Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
APPARATUS, METHOD AND COMPUTER PROGRAM FOR SELECTING CODEWORDS
Document Type and Number:
WIPO Patent Application WO/2012/107904
Kind Code:
A1
Abstract:
From a set of X precoding matrix codewords, there is selected (402) a subset of N codewords so that each selected codeword is optimized for both a cross-polarized antenna array and a co-polarized linear antenna array. Each n th one of the N codewords is associated (404) with a respective n th group of physical resource blocks PRBs which are wirelessly transmitted downlink. N and X are integers, X>N, n indexes through N, and each n th group of PRBs comprises at least one PRB. In various embodiments: each selected codeword is characterized in steering energy in a respective direction and associated with the PRB groups such that no pair of the codewords is associated with adjacent ones of the PRB groups steers energy in the same direction(406); and each of the selected N codewords is formed as a product of two matrices and the selected codewords are cycled among the groups of PRBs (408).

Inventors:
KOIVISTO TOMMI (FI)
ENESCU MIHAI (FI)
Application Number:
PCT/IB2012/050614
Publication Date:
August 16, 2012
Filing Date:
February 10, 2012
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
RENESAS MOBILE CORP
KOIVISTO TOMMI (FI)
ENESCU MIHAI (FI)
International Classes:
H04B7/04
Domestic Patent References:
WO2009091307A12009-07-23
Other References:
NOKIA ET AL: "Detailed PUCCH CSI feedback signaling design", 3GPP DRAFT; R1-105533, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG1, no. Xi'an; 20101011, 6 October 2010 (2010-10-06), XP050450749
NOKIA ET AL: "Double codebook design principles", 3GPP DRAFT; R1-103804, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG1, no. Dresden, Germany; 20100628, 23 June 2010 (2010-06-23), XP050449423
QUALCOMM INCORPORATED: "Remaining details of feedback for TM9", 3GPP TSG-RAN WG1 #63BIS, no. R1-110338, 17 January 2011 (2011-01-17) - 21 January 2011 (2011-01-21), Dublin, Ireland, pages 1 - 10, XP002678667, Retrieved from the Internet [retrieved on 20120620]
"3rd Generation Partnership Project; Technical Specification Group Radio Access Network; Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and modulation (Release 10)", 3GPP STANDARD; 3GPP TS 36.211, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, no. V10.0.0, 22 December 2010 (2010-12-22), pages 1 - 103, XP050462365
LG ELECTRONICS: "Link Performance Comparison between Open-loop SM Candidates", 3GPP DRAFT; R1-080273, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG1, no. Sevilla, Spain; 20080109, 9 January 2008 (2008-01-09), XP050108795
QUALCOMM INCORPORATED: "Remaining aspects of PRB bundling", 3GPP DRAFT; R1-104796 PRB BUNDLING, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG1, no. Madrid, Spain; 20100823, 17 August 2010 (2010-08-17), XP050449988
TEXAS INSTRUMENTS: "Codebook-Based Feedback Issues on 8Tx Downlink MIMO", 3GPP DRAFT; R1-101090 TI 8TX MIMO, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG1, no. San Francisco, USA; 20100222, 16 February 2010 (2010-02-16), XP050418653
3GPP TSG-RAN1 #63BIS MEETING, 17 January 2011 (2011-01-17)
Attorney, Agent or Firm:
FLINT, Adam (Fulwood PlaceLondon, Greater London WC1V 6HU, GB)
Download PDF:
Claims:
CLAIMS

1. Apparatus comprising a processing system constructed and arranged to cause the apparatus to at least:

select a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co-polarized linear antenna array; and

associate each nth one of the N codewords with a respective nth group of physical resource blocks;

in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each nth group of physical resource blocks comprises at least one physical resource block.

2. Apparatus according to claim 1, in which each selected codeword is characterized in steering energy in a respective direction; and

each nth one of the N codewords is associated with its respective nth group of physical resource blocks such that no pair of the codewords associated with adjacent ones of the physical resource block groups steers energy in the same direction. 3. Apparatus according to claim 1 or claim 2, in which each of the selected N codewords is formed as a product of two matrices, and the codewords are cycled among the physical resource blocks.

4. Apparatus according to claim 3, in which N is an integer multiple of 2Y, and one of the matrices comprises a co-phasing term to concatenate the two matrices, each of which is a Y-transmission antenna matrix, into a 2Y-transmission antenna matrix, in which Y is an integer greater than one.

5. Apparatus according to claim 4, in which no pair of the codewords associated with adjacent ones of the physical resource block groups is formed by the same co- phasing term.

6. Apparatus according to any of claims 3 to 5, in which the apparatus comprises a user equipment comprising at least one receive antenna and a transmitter, in which: the at least one receive antenna is configured to receive at least some of the selected codewords which are disposed in the respective groups of physical resource blocks; and

the transmitter is configured to transmit a channel quality indicator computed by the processing system utilizing a downlink transmission scheme which corresponds to the cycled association of the selected codewords with the respective groups of physical resource blocks.

