^)ℂ ^^ ^^on subcarrier

[0103] The channel model at the receiver of UE ^^ (depicted as UE 102) on subcarrier ^^ isgiven as ^^^^,^^ = ^^^^,^^^^^^^^^^ + ^^^^,^^ .

[0104] It is assumed that the channels ^^^^,^^ are estimated on the uplink using SRS for a TDD(time-division duplex) system, with ^^ ^^,^^denoting the DL channel estimate from the gNB to UE ^^. It is also assumed that the channel estimates are obtained for a certain PRB granularity, i.e., one channel estimate per few PRBs (for example, 2 or 4 PRBs). Each chiplet 402 may be configured to estimate the channels at SRS frequency meaning that the operation is performed at the same periodicity as that of UE SRS transmissions which is approximately once every 20 slots. For example, at operation 300, each chiplet 402 may be configured to estimate UL channel based on reference signals from all the co-scheduled UEs ^^.Each chiplet 402 may be configured to compute, for example, UL SRS CSI ^^^^.

[0105] At operation 302, each chiplet 402 may be configured to compute a partial precoder,such as SVD-ZF. The partial precoder may be referred to as a sub-panel precoder. For example, each chiplet 402 may comprise a sub-panel of a base station, wherein the sub-panelcomprises a plurality of antennas 404. Let ^^ ^^,^^,^^the channel from chiplet ^^ toUE ^^, ^^ = 1, ⋯ , ^^^^, ^^ = 1, ⋯ , ^^^^. ^^^^ may refer to the total number of chiplets 402 at gNB.Chiplet ^^ computes its own ^^ ^^,^^,^^ ∈, where ^^^^,^^,^^is in the null-space of ^^^^,\^^,^^∈ ≥^^(^^). Each chiplet 402 computes(^^) (^^)right singular vectors ^^ ^^,^^,^^ ∈ ℂ^^^^×^^^^and thenon-zero real-valued singular valuesof ^^^^,^^,^^^^ . Singular vectors of a matrix may describe a direction of maximum action, and the corresponding singular values may describe a magnitude of the action, e.g., a direction and strength of an effective channel. It is to be noted that this principle applies even for arbitrary ZF-precoders (for e.g., EZF precoders) wherein ^^^^,^^,^^represents the ZF- precoder used.

[0106] Each chiplet 402 can send the singular values at operation 304 to the CP 400 for UE^^, ∀^^ = 1, ⋯ , ^^ . The CP 400 then identifies thelargest elementsFor example, at operation 306, the CP 400 may be configured to sort the singular values sent by the chiplets 402 at 304, and identify column indices for each sub-panel precoder. Thereafter, at operation 308, the CP 400 may be configured to transmit relative positions ofthe column indices to the chiplets 402 concerned. For example,= 4 and ^^ >^^ ^^,^^,1,2 > ^^ ^^,^^,2,3are the largest 4 values in that order, then, the CP 400 sends theindices {1,2} to chiplet 3, to chiplet 2 and chiplet 1 the index {1}. It also sends to chiplet 3 the data symbols of UE ^^ corresponding to layers 1 and 3, to chiplet 2 the data symbols of UE ^^ corresponding to layer 2, and to chiplet 1 the data symbols of UE ^^ corresponding to layer 4. The CP may perform transmission of the indication of the indices and / or data symbols at every slot.

