Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02731210 2011-01-17
WO 2010/018969 1
PCT/KR2009/004468
METHOD AND APPARATUS FOR TRANSMITTING UPLINK SIGNALS USING
MULTI-ANTENNA
BACKGROUND OF THE INVENTION
Field of the Invention
[0001] The present invention relates to a wireless mobile
communication system, and more particularly to a
communication system based on a Multiple Input Multiple
Output (MIMO) scheme.
Discussion of the Related Art
[0002] MIMO technology is an abbreviation for Multiple
Input Multiple Output technology.
MIMO technology uses a
plurality of transmission (Tx) antennas and a plurality of
reception (Rx) antennas to improve the efficiency of
transmission and reception (Tx/Rx) of data. In other words,
MIMO technology allows a transmission end or reception end
of a wireless communication system to use multiple antennas
(hereinafter referred to as a multi-antenna), so that the
capacity or performance can be improved. For convenience of
description, the term "MIMO" can also be considered to be a
multi-antenna technology.
[0003] In more detail, MIMO technology is not dependent
on a single antenna path to receive a single total message.
Instead, the MIMO technology collects a plurality of data
CA 02731210 2011-01-17
WO 2010/018969 2
PCT/KR2009/004468
fragments received via several antennas, merges the
collected data fragments, and completes total data.
As a
result, MIMO technology can increase a data transfer rate
within a predetermined-sized cell region, or can increase
system coverage while guaranteeing a specific data transfer
rate. Under this situation, MIMO technology can be widely
applied to mobile communication terminals, repeaters, or the
like.
MIMO technology can extend the range of data
communication, so that it can overcome the limited amount of
transmission (Tx) data of mobile communication systems.
[0004]
FIG. 1 is a block diagram illustrating a general
MIMO communication system.
[0005] Referring to FIG. 1, the number of transmission
(Tx) antennas in a transmitter is NT, and the number of
reception (Rx) antennas in a receiver is NR. In
this way,
theoretical channel transmission capacity of the MIMO
communication system when both the transmitter and the
receiver use a plurality of antennas is greater than that of
another case in which only the transmitter or the receiver
uses several antennas. The theoretical channel transmission
capacity of the MIMO communication system increases in
proportion to the number of antennas.
Therefore, data
transfer rate and frequency efficiency are greatly increased.
Provided that a maximum data transfer rate acquired when a
single antenna is used is set to Ro, a data transfer rate
CA 02731210 2011-01-17
WO 2010/018969 3
PCT/KR2009/004468
acquired when multiple antennas are used can theoretically
increase by a predetermined amount that corresponds to the
maximum data transfer rate (R0) multiplied by a rate of
increase Ri. The rate of increase (Ri) can be represented by
the following equation 1.
[0006] [Equation 1]
R. =min(NT,NR)
[0007] For example, provided that a MIMO system uses four
transmission (Tx) antennas and four reception (Rx) antennas,
the MIMO system can theoretically acquire a high data
transfer rate which is four times higher than that of a
single antenna system. After the above-mentioned
theoretical capacity increase of the MIMO system was
demonstrated in the mid-1990s, many developers began to
conduct intensive research into a variety of technologies
which can substantially increase a data transfer rate using
the theoretical capacity increase.
Some of the above
technologies have been reflected in a variety of wireless
communication standards, for example, a third-generation
mobile communication or a next-generation wireless LAN, etc.
[0008] The above-mentioned MIMO technology can be
classified into a spatial diversity scheme (also called a
Transmit Diversity scheme) and a spatial multiplexing scheme.
The spatial diversity scheme increases transmission
reliability using symbols passing various channel paths.
CA 02731210 2011-01-17
WO 2010/018969 4
PCT/KR2009/004468
The spatial multiplexing scheme simultaneously transmits a
plurality of data symbols via a plurality of transmission
(Tx) antennas, so that it increases a transfer rate of data.
In addition, the combination of the spatial diversity scheme
and the spatial multiplexing scheme has also been recently
developed to properly acquire unique advantages of the two
schemes.
[0009] In association with the MIMO technology, a variety
of MIMO-associated technologies have been intensively
researched by many companies or developers, for example,
research into an information theory associated with a MIMO
communication capacity calculation under various channel
environments or multiple access environments, research into
radio frequency (RF) channel measurement and modeling of the
MIMO system, and research into a space-time signal
processing technology for increasing
transmission
reliability and data transfer rate.
[0010] In a 3rd Generation Partnership Project Long Term
Evolution (3GPP LTE) system, the above-mentioned MIMO scheme
is applied to only downlink signal transmission of the 3GPP
LTE system.