7. Apparatus according to any of claims 1 to 5, in which the apparatus comprises a network access node comprising a plurality of transmit antennas and a transmitter, in which the plurality of transmit antennas are configured with the transmitter to transmit the groups of physical resource blocks in which the respective ones of the selected codewords are disposed.

8. Apparatus according to any of claims 1 to 5, in which the apparatus comprises a modem.

9. A method comprising:

selecting a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co -polarized linear antenna array; and

associating each nth one of the N codewords with a respective nth group of physical resource blocks;

in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each nth group of physical resource blocks comprises at least one physical resource block.

10. A method according to claim 9, in which each selected codeword is characterized in steering energy in a respective direction; and

each nth one of the N codewords is associated with its respective nth group of physical resource blocks such that no pair of the codewords associated with adjacent ones of the physical resource block groups steers energy in the same direction.

11. A method according to claim 9 or claim 10, in which each of the selected N codewords is formed as a product of two matrices, and the selected codewords are cycled among the groups of physical resource blocks.

12. A method according to claim 11, in which N is an integer multiple of 2Y, and one of the matrices comprises a co-phasing term to concatenate the two matrixes, each of which is a Y-transmission antenna matrix, into a 2Y-transmission antenna matrix, in which Y is an integer greater than one.

13. A method according to claim 12, in which no pair of the codewords associated with adjacent ones of the physical resource block groups is formed by the same co- phasing term. 14. A method according to any of claims 11 to 13, in which the method is executed by a user equipment and the method comprises the user equipment:

receiving at least some of the selected codewords which are disposed in the respective groups of physical resource blocks;

computing a channel quality indicator by utilizing a downlink transmission scheme which corresponds to the cycled association of the selected codewords with the respective groups of physical resource blocks; and

transmitting the channel quality indicator.

15. A method according to any of claims 9 to 13, in which the method is executed by a network access node, the method comprising the network access node transmitting from a plurality of transmit antennas the groups of physical resource blocks in which the respective ones of the selected codewords are disposed.

16. A method according to any of claims 9 to 13, in which the method is executed by a modem.

17. A computer program comprising:

code for selecting a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co -polarized linear antenna array; and

code for associating each nth one of the N codewords with a respective nth group of physical resource blocks;

in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each nth group of physical resource blocks comprises at least one physical resource block.

18. A computer program according to claim 17, in which each selected codeword is characterized in steering energy in a respective direction; and

the code for associating is for associating each nth one of the N codewords with its respective nth group of physical resource blocks such that no pair of the codewords associated with adjacent ones of the physical resource block groups steers energy in the same direction.

19. A computer program according to claim 18, in which N is an integer multiple of 2Y, each of the selected N codewords is formed as a product of two matrices, and one of the matrices comprises a co-phasing term to concatenate the two matrices, each of which is a Y-transmission antenna matrix, into a 2Y-transmission antenna matrix, where Y is an integer greater than one.

20. A computer program according to claim 19, in which no pair of the codewords associated with adjacent ones of the physical resource block groups is formed by the same co -phasing term. 21. Apparatus comprising:

means for selecting a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co -polarized linear antenna array; and

means for associating each nth one of the N codewords with a respective nth group of physical resource blocks;

in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each nth group of physical resource blocks comprises at least one physical resource block. 22. Apparatus according to claim 21, in which each selected codeword is characterized in steering energy in a respective direction; and

each nth one of the N codewords is associated with its respective nth group of physical resource blocks such that no pair of the codewords associated with adjacent ones of the physical resource block groups steers energy in the same direction.

23. Apparatus according to claim 21 or claim 22, in which each of the selected N codewords is formed as a product of two matrices, and the selected codewords are cycled among the groups of physical resource blocks. 24. Apparatus according to claim 23, in which N is an integer multiple of 2Y, and one of the matrices comprises a co-phasing term to concatenate the two matrixes, each of which is a Y-transmission antenna matrix, into a 2Y-transmission antenna matrix, in which Y is an integer greater than one.

25. Apparatus according to claim 24, in which no pair of the codewords associated with adjacent ones of the physical resource block groups is formed by the same co- phasing term. 26. Apparatus according to any of claims 23 to 25, in which the apparatus comprises a user equipment comprising:

means for receiving at least some of the selected codewords which are disposed in the respective groups of physical resource blocks;

means for computing a channel quality indicator by utilizing a downlink transmission scheme which corresponds to the cycled association of the selected codewords with the respective groups of physical resource blocks; and

means for transmitting the channel quality indicator.

27. Apparatus according to any of claims 21 to 25, in which the apparatus comprises a network access node, comprising:

means for transmitting from a plurality of transmit antennas the groups of physical resource blocks in which the respective ones of the selected codewords are disposed. 28. Apparatus according to any of claims 21 to 25, in which the apparatus comprises a modem.

Description:
APPARATUS, METHOD AND

COMPUTER PROGRAM FOR SELECTING CODEWORDS

Technical Field

The present invention relates to apparatus, a method and a computer program for selecting codewords. The exemplary and non-limiting embodiments of this invention relate generally to wireless communication systems, methods, devices and computer programs and, in particular embodiments, relate to selecting precoders/codewords for downlink MIMO transmissions such as for example in wireless systems without feedback of precoder information in which the downlink transmission scheme must be estimated in the absence of downlink transmission diversity in order to provide channel quality feedback on the uplink.