[0107] Dropping the subscript denoting the frequency, let ℐ^^,^^ be the set of indices sent bythe CP 400 to chiplet ^^ for UE ^^. Further, let ^^^^,^^be the data symbol vector for UE ^^ 102 andchiplet ^^. Then, chiplet ^^ first chooses the columns of ^^ ^^,^^ corresponding to ℐ^^,^^ and performstransmit data beamforming at operation 310 using part of the SCD-ZF precoder associated with the received indices and data symbols Wi,ssi,s. The transmit data beamforming may be computed as ^^^^^^^^∑ ^^ ^^ ^^ ∈^^^^×1 ^^,^^ ^^,^^ ^^,^^ ℂ, ^^=1 where ^^^^,^^is a matrix whose columns are chosen from the columns of ^^^^,^^using the set of indices ℐ^^,^^. This transmit-beamformed vector can be then transmitted over the wireless channel at 406. In case a chiplet receives {0} (or other predetermined index indicating that none of the values received from the chiplet was received) as the set of indices for at least one of the served UE, it can recalculate the ZF-precoder after removing the effect of that UE. This may improve performance at the cost of additional computations. The disclosed technique effectively assigns a single chiplet to each layer of each UE. Note that under the assumption of ideal CSI knowledge, this operation is zero-forcing.

[0108] FIG. 5 illustrates an example of throughput performance comparison betweendifferent types of signal processing techniques. Plot 500 shows arithmetic mean goodputs with 90% confidence interval, plot 502 illustrates geometric mean goodputs with 90% confidence level, with bottom 5% users removed, plot 504 illustrates 5thpercentile goodputs, and plot 506 illustrates 10thpercentile goodputs for the signal processing approaches 508, 510, 512, 524, 516 under comparison.

[0109] In each plot, the bars from left to right illustrate the following signal processingapproaches: full-panel Moore-Penrose Inverse ZF (MPI-ZF) processing 508, full-panel SVD-ZF processing 510, 4 chiplets with distributed SVD-ZF processing 512, 4 chiplets with naive SVD-ZF processing 514, 4 chiplets with independent SVD-ZF processing 516.

[0110] Full-panel MPI-ZF refers to full-panel processing with the ZF precoder calculatedbased on the Moore-Penrose inverse of the composite channel. In full-panel MPI-ZFapproach, the CP jointly designs a precoder based on channel information transmitted by chiplets to the CP. Full-panel SVD-ZF refers to full-panel processing with SVD-ZF precoding. 4 chiplets with naïve SVD-ZF processing refers to a case where there are 4 chiplets and each chiplet designs an SVD-ZF precoder for all the layers of all UE. Finally, 4 chiplets with independent SVD-ZF processing refers to a case where the set of UEs is divided into 4 disjoint subsets, each subset assigned to one chiplet. Each chiplet only serves the subset of UEs assigned to it using SVD-ZF precoding.

[0111] The plots 500, 502, 504, 506 show example simulation results of a multi-cell, multi-link-level simulation (MCMLLS). Letdenote the ^^^^ℎ row of ^^^^, ^^ = 1, ⋯ , ^^^^, ^^ =1, ⋯ , ^^^^ . the columns of ^^^^ are normalized to unity. Let ^^^^^^^^ denote the per antenna powerconstraint for the entire bandwidth of ^^^^subcarriers. Then, the precoder is taken to be^^^^ ^ √^^ ^^^^ withThis operation may not destroy the zero-forcing effect while being power compliant. In the case of full-panel centralized processing, the CP calculates the value of ^^ and scales the precoder weights using√^^ before sending the precoder matrix to the chiplets. In the case of independent chiplet processing, each chiplet calculates the value of ^^ and applies it to its precoder. For the case of distributed signal processing as described in this disclosure, there are two possibilities: 1) Each chiplet individually calculates the value of ^^ and applies it to its precoder.2) The CP assigns the value of ^^ for each chiplet (one per chiplet) which it transmits toeach chiplet. In the simulations, the first technique is used.

[0112] Based on plot 500, it can be seen that an average arithmetic UE goodput isapproximately 14.3 Mbps for full-panel MPI-ZF 508, 20.3 Mbps for full-panel SVD-ZF 510, 15.8 Mbps for 4 chiplets with SVD-ZF 512, 12.8 Mbps for 4 chiplets with naïve SVD-ZF and 9.0 Mbps for 4 chiplets with independent SVD-ZF.