The MIMO technology may also be applied to
uplink signal transmission.
In this case, a transmitter
structure is changed to implement the MIMO technology, so
that a Peak power to Average Power Ratio (PAPR) or Cubic
Metric (CM) characteristics may be deteriorated. Therefore,
CA 02731210 2013-10-15
74420-482
there is needed a new technology capable of effectively
applying the MIMO scheme to uplink signal transmission.
SUMMARY OF THE INVENTION
[0011] Accordingly, an embodiment of the present invention
5 is directed to a method and apparatus for transmitting uplink
signals via multiple antennas that substantially obviate one or
more problems due to limitations and disadvantages of the
related art.
[0012] An embodiment of the present invention may provide a
technology for effectively carrying out uplink signal
transmission according to a MIMO scheme.
[0013] Additional features will be set forth in part in the
description which follows and in part will become apparent to
those having ordinary skill in the art upon examination of the
following or may be learned from practice of the invention.
Embodiments of the invention may be realized and attained by
the structure particularly pointed out in the written
description and claims hereof as well as the appended drawings.
[0013a] In an aspect, there is provided a method for
controlling a user equipment (UE) to transmit uplink signals
via multiple antennas, the method comprising: mapping the
uplink signals to a predetermined number of layers; performing
Discrete Fourier Transform (DFT) spreading upon each of the
predetermined number of layers; precoding each of the DFT-
spread layers by using a specific precoding matrix selected
from among a prestored codebook, wherein every precoding matrix
in the prestored codebook is established in a manner that each
one of the multiple antennas transmits no more than one layer
CA 02731210 2013-10-15
74420-482
5a
of the predetermined number of layers, and wherein the
prestored codebook comprises a precoding matrix established in
a manner that one layer is transmitted via two or more
antennas; and transmitting the precoded signals to a base
station (BS) via the multiple antennas by performing a
predetermined process for constructing a Single Carrier-
Frequency Division Multiple Access (SC-FDMA) symbol upon the
precoded signals.
[0013b] There is also provided a user equipment (UE) for
transmitting uplink signals via multiple antennas comprising:
multiple antennas for transmitting and receiving signals; a
memory for storing a codebook, wherein every precoding matrix
in the codebook is established in a manner that each of the
multiple antennas transmits no more than one layer of the
predetermined number of layers, and wherein the prestored
codebook comprises a precoding matrix established in a manner
that one layer is transmitted via two or more antennas; and a
processor connected to the multiple antennas and the memory so
as to process transmission of the uplink signals, wherein the
processor includes: a layer mapper for mapping the uplink
signals to a predetermined number of layers corresponding to a
specific rank; a Discrete Fourier Transform (DFT) module for
performing DFT spreading upon each of the predetermined number
of layers; a precoder for precoding each of the DFT-spread
layer signals received from the DFT module by selecting a
specific precoding matrix from among the codebook stored in the
memory; and a transmission module for performing a
predetermined process for constructing a Single Carrier-
Frequency Division Multiple Access (SC-FDMA) symbol upon the
precoded signals, and transmitting the processed signals to a
base station (BS) via the multiple antennas.
CA 02731210 2013-10-15
' 74420-482
5b
[0014] A method for enabling a user equipment (UE) to
transmit uplink signals via multiple
CA 02731210 2011-01-17
WO 2010/018969 6
PCT/KR2009/004468
antennas includes mapping the uplink signals to a
predetermined number of layers, performing Discrete Fourier
Transform (DFT) spreading upon each of the predetermined
number of layer signals, precoding the DFT-spread layer
signals by selecting a specific precoding matrix established.
in a manner that one layer signal is transmitted to each of
the multiple antennas from among a prestored codebook, and
performing a predetermined process for constructing a Single
Carrier-Frequency Division Multiple Access (SC-FDMA) symbol
upon the precoded signals, and transmitting the processed
signals to a base station (BS) via the multiple antennas.
[0015] The specific precoding matrix may be a precoding
matrix established in a manner that the multiple antennas
have uniform transmission power therebetween.
The specific
precoding matrix may be a precoding matrix established in a
manner that the predetermined number of layers have uniform
transmission power therebetween
[0016] The codebook may include a first type precoding
matrix, wherein the first type precoding matrix may be
1 0
x0
0 1
configured in a form of - Y- , as a Rank 2 precoding matrix
utilized when the number of the multiple antennas is 4 and a
rank is set to 2, and may satisfy a condition of
CA 02731210 2011-01-17
WO 2010/018969 7
PCT/KR2009/004468
{
1.-E j i.¨ j i ¨1¨j. -1+31
11 , i E 1, ¨,¨ , j, , 1, __ r- , j, ,---
V2 42 42 112
- . The Rank 2 precoding
matrix may further include a precoding matrix generated when
positions of individual rows of the first type precoding
matrix are changed.