Background

The following abbreviations which may be found in the specification and/or the drawing figures are defined as follows:

3 GPP Third Generation Partnership Project

CSI channel state information

CSI-RS channel state information reference symbols

CRS common reference symbols

CQI channel quality indication

DFT discrete Fourier transform

DL downlink

DM-RS demodulation reference symbols

eNB node B/base station in an E-UTRAN system

E-UTRAN evolved UTRAN (LTE)

FSTD frequency- switched transmit diversity

HARQ hybrid automatic repeat request

LTE Long Term Evolution

MIMO multiple input multiple output

PDSCH physical downlink shared channel PMI precoding matrix indication

PRB physical resource block

RI rank indicator

RS reference signal(s)

RX receive or receiver

SFBC space-frequency block code

TDD time division duplex

TX transmit or transmitter

UE user equipment

UL uplink

ULA uniform linear arrays

UTRAN Universal Terrestrial Radio Access Network

XP cross-polarized arrays Some of the changes in LTE Release 10 wireless protocol over previous releases include DL and UL MIMO, enhanced use of relays, bandwidth extensions via carrier aggregation and enhanced inter-cell interference coordination. Relevant to these teachings is DL MIMO; Release 10 supports feedback-based closed- loop spatial multiplexing with up to eight TX antennas (and hence 8x8 MIMO in the context of eight RX antennas) whereas Releases 8/9 supported this for only up to 4 TX antennas. That is, the closed loop precoding for Releases 8 and 9 utilizing a 4 TX antenna codebook has been extended to 8 TX antenna precoding in Release 10. Specifically, in Release 10 there is a special CSI-RS based operation in which the UE computes feedback and sends it to the eNB to support spatial multiplexing. Release 10 defines PMI -based feedback as well as non-PMI -based feedback in which the UE reports only CQI. It follows for those CQI-only reports that since the UL feedback is not PMI- based, the UE must make some assumption about the eNB's transmission scheme on the PDSCH in order to compute CQI. As an overview, closed-loop spatial multiplexing and multi-user MIMO in the

DL for LTE are based on UE feedback, where the UE computes and reports CSI to the eNB in the form of a precoding matrix. LTE Release 8 defines codebooks for two and four TX antennas for feeding back this precoding matrix, and the feedback itself takes the form of an index (termed PMI) of the UE's preferred precoding matrix in the codebook. Release 10 extends this to 8 TX antennas, and defines PMI-based feedback and the codebook for that. In contrast to 2TX and 4TX schemes in Releases 8/9, for the 8 TX antenna case in Release 10 the precoding matrices are formed as a product of two matrices. So the final precoders are of the form W=WiW 2 , where Wi presents the long-term/wideband properties of the radio channel and W 2 captures the short-term/frequency-selective properties of the radio channel. The codebooks are specified in TS 36.21 1 VIO.0.0, section 6.3.4, and in the LTE specifications the double codebook is captured in tables indexed by two indices il and i2. In essence, il corresponds to the long-term/wideband part whereas i2 corresponds to the short- term/frequency-selective part. Additionally, transmit diversity may be used as the PDSCH transmission scheme in Release 8/9, in which case the UE only reports CQI on the assumption that the eNB would transmit with SFBC transmit diversity for 2TX antennas and SFBC- FSTD transmit diversity for 4TX antennas. The CQI which the UE reports is conditioned on some specific transmission scheme. Release 10 does not support transmit diversity in the case of 8TX antennas.

In addition to UE-based feedback operation of closed-loop spatial multiplexing, there is the LTE TDD specific operation of non-PMI based closed-loop spatial multiplexing. In this case, the UE performs UL sounding in order to estimate the eNB's DL channel. This exploits the radio channel's reciprocity property which is a property specific to TDD. Non-PMI feedback has been defined in both Release 8 (for one stream operation) and Release 9 (for two stream operation). For the non-PMI feedback based operation, inter-cell interference is captured in the CQI report which in Release 9 is derived similarly to transmit diversity. In Release 8 and 9, for CQI computation, the UE is assuming 2 TX SFBC operation on 2 common reference symbols CRS ports, or 4Tx SFBC-FSTD operation on 4 CRS ports. While only PMI-based feedback has been specified for Release 10 to date, recent discussions concern PMI feedback disabling, meaning the UE would report only CQI. To compute the CQI, the UE needs to know or assume the eNB 's DL transmission scheme, which in the transmission diversity examples above for Release 8/9 is termed the PDSCH reference transmission scheme. As in those examples, the UE bases its assumption for the reference transmission scheme on transmit diversity when the UE reports only CQI. But for the 8 TX antenna transmissions of Release 10, there is no transmit diversity scheme defined. To ensure correct operation at the UE side when computing the CQI, it should also be possible to transmit with the reference transmission scheme. This would ensure the RAN4 testability of the feature.