[0113] Based on plot 502, it can be seen that an average geometric UE goodput isapproximately 11.5 Mbps for full-panel MPI-ZF 508, 16.3 Mbps for full-panel SVD-ZF 510, 12.0 Mbps for 4 chiplets with distributed SVD-ZF 512, 9.7 Mbps for 4 chiplets with naïve SVD-ZF and 6.5 Mbps for 4 chiplets with independent SVD-ZF.

[0114] Based on the plots, full-panel SVD-ZF 510 has the best performance, but thedistributed signal processing technique of this disclosure (illustrated by the 4 chiplets with distributed SVD-ZF 512) is better than all other baseline approaches 508, 514, 516. The distributed signal processing technique described herein provides better results than the full- panel MPI-ZF 508 which is one of the most popular ZF precoders. In particular, the distributed signal processing technique provides improvement over the independent chiplet processing method 516 by a significant 70% in arithmetic goodput.

[0115] FIG. 6 illustrates an example of computational complexity comparison betweendifferent types of signal processing techniques.

[0116] Chart 600 shows complexity comparison of signal processing performed with 4chiplets and independent SVD-ZF, 4 chiplets and distributed SVD-ZF and full-panel SVD- ZF. Bars 602, 604 and 606 show computation complexity involved for channel estimation per chiplet, bars 608, 610 and 612 show precoder calculation per chiplet and bar 614 shows precoder calculation at central processor, in terms of floating-point operation (FLOP) per PRB for the three signal processing techniques. It is noted that there is no precoding computation at the CP for independent chiplet processing and for the disclosed technique the precoding computation at the CP may only involve sorting which is negligible and not visible on the chart 600. While independent chiplet processing has the lowest computational complexity based on bars 602 and 608 as may be expected, the disclosed technique has much lower computational complexity (bars 604 and 610) than full-panel SVD-ZF (bars 606, 612 and 614) in addition to the signal processing operations being distributed.

[0117] FIG. 7 illustrates an example of comparison of link rate requirements betweendifferent types of signal processing techniques.

[0118] In chart 700, link requirements for CSI transfer, precoder transfer, and data symboltransfer are shown separately. It is further assumed that complex numbers need 20 bits of precision, real symbols need 10 bits of precision, and data symbols which take values from a complex QAM (quadrature amplitude modulation) constellation (maximum size 1024) need 10 bits. The PRB granularity for channel estimation is assumed to be 4 (1 channel estimate for every 4 PRBs). It is also assumed that the SRS signals of different users are spread in time so that the chiplets don’t have to send the channel estimates all in one slot. It is assumed that 12 UEs are co-scheduled with 2 layers each. The maximum load requirement in a slot would be during the transfer of the precoder and the data symbols to the chiplets.

[0119] Chart 700 shows the link rate requirements for data symbol transfer when signalprocessing is performed with 4 chiplets and independent SVD-ZF at 702, with 4 chiplets andnaïve SVD-ZF at 704, with 4 chiplets and distributed SVD-ZF at 706, with full-panel SVD- ZF at 708 and with full-panel MPI-ZF at 710. Further, chart 700 shows the link rate requirements for precoder transfer when signal processing is performed with 4 chiplets and distributed SVD-ZF at 712, with full-panel SVD-ZF at 714 and with full-panel MPI-ZF at 716. Chart 700 also shows the link rate requirements for CSI (channel) transfer when signal processing is performed with 4 chiplets and distributed SVD-ZF at 718, with full-panel SVD-ZF at 720 and with full-panel MPI-ZF at 722.

[0120] Hence, it can be clearly seen based on the chart 700 that the disclosed techniqueillustrated with the 4 chiplets and distributed SVD-ZF at 704 may need the same link capacity as independent chiplet processing at 702, and nearly an order of magnitude lower than that of full-panel processing at 708.

[0121] FIG. 8 illustrates an example of a method 800 for distributed signal processing.Method 800 may be performed by an apparatus, such a sub signal processing device of a base station, such as the gNB 104. The sub signal processing device may be, for example, a chiplet communicatively coupled with a central processor of the gNB 104.