[0017] The Rank 2 precoding matrix may further include a
-1 0.-
if
A' 0
0
second type precoding matrix configured in a form of
and a third type precoding matrix configured in a form of
1 0
0 1
0. r.
.. .
!Fr a :i
_.....,, where individual rows of the precoding matrix may
respectively correspond to four antennas of the multiple
10 antennas, and individual columns may respectively correspond
to layers.
[0018] The Rank 2 precoding matrix may further include a
precoding matrix generated when positions of individual
columns of the first type precoding matrix are changed.
[0019] The codebook may include a first type precoding
matrix, wherein the first type precoding matrix, serving as a
Rank 3 precoding matrix utilized when the number of the
multiple antennas is 4 and a rank is set to 3, is configured
CA 02731210 2011-01-17
WO 2010/018969 8 PCT/KR2009/004468
[ -1 0- 47
0100: :
0 0:- I '
. :r,
in a form of ::.... Jr 6 o
, and satisfies a condition of
-
1 ................................ .
:.. 14 t i'..,1 i.. -1+) .
17:','.=E 1, ¨, _ . i::1;: ,¨. ,:.¨i;i; = ' ,... . ,::¨ J.
--:. ,... =
. . . : . 42: . .42: . .42 . 42
......................... - : .
The Rank 3 precoding
matrix may further include a precoding matrix generated when
positions of individual rows of the first type precoding
matrix are changed. The Rank 3 precoding matrix may further
include a precoding matrix generated when positions of
individual columns of the first type precoding matrix are
changed. That is, the codebook may includes a precoding
matrix configured to alternatively map a first layer to first
and second antennas and second and third layers to third and
fourth antennas, respectively, as the precoding matrix used
for the case when the number of antennas is 4 and the rank is
3.
[0020] When the number of antennas is 4, the Rank is 3,
and the number of codewords is 2, one of the codeword is
mapped to a single layer, and the other codeword is mapped to
two layers. The precoding matrix can be configured so that
the total transmission power from the layer perspective may
be different in order to enforce uniform transmission power
between multiple antennas. In such a case the precoding
matrix column which has larger effective transmission power
CA 02731210 2011-01-17
WO 2010/018969 9
PCT/KR2009/004468
is mapped to the layer which is solely mapped to a single
codeword. Thus in case of precoding matrix in the form of
/ 0 =0
0 1 0
0 0
Jr 0: 0.
, the first column is mapped to the layer which is
solely mapped to a single codeword, and the second and third
column is mapped to layers which is mapped to the other
codeword.
[0021] The codebook may include a different number of
precoding matrices for each rank.
[0022] Each of the uplink signals may be entered in units
of a codeword, and the mapping step of the uplink signals to
the predetermined number of layers may periodically change a
layer mapped to a specific codeword to another layer. One
example of this periodicity can be I SC-FDMA symbol.
[0023] In another aspect of the present invention, a user
equipment (UE) for transmitting uplink signals via multiple
antennas includes multiple antennas for transmitting and
receiving signals, a memory for storing a codebook having a
precoding matrix established in a manner that one layer
signal is transmitted to the multiple antennas, and a
processor connected to the multiple antennas and the memory
so as to process the uplink signal transmission. The
processor includes a layer mapper for mapping the uplink
CA 02731210 2011-01-17
WO 2010/018969 10
PCT/KR2009/004468
signals to a predetermined number of layers corresponding to
a specific rank, a Discrete Fourier Transform (DFT) module
for performing DFT spreading upon each of the predetermined
number of layer signals, a precoder for precoding each of the
DFT-spread layer signals received from the DFT module by
selecting a specific precoding matrix established in a manner
that one layer signal is transmitted to each of the multiple
antennas from among a codebook stored in the memory, and a
transmission module for performing a predetermined process
for constructing a Single Carrier-Frequency Division Multiple
Access (SC-FDMA) symbol upon the precoded signals, and
transmitting the processed signals to a base station (BS) via
the multiple antennas.
[0024] In this case, the memory may store the codebook.
The processor may perform the antenna shift and/or the layer
shift either in a different way from the precoding of a
,
precoder or through row permutation and/or column permutation
of a precoding matrix.
[0025] It is to be understood that both the foregoing
general description and the following detailed description of
the present invention are exemplary and explanatory and are
intended to provide further explanation of the invention as
claimed.