One potential solution for the reference transmission scheme is to employ precoder cycling, in which the precoder is changed from PRB to PRB but kept constant within each PRB to allow proper channel estimation for demodulation at the UE side. Precoder cycling is known from Release 8 where it is used for open-loop spatial multiplexing in transmission mode 3 (see 3GPP TS 36.21 1 VIO.0.0, section 6.3.4.2.2). One specific proposal to utilize precoder cycling in Release 10 is at document Rl-1 10338 by Qualcomm, entitled Remaining details for feedback for TM9 (3 GPP TSG-RANl #63bis meeting; 17-21 January 201 1 ; Dublin, Ireland) which proposes a new feedback mode based on precoder cycling per PRB. In the context of Release 10 8Tx codebook utilization, for a rank 1 transmission there are 256 final codewords W in total in the codebook; 16 Wi codewords multiplied by 16 W 2 codewords. The inventors consider this far too many precoders to cycle in a practical system, and hence a method is needed for downselecting the codewords to be cycled from the full codebook.

In that same meeting was proposed utilizing a fixed precoder for calculating CQI. The inventors do not consider this as sufficiently reliable. Specifically, the fixed beam would need to be known by both the UE and the eNB. If that could somehow be resolved it appears there would be a high probability that this fixed beam would steer the energy into the wrong direction as compared to the UE signal space, leading to a large number of pessimistic CQI values. It is not expected that such high CQI errors can be reasonably corrected with open loop link adaptation techniques which are typically utilized for CQI refinement.

Summary

According to a first aspect of the present invention, there is provided apparatus comprising a processing system constructed and arranged to cause the apparatus to at least: select a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co-polarized linear antenna array; and associate each n th one of the N codewords with a respective n th group of physical resource blocks; in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each n th group of physical resource blocks comprises at least one physical resource block.

The processing system may comprise at least one processor and at least one memory storing a computer program, the at least one memory with the computer program being configured with the at least one processor to cause the apparatus to at least operate as described above.

According to a second aspect of the present invention, there is provided a method comprising: selecting a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co-polarized linear antenna array; and associating each n th one of the N codewords with a respective n th group of physical resource blocks; in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each n th group of physical resource blocks comprises at least one physical resource block. According to a third aspect of the present invention, there is provided a computer program, in which the computer program comprises: code for selecting a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co-polarized linear antenna array; and code for associating each n th one of the N codewords with a respective n th group of physical resource blocks; in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each n th group of physical resource blocks comprises at least one physical resource block. There may be provided a computer-readable memory storing a computer program as described above.

According to a fourth aspect of the present invention, there is provided apparatus comprising: means for selecting a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co -polarized linear antenna array; and means for associating each n th one of the N codewords with a respective n th group of physical resource blocks; in which N is an integer greater than one, X is an integer greater than N, n indexes through N, and each n th group of physical resource blocks comprises at least one physical resource block.

Specific examples of embodiments of the present invention described below address how to enable CQI-only feedback for the 8 TX antenna scheme of Release 10 and particularly a reference transmission scheme from which the UE can compute that CQI. As a reference transmission scheme it is preferable to be specified in wireless standards, but the broader teachings herein are useful also as a standard-transparent transmit diversity scheme for DL transmissions and therefore need not be specified in a written wireless protocol since these broader teachings can be implemented purely within the access node/eNB. Further features and advantages of the invention will become apparent from the following description of preferred embodiments of the invention, given by way of example only, which is made with reference to the accompanying drawings. Brief Description of the Drawings

Figure 1 shows a chart showing schematically a set of 8 precoders selected according to an exemplary embodiment of the invention in which adjacent beams are cycled in a consecutive manner; Figure 2 shows a chart similar to Figure 1 but with 16 precoders in the selected set;

Figure 3 shows a chart similar to Figure 1 in which adjacent beams are cycled in a randomized manner for increased diversity in CQI computation;

Figure 4 shows a schematic logic flow diagram that illustrates the operation of a method, and a result of execution of computer program instructions embodied on a computer readable memory, in accordance with the exemplary embodiments of this invention; and

Figure 5 shows a simplified block diagram of an example of a UE in communication with a wireless network illustrated as an eNB and a serving gateway SGW, which are exemplary electronic devices suitable for use in practising the exemplary embodiments of this invention.

Detailed Description

Exemplary embodiments of these teachings address the assumed PDSCH transmission scheme when computing the CQI in non-PMI-based feedback modes in case of 8 TX antennas at the eNB. As noted above these teachings may also be utilized in standard-transparent demodulation reference signals (DM-RS) based transmit diversity schemes. To better understand the advantages of these exemplary embodiments for Release 10 deployment, first consider the nature of the Release 10 codebooks. Wi is in essence a group of 4TX beams when applied on a 4TX antenna array, with elements spaced by half of the wavelength. In other words, each beam (which results from the 4 individual antenna weights of the 4 TX antennas) is steering the energy towards a certain physical direction. This structure is particularly appealing for 8TX cross polarized antennas where there are 4 antenna branches per polarization. W 2 is a set of matrices which selects the optimum beams from Wi and which uses a co- phasing term to concatenate two 4TX matrices into a 8TX precoding matrix. The codebooks are chosen to optimize the precoding performance in typical 8TX antenna arrays which may be either cross-polarized arrays of 4TX antennas per polarization, or co-polarized uniform linear arrays of 8TX antennas (ULA). Now consider the precoder cycling solution noted in the background section.