[0122] At 802, the method may comprise estimating uplink channel based on referencesignals from one or more user equipment.

[0123] At 804, the method may comprise computing a precoding matrix based on theestimated uplink channel for the one or more user equipment.

[0124] At 806, the method may comprise obtaining, for the one or more user equipment, aset of pairs of complex vectors and values indicative of strength of effective channel between the first apparatus and the respective user equipment based on the computed precoding matrix.

[0125] At 808, the method may comprise transmitting, to a central processor, the obtainedone or more sets of the values indicative of strength of effective channel for the one or more user equipment.

[0126] At 810, the method may comprise receiving, from the central processor, anindication of which values from the transmitted one or more sets of values are determined to be used for transmit beamforming for the one or more user equipment.

[0127] At 812, the method may comprise determining, based on the received indication,one or more associated complex vectors of the precoding matrix to be used for transmit beamforming for the one or more user equipment.

[0128] FIG. 9 illustrates an example of a method 900 for distributed signal processing.Method 900 may be performed by an apparatus, such as a central processor of a base station, such as the gNB 104.

[0129] At 902, the method may comprise receiving, from a plurality of signal processingdevices of a base station, one or more sets of values computed independently by the plurality of signal processing devices for one or more user equipment, wherein the values are indicative of strength of effective channel between the respective signal processing device and the one or more user equipment.

[0130] At 904, the method may comprise determining, for each of the one or more userequipment, one or more largest values based on the one or more sets of values received from the plurality of signal processing devices.

[0131] At 906, the method may comprise transmitting, to the plurality of signal processingdevices, an indication of which values received from the respective signal processing device for the one or more user equipment are to be used for transmit beamforming based on the determined one or more largest values for the one or more user equipment.

[0132] Further features of the methods directly result for example from functionality ofgNB(s) 104, 106 and / or 108, as described throughout the description, claims, and drawings, and are therefore not repeated here. An apparatus, for example a device such as a base station (e.g., gNB 104, 106 and / or 108), a central processor of the base station or a signal processing module of the base station, may be configured to perform or cause performance of any aspect of the method(s) described herein. Further, a computer program, a computer program product, or a (non-transitory) computer-readable medium may comprise instructions for causing, when executed by an apparatus, the apparatus to perform any aspect of the method(s) described herein. Further, an apparatus may comprise means for performing any aspect of the method(s) described herein. According to an example embodiment, the means comprises at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to perform any aspect of the method(s).

[0133] Any range or device value given herein may be extended or altered without losingthe effect sought. Also, any embodiment may be combined with another embodiment unless explicitly disallowed.

[0134] Although the subject matter has been described in language specific to structuralfeatures and / or acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, thespecific features and acts described above are disclosed as examples of implementing the claims and other equivalent features and acts are intended to be within the scope of the claims.

[0135] It will be understood that the benefits and advantages described above may relateto one embodiment or may relate to several embodiments. The embodiments are not limited to those that solve any or all of the stated problems or those that have any or all of the stated benefits and advantages. It will further be understood that reference to 'an' item may refer to one or more of those items.

[0136] The steps or operations of the methods described herein may be carried out in anysuitable order, or simultaneously where appropriate. Additionally, individual blocks may be deleted from any of the methods without departing from the scope of the subject matter described herein. Aspects of any of the example embodiments described above may be combined with aspects of any of the other example embodiments described to form further example embodiments without losing the effect sought.

[0137] The term 'comprising' is used herein to mean including the method, blocks, orelements identified, but that such blocks or elements do not comprise an exclusive list and a method or apparatus may contain additional blocks or elements.

[0138] As used herein, “at least one of the following: ”and “at least one of ” and similar wording, where the list of two or more elements are joined by “and” or “or”, mean at least any one of the elements, or at least any two or more of the elements, or at least all the elements.