Since as stated there it is too burdensome to cycle among the 256 final codewords W in total within the codebook (16 Wi codewords and 16 W 2 codewords), the problem becomes which and how many codewords to select for cycling, and how to select them. The examples below variously show a set of 8 and a set of 16 codewords selected for precoder cycling. In these examples they are selected from the Release 10 codebook, but this is simply for a specific illustration as these teachings are not limited to the CQI-only feedback mechanism for 8 TX antennas of Release 10 in specific or to the LTE system in general. They are selected to overcome the problem stated above with the fixed precoder which might steer the beam energy too frequently in the wrong direction.

The Release 10 codebook has been designed to operate for both cross- polarized arrays XP and uniform linear arrays ULA, and so contains codewords for both configurations. But the 8TX antenna codebook is based on 4TX beams and also co-phasing terms to combine the 4TX beams into 8TX codewords, which the inventors have determined make it particularly advantageous for cross-polarized antenna arrays. Additionally, some of the beam/co-phasing combinations W=WiW 2 also directly form 8TX antenna beams, and are therefore particularly useful for co- polarized ULA antenna arrays. There is a well known DFT butterfly technique which is discussed further below and which can be employed in this case to obtain 8TX DFT vectors (beams) from two 4TX DFT vectors. Note that this is not true for all possible combinations W=WiW 2 but only a specific subset of all 8TX ULA codewords obtained as a combination W=WiW 2 as is discussed below.

If we consider all the potential precoding matrix codewords as the set X (X=256 in the above example), then the selected subset of N codewords in the below examples is N=8 or N=16 selected codewords, and so X>N since less than all of the X=256 possible codewords are optimized for both cross-polarized arrays XPs and uniform linear arrays ULAs. Now consider an exemplary but non-limiting technique for how the selection of the optimized codewords may be done, directed specifically to how the 8 TX antenna array codebook for LTE Release 10 is formed. In cross-polarized antenna arrays (XP), the 8TX array is formed as two 4TX sub-arrays, each of which uses different polarization. The antennas of each sub-array are in practice typically spaced by half of the wavelength, with wavelength represented by λ as is conventional.

< >

λ/2

In co-polarized uniform linear antenna arrays (ULA), the 8TX array is formed simply by having eight antennas, spaced typically in practice by half of the wavelength, each using the same polarization.

For an antenna array with elements spaced by half of the wavelength λ, an optimum codebook is formed with DFT (Discrete Fourier Transform) vectors. A DFT vector is expressed as:

where M is vector length (4 or 8 in this case), N=QM is the number of beams, Q is the spatial oversampling factor, n is beam (rank-1 codeword) index and i is the imaginary unit. Following this we can write for example DFT-4 vectors (rank-1 codewords) with oversampling factor =4 as:

These would constitute optimum rank-1 precoders for each 4-Tx sub-array of the 8TX cross-polarized antenna array. To get full 8TX codewords, one needs to have a phase combiner to concatenate two DFT-4 vectors into one 8TX codeword, and so the final codewords are formed as: with the phase factor φ selected from e.g. a QPSK codebook as in LTE Release 10, φ E {1, -1, £,—£}. On the other hand, for 8TX ULA the optimum codewords are DFT-8 vectors. With oversampling factor Q=2 these vectors are expressed as:

Comparing the two previous equations, it can be seen that whenever for

codeword n the XP combiner phase is selected as φ =— n, the resulting 8-Tx codeword corresponds to a DFT-8 vector, and hence is optimum for an 8Tx ULA array (when the antennas are spaced by half of the wavelength).

In other words, the codeword is optimized for both XP and ULA when the

(4,16)

combination of beam v. and phase e l<p is selected such that the resulting codeword matches DFT-8. In this case the codeword corresponding to each 4-TX sub-array is a DFT-4 vector and the full codeword corresponding to the full 8Tx array is a DFT-8 vector v^ ' 16 ^ .

Regarding 8TX double codebook used in LTE Release 10, only well-selected combinations of indices il and i2 fulfil the above condition, i.e. that the phase is selected such as to make the full 8-TX codeword to be a DFT-8 vector. In an embodiment discussed in more detail below, precoder cycling is done through the precoders that fulfil this condition. Note that the needed phase depends on the beam (DFT vector) index n, and hence for different beams a different phase has to be chosen. In this context, the DFT butterfly technique as mentioned above is a technique for selecting a phase combiner such that two DFT-4 vectors are used to form a DFT-8 vector, as detailed in the above analysis. More generally the DFT butterfly refers to an efficient way of implementing DFTs/FFTs.