[0139] Although subjects may be referred to as ‘first’ or ‘second’ subjects, this does notnecessarily indicate any order or importance of the subjects. Instead, such attributes may be used solely for the purpose of making a difference between subjects.

[0140] As used in this application, the term ‘circuitry’ may refer to one or more or all ofthe following: (a) hardware-only circuit implementations (such as implementations in only analog and / or digital circuitry) and (b) combinations of hardware circuits and software, such as (as applicable):(i) a combination of analog and / or digital hardware circuit(s) with software / firmware and (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions) and (c) hardware circuit(s) and or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software maynot be present when it is not needed for operation. This definition of circuitry applies to all uses of this term in this application, including in any claims.

[0141] As a further example, as used in this application, the term circuitry also covers animplementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.

[0142] It will be understood that the above description is given by way of example onlyand that various modifications may be made by those skilled in the art. The above specification, examples and data provide a complete description of the structure and use of exemplary embodiments. Although various embodiments have been described above with a certain degree of particularity, or with reference to one or more individual embodiments, those skilled in the art could make numerous alterations to the disclosed embodiments without departing from scope of this specification.

Claims

CLAIMS 1. A first apparatus, comprising:at least one processor; and at least one memory comprising instruction which, when executed by the at least one processor, cause the apparatus at least to: estimate uplink channel based on reference signals from one or more user equipment; compute a precoding matrix based on the estimated uplink channel for the one or more user equipment; obtain, for the one or more user equipment, a set of pairs of complex vectors and values indicative of strength of effective channel between the first apparatus and the respective user equipment based on the computed precoding matrix; transmit, to a central processor, the obtained one or more sets of the values indicative of strength of effective channel for the one or more user equipment; receive, from the central processor, an indication of which values from the transmitted one or more sets of values are determined to be used for transmit beamforming for the one or more user equipment; and determine, based on the received indication, one or more associated complex vectors of the precoding matrix to be used for transmit beamforming for the one or more user equipment.

2. The first apparatus of claim 1, wherein the first apparatus is further caused to:receive, from the central processor, data symbols associated with the indication; and perform transmit beamforming using at least part of the precoding matrix based on the determined complex vectors and the received data symbols.

3. The first apparatus of any of the preceding claims, wherein the number of computedcomplex vectors and values indicative of strength of the effective channel within a set corresponds to a number of data layers for the respective user equipment.

4. The first apparatus of any of the preceding claims, wherein the uplink channelestimates are obtained for a certain physical resource block granularity.

5. The first apparatus of any of the preceding claims, wherein the one or more sets ofpairs of complex vectors and values are determined for a certain physical resource block granularity.

6. The first apparatus of any of the preceding claims, wherein the reference signalscomprise sounding reference signals.

7. The first apparatus of any of the preceding claims, wherein the values indicative ofstrength of the effective channel comprise singular values generated by the precoding matrix.

8. The first apparatus of any of the preceding claims, wherein the precoding matrix is asingular value decomposition based zero forcing precoder.

9. The first apparatus of any of claims 1 to 6, wherein the precoding matrix comprisesan eigen zero forcing precoder.

10. The first apparatus of any of the preceding claims, further caused to:receive, from the central processor, a power scaling factor, wherein the power scaling factor is a positive real number; and scale at least the complex vectors determined to be used for transmit beamforming based on the power scaling vector.

11. The first apparatus of any of the preceding claims, further caused to:calculate a power scaling factor based on an antenna power constraint; and scale at least the complex vectors determined to be used for transmit beamforming based on the power scaling vector.

12. The first apparatus of any of the preceding claims, wherein a length of the complexvector corresponds to a number of antennas of the first apparatus.

13. The first apparatus of any of the preceding claims, wherein the received indicationcomprises one or more indices of values comprised in the transmitted one or more sets of values.

14. The first apparatus of any preceding claim, further caused to:detect that the transmitted set of values for at least one of the one or more user equipment is being excluded based on the received indication; and recompute the precoding matrix after removing the at least one user equipment with the excluded set of values from consideration.