In an exemplary embodiment of the invention particular to LTE Release 10, there is selected the precoder indices il and i2 such that the resulting precoders that are cycled through correspond exactly to the ones that are optimal for both cross- and co-polarized arrays. Each of the selected codewords is associated with a PRB, or a group of PRBs. For clarity, consider that the eNB disposes each n th selected codeword within a PRB group with which that codeword is associated, and each n th one of the N PRB groups may have one or more than one PRB. Thus for every PRB group there is a pre-defined precoder which is selected from this subset of {il, il) index pairs which correspond to the selected codewords, each of which is optimized for both ULA and XP antenna arrays. So long as both the UE and the eNB know in advance what these pairs are, the UE will know the precoders which the eNB will use in each PRB group it transmits and thus the eNB's transmission scheme. Knowing the transmission scheme enables the UE to accurately compute CQI, without having to initially report or suggest a PMI for the eNB to use.

From the optimization analysis above, for the XP case the optimum codeword for each 4-TX subarray is a DFT-4 vector, and for the ULA case the optimum 8-TX codeword is a DFT-8 vector. So a codeword which is optimized for both XP and ULA fulfils both conditions, i.e. that each 4-TX part of the full 8-TX codeword is a DFT-4 vector and that the full 8-TX codeword is a DFT-8 vector. Specifically for the Release 10 LTE double codebook, this is fulfilled in practice by properly selecting correct (il,i2) codeword combinations.

Conceptually, each beam steers the radiated (transmission) energy towards a certain direction, which may or may not be the correct direction to reach the intended UE. To resolve this issue, the precoders are selected and disposed such that the precoder beam direction in consecutive PRB groups are as different as possible. The cycling of precoders therefore randomizes the beam directions. This does not mean that the decision of which precoder to associate with which PRB group is itself a random decision; apart from the first PRB group it is not since precoder directions of adjacent PRB groups should be different. For example, adjacent beams should not be selected to lie in adjacent PRB groups. This aspect of the exemplary embodiments ensures that even where the beam energy of one PRB group's precoder points in a direction which the UE cannot easily receive, the beam energy of the precoders in both adjacent PRB groups is likely to be properly received at that same UE. This arrangement should provide maximum diversity.

Figures 1 to 3 illustrate specific examples for the 8 TX antenna codebook used in LTE Release 10. Index il selects the beam from the first channel matrix Wi which represents the long-term/wideband properties of the radio channel, and index i2 selects the co-phasing from the second channel matrix W 2 which represents the short- term/frequency-selective properties of the radio channel. The individual precoders are the product W=WiW 2 in which Wi and W 2 are identified by the index pair {il, il), and the selected set of precoders according to the embodiments shown in Figures 1 to 3 are indicated by numbering at the shaded intersection of the index pair. Note that any of the shaded intersections lying along the diagonals are appropriate candidates for precoders according to the exemplary embodiments; the selected set of N codewords at Figure 2 includes all 16 viable precoder candidates whereas the selected sets of N codewords at Figures 1 and 3 have only 8 of the total 16 viable precoder candidates. Regardless, each n th codeword selected from the overall set X=256 for the subset of size N is optimized for XP and ULA.

A precoder/codeword is a viable candidate if it is optimal for both cross- polarized and co-polarized arrays, but not all of the viable precoders/codewords need be selected for a given implementation as Figures 1 and 3 illustrate. In other embodiments, the size and makeup of the underlying matrices Wi and W 2 may be different from that in Release 10 and so the viable precoder/codeword candidates may be more or fewer than 16, and may or may not line up along diagonals as clearly as in Figures 1 to 3.

In the Figure 1 example, the selected 8 precoders are cycled through in a consecutive manner according to the numbers at the index pair intersection. So for example if we begin with disposing the precoder defined by index pair (0, [1;1]) in PRB groupl, then in the next consecutive PRB_group2 would be disposed the precoder defined by index pair (4, [1;-1]), and the next consecutive PRB_group3 will carry the precoder defined by index pair (8, [1;1]), and so forth.

The example of Figure 2 is similar to that of Figure 1 but all 16 of the precoders which are optimized for both cross-polarized and co-polarized arrays are used for cycling, and the cycling among the precoders is consecutive. So for example the precoder defined by index pair (0, [1;1]) is in PRB groupl, the precoder defined by index pair (2, [l;i]) is in the next consecutive PRB_group2, followed by the precoder defined by index pair (4; [1,-1]) in the next consecutive PRB_group3, and so forth.

Figure 3 shows an example with 8 precoders in the selected set, identical to those at Figure 1. But in the example of Figure 3, the precoders and thus the beam directions are cycled through in a randomized way to provide maximum diversity in CQI computation. For the example of Figure 3, the precoder defined by index pair (0, [1;1]) is put in PRB groupl (identical to figure 1), then in the next consecutive PRB_group2 would be disposed the precoder defined by index pair (20, [1;-1]), the next consecutive PRB_group3 has the precoder defined by index pair (8, [1 ; 1 ]), followed by PRB_group4 in which is disposed the precoder defined by index pair (28, [1;-1]), and so forth.

Note that for each of the examples of Figures 1 to 3, no pair of precoders/codewords which would be placed in adjacent PRB groups is formed by the same co-phasing term from the W 2 matrix. From the UE perspective, for non-PMI feedback the UE would assume precoding according to the description above, and report CQI for that in the UL. The eNB may then utilize the CQI for transmitting in the DL with the same precoding cycling scheme. Or alternatively the eNB may utilize that CQI for transmitting DL with TDD reciprocity-based beamforming, in which case the eNB may make an adjustment to the CQI reported by the UE to account for the beamforming gain and the number of beams While the above examples are specific to rank 1 transmissions in LTE Release

10, these teachings are of course equally advantageous for use in higher rank transmissions (open loop spatial multiplexing). For deployment in such higher rank transmission scenarios, in an exemplary embodiment the UE reports rank indicator (RI) in addition to the CQI.

While the above examples are specific to 8Tx double codebook transmissions in LTE Release 10, these teachings are of course equally advantageous for use when different number of antennas, less or greater than 8, are considered while optimization for ULA and XP antennas arrays is desired. The double codebook implementation may be stated more generally in that one matrix has a co-phasing term to concatenate two Y-transmission antenna matrices into a 2Y-transmission antenna matrix, and the selected set of precoders is then N where N is an integer multiple of 2Y and Y is an integer greater than one. Exemplary embodiments of these teachings exhibit the technical effect of providing a testable reference transmission scheme for CQI computations in non- PMI-based LTE feedback modes. As compared to fixed precoding, another technical effect of these embodiments is improved performance and utilization beyond CQI calculation purposes for the actual DL (e.g. PDSCH) transmission. For such DL transmissions, it is not necessary to stipulate the selected precoder set or the cycling order in a wireless standard/protocol, unlike the CQI reporting functionality for which those parameters should be commonly understood in advance among the UE and eNB.

Figure 4 is a logic flow diagram which describes an exemplary embodiment of the invention in a manner which may be from the perspective of the UE or from the eNB. Figure 4 may be considered to illustrate the operation of a method, and a result of execution of a computer program stored in a computer readable memory, and a specific manner in which components of an electronic device are configured to cause that electronic device to operate. The various blocks shown in Figure 4 may also be considered as a plurality of coupled logic circuit elements constructed to carry out the associated function(s), or specific result of strings of computer program code stored in a memory.

Such blocks and the functions they represent are non-limiting examples, and may be practised in various components such as integrated circuit chips and modules, and the exemplary embodiments of this invention may be realized in an apparatus that is embodied as an integrated circuit. The integrated circuit, or circuits, may comprise circuitry (as well as possibly firmware) for embodying at least one or more of a data processor or data processors, a digital signal processor or processors, baseband circuitry and radio frequency circuitry that are configurable so as to operate in accordance with the exemplary embodiments of this invention.

At block 402 there is selected a subset of N codewords from a set of X precoding matrix codewords, each selected codeword being optimized for both a cross-polarized antenna array and a co-polarized linear antenna array. At block 404 each n th one of the N codewords is associated with a respective n th group of physical resource blocks. As in the above examples, those PRB groups are wirelessly transmitted downlink by the eNB. In this characterization of the exemplary embodiments, N is an integer greater than one, X is an integer greater than N, n indexes through N, and each n th group of physical resource blocks comprises at least one physical resource block. By example, N=8 in the example of Figures 1 and 3, and N=16 in the example of Figure 2. As noted above, each of the selected N codewords is optimized for both a cross-polarized antenna array and a co -polarized linear antenna array, and in the examples for LTE Release 10 each of the selected N codewords is formed as a product of two matrices and the codewords are cycled among the physical resource blocks.

The remainder of Figure 4 illustrates more specific implementations for blocks 402 and 404. Block 406 recites as detailed above that each selected codeword is characterized in steering energy in a respective direction, and each n th one of the N codewords is associated with a respective n th group of PRBs such that no pair of the codewords associated with adjacent ones of the PRB groups steers energy in the same direction. Block 408 details the cycling for the dual-matrix examples detailed above; each of the selected N codewords is formed as a product of two matrices, and the selected codewords are cycled among the groups of physical resource blocks. At block 410 which is specific for the 8TX antenna matrix from the above examples, N is an integer multiple of eight, and one of the matrices of block 408 comprises a co- phasing term to concatenate two 4-transmission antenna matrices into an 8-transmission antenna matrix. Block 410 might be stated more generally that the matrices are each Y-transmission antenna matrices, the concatenation results in a 2Y-transmission antenna matrix, and N is an integer multiple of 2Y where Y is itself an integer greater than one. Further detail of block 410 is shown at block 412, in which no pair of the codewords associated with adjacent ones of the physical resource block groups is formed by the same co-phasing term. For embodiments in which Figure 4 is from the perspective of a user equipment, such a UE may also include at least one receive antenna and a transmitter. The transmitter is configured to transmit a CQI computed by the UE's processor by utilizing a DL transmission scheme which corresponds to the cycled association of the selected codewords with the respective groups of PRBs. For embodiments in which Figure 4 is from the perspective of an eNB or more generally a network access node, such an access node may comprise a plurality of transmit antennas and a transmitter, in which the plurality of transmit antennas are configured with the transmitter to transmit the groups of PRBs in which the respective ones of the selected codewords are disposed.