15. The first apparatus of claim 14, wherein the excluded set of values is detected basedon the received indication comprising a certain index indicative of the excluded set of values for the at least one user equipment.

16. A second apparatus, comprising:at least one processor; and at least one memory comprising instructions which, when executed by the at least one processor, cause the second apparatus at least to: receive, from a plurality of signal processing devices of a base station, one or more sets of values computed independently by the plurality of signal processing devices for one or more user equipment, wherein the values are indicative of strength of effective channel between the respective signal processing device and the one or more user equipment; determine, for each of the one or more user equipment, one or more largest values based on the one or more sets of values received from the plurality of signal processing devices; transmit, to the plurality of signal processing devices, an indication of which values received from the respective signal processing device for the one or more user equipment are to be used for transmit beamforming based on the determined one or more largest values for the one or more user equipment.

17. The second apparatus of claim 16, wherein the number of the one or more largestvalues correspond to a number of data layers of the respective user equipment.

18. The second apparatus of claim 16 or 17, further caused to:determine a composite set of values for each of the one or more user equipment based on the one or more sets of values received from the plurality of signal processing devices; sort the composite set of values in an order of magnitude; and determine the one or more largest values based on the sorted composite set of values.

19. The second apparatus of any of claims 16 to 18, further caused to:transmit, to the plurality of signal processing devices, data symbols associated with the indication.

20. The second apparatus of claim 19, wherein the transmitted data symbols aredetermined based on a number of the indicated values and an order of magnitude of the indicated values within the determined one or more largest values for the respective user equipment.

21. The second apparatus of any of claims 16 to 20, wherein the second apparatuscomprises one of a radio unit of the base station, a distributed unit of the base station, one of the plurality of signal processing devices, or the plurality of signal processing devices wherein each of the signal processing devices is configured to serve as a central processor for non-overlapping frequency bands.

22. The second apparatus of any of the preceding claims, further caused to:determine a power scaling factor for each of the plurality of signal processing devices, wherein the power scaling factor is a positive real number; and transmit, to the plurality of signal processing devices, the determined power scaling factors.

23. The second apparatus of any of claims 16 to 22, wherein the indication comprises atleast one of: -one or more indices of values comprised in the determined one or more largestvalues that were received from the signal processing device; or -a certain index indicative that none of the values received from the signalprocessing device for at least one of the one or more user equipment were included in the determined one or more largest values for the at least one user equipment.

24. A base station, comprising:a plurality of the first apparatuses according to any of claims 1 to 15; and the second apparatus according to any of claims 16 to 23.

25. A computer-implemented method, comprising:estimating uplink channel based on reference signals from one or more user equipment; computing a precoding matrix based on the estimated uplink channel for the one or more user equipment; obtaining, for the one or more user equipment, a set of pairs of complex vectors and values indicative of strength of effective channel between the first apparatus and the respective user equipment based on the computed precoding matrix; transmitting, to a central processor, the obtained one or more sets of the values indicative of strength of effective channel for the one or more user equipment; receiving, from the central processor, an indication of which values from the transmitted one or more sets of values are determined to be used for transmit beamforming for the one or more user equipment; and determining, based on the received indication, one or more associated complex vectors of the precoding matrix to be used for transmit beamforming for the one or more user equipment.

26. A computer-implemented method, comprising:receiving, from a plurality of signal processing devices of a base station, one or more sets of values computed independently by the plurality of signal processing devices for one or more user equipment, wherein the values are indicative of strength of effective channel between the respective signal processing device and the one or more user equipment; determining, for each of the one or more user equipment, one or more largest values based on the one or more sets of values received from the plurality of signal processing devices; and transmitting, to the plurality of signal processing devices, an indication of which values received from the respective signal processing device for the one or more user equipment are to be used for transmit beamforming based on the determined one or more largest values for the one or more user equipment.

Images