More generally, Figure 4 may be considered to reflect a modem or operation of a modem which may be apart from or disposed in the above UE or eNB. Reference is now made to Figure 5 for illustrating a simplified block diagram of examples of various electronic devices and apparatus that are suitable for use in practising the exemplary embodiments of this invention. In Figure 5, a wireless network (eNB 22 and mobility management entity MME/serving gateway SGW 24) is adapted for communication over a wireless link 21 with an apparatus, such as a mobile terminal or UE 20, via a network access node, such as a base or relay station or more specifically an eNB 22. The network may include a network control element MME/SGW 24, which provides connectivity with further networks (e.g. a publicly switched telephone network PSTN and/or a data communications network/Internet). The UE 20 includes processing means such as at least one data processor (DP)

20A, storing means such as at least one computer-readable memory (MEM) 20B storing at least one computer program (PROG) 20C, communicating means such as a transmitter TX 20D and a receiver RX 20E for bidirectional wireless communications with the eNB 22 via one or more antennas 20F (8 RX antennas being schematically shown but there may be as few as one RX antenna in certain embodiments). Also stored in the MEM 20B at reference number 20G is an algorithm for selecting the set of precoders and the association with the PRB groups, or for the case such selection may be standardized simply the set of precoders and the PRB or grouped PRB associations as detailed in the examples above. The eNB 22 also includes processing means such as at least one data processor (DP) 22A, storing means such as at least one computer-readable memory (MEM) 22B storing at least one computer program (PROG) 22C, and communicating means such as a transmitter TX 22D and a receiver RX 22E for bidirectional wireless communications with the UE 20 via one or more antennas 22F (8 TX antennas being schematically shown as in the above examples though these teachings may be utilized with 4 or some other number of TX antennas). There is a data and/or control path 25 coupling the eNB 22 with the MME/SGW 24, and another data and/or control path 23 coupling the eNB 22 to other eNBs/access nodes. The eNB 22 stores the algorithm 22G for selecting the set of precoders and the association with the PRB groups, or for the case such selection may be standardized simply the set of precoders and their respective associated PRB or PRB group such as detailed in the examples above.

Similarly, the MME/SGW 24 includes processing means such as at least one data processor (DP) 24A, storing means such as at least one computer-readable memory (MEM) 24B storing at least one computer program (PROG) 24C, and communicating means such as a modem 24H for bidirectional wireless communications with the eNB 22 via the data/control path 25. While not particularly illustrated for the UE 20 or eNB 22, those devices are also assumed to include as part of their wireless communicating means a modem which may be inbuilt on an RF front end chip within those devices 20, 22 and which also carries the TX 20D/22D and the RX 20E/22E.

At least one of the PROGs 20C in the UE 20 is assumed to include program instructions that, when executed by the associated DP 20A, enable the device to operate in accordance with the exemplary embodiments of this invention, as detailed above. The eNB 22 may also have software stored in its MEM 22B to implement certain aspects of these teachings as detailed above. In these regards, the exemplary embodiments of this invention may be implemented at least in part by computer software stored on the MEM 20B, 22B which is executable by the DP 20A of the UE 20 and/or by the DP 22A of the eNB 22, or by hardware, or by a combination of tangibly stored software and hardware (and tangibly stored firmware). Electronic devices implementing these aspects of the invention need not be the entire UE 20 or eNB 22, but exemplary embodiments may be implemented by one or more components of same such as the above described tangibly stored software, hardware, firmware and DP, a system-on-a-chip SOC or an application specific integrated circuit ASIC.

In general, the various embodiments of the UE 20 can include but are not limited to personal portable digital devices having wireless communication capabilities, including but not limited to cellular telephones, navigation devices, laptop/palmtop/tablet computers, digital cameras and music devices, and Internet appliances.

Various embodiments of the computer readable MEMs 20B and 22B include any data storage technology type which is suitable to the local technical environment, including but not limited to semiconductor based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory, removable memory, disc memory, flash memory, DRAM, SRAM, EEPROM and the like. Various embodiments of the DPs 20A and 22A include but are not limited to general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs) and multi-core processors.

Various modifications and adaptations to the foregoing exemplary embodiments of this invention may become apparent to those skilled in the relevant arts in view of the foregoing description. While the exemplary embodiments have been described above in the context of the LTE Release 10 system, it should be appreciated that the exemplary embodiments of this invention are not limited for use with only this one particular type of wireless communication system, and that they may be used to advantage in other wireless communication systems such as for example UTRAN, GERAN and GSM and others. The above embodiments are to be understood as illustrative examples of the invention. Further embodiments of the invention are envisaged. It is to be understood that any feature described in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodiments, or any combination of any other of the embodiments. Furthermore, equivalents and modifications not described above may also be employed without departing from the scope of the invention, which is defined in the accompanying claims.