CA 02731210 2011-01-17
WO 2010/018969 11
PCT/KR2009/004468
BRIEF DESCRIPTION OF THE DRAWINGS
[0026] The accompanying drawings, which are included to
provide a further understanding of the invention and are
incorporated in and constitute a part of this application,
illustrate embodiment(s) of the invention and together with
the description serve to explain the principle of the
invention. In the drawings:
[0027] FIG. 1 is a conceptual diagram illustrating a
general MIMO communication system.
[0028] FIGS. 2 and 3 illustrate a general structure of a
transmitter based on a MIMO technology.
[0029] FIG. 4 is a conceptual diagram illustrating a
method for precoding information of each layer and
transmitting the precoded information via an antenna.
[0030] FIG. 5 is a conceptual diagram illustrating a
general SC-FDMA scheme.
[0031] FIG. 6 is a conceptual diagram illustrating a
method for mapping a codeword to several layers.
[0032] FIG. 7 is a conceptual diagram illustrating a
method for performing a DFT upon each layer after performing
codeword-to-layer mapping (i.e., codeword-layer mapping) so
as to prevent a CM value for each antenna from being
increased.
CA 02731210 2011-01-17
WO 2010/018969 12
PCT/KR2009/004468
[0033] FIG. 8 is a conceptual diagram illustrating a
method for performing permutation on the position of a row or
column of a precoding matrix.
[0034] FIG. 9 is a conceptual diagram illustrating a
chordal distance.
[0035] FIG. 10 is a block diagram illustrating a general
base station (BS) and a general user equipment (UE).
[0036] FIGS. 11 and 12 illustrate an SC-FDMA scheme for
transmitting an uplink signal in a 3GPP LTE system and an
OFDMA scheme for transmitting a downlink signal in the 3GPP
LTE system.
[0037] FIG. 13 is a block diagram illustrating a processor
for enabling a base station (BS) to transmit a downlink
signal using a MIMO scheme in a 3GPP LTE system.
[0038] FIG. 14 illustrates a processor of a UE according
to one embodiment of the present invention.
DETAILED DESCRIPTION OF THE INVENTION
[0039] Reference will now be made in detail to the
preferred embodiments of the present invention, examples of
which are illustrated in the accompanying drawings. Wherever
possible, the same reference numbers will be used throughout
the drawings to refer to the same or like parts.
[0040] The detailed description, which will be given below
with reference to the accompanying drawings, is intended to
CA 02731210 2011-01-17
WO 2010/018969 13
PCT/KR2009/004468
explain exemplary embodiments of the present invention,
rather than to show the only embodiments that can be
implemented according to the present invention.
The
following detailed description includes specific details in
order to provide a thorough understanding of the present
invention. However, it will be apparent to those skilled in
the art that the present invention may be practiced without
such specific details.
For example, the following
description will be given centering on specific terms, but
the present invention is not limited thereto and any other
terms may be used to represent the same meanings.
Also,
wherever possible, the same reference numbers will be used
throughout the drawings to refer to the same or like parts.
[0041] Peak power to Average Power Ratio (PAPR) is a
parameter indicating characteristics of a waveform. PAPR is
a specific value acquired when a peak amplitude of the
waveform is divided by a time-averaged Root Mean Square
(RMS) value of the waveform. PAPR is a dimensionless value.
In general, a PAPR of a single carrier signal is better than
that of a multi-carrier signal.
[0042] An LTE-Advanced scheme can implement MIMO
technology using Single Carrier - Frequency Division
Multiple Access (SC-FDMA) so as to maintain a superior CM
property. When using general precoding, a signal including
information corresponding to several layers is multiplexed
CA 02731210 2011-01-17
WO 2010/018969 14
PCT/KR2009/004468
and transmitted via a single antenna, so that the signal
transmitted via this antenna may be considered to be a kind
of multi-carrier signal. PAPR is associated with a dynamic
range that must be supported by a power amplifier of a
transmitter, and a CM value is another value capable of
being used as a substitute for the PAPR.
[0043] FIG. 2 shows a general structure of a transmitter
based on a MIMO technology.
[0044]
In FIG. 2, one or more codewords are mapped to a
plurality of layers. In
this case, mapping information is
mapped to each physical antenna by a precoding process, and
is then transmitted via each physical antenna.
[0045]
FIG. 3 is a detailed block diagram illustrating the
MIMO-based transmitter shown in FIG. 2.
[0046] The term 'codeword' indicates that Cyclic
Redundancy Check (CRC) bits are attached to data information
and are then encoded by a specific coding method. There are
a variety of coding methods, for example, a turbo code, a
tail biting convolution code, and the like. Each codeword is
mapped to one or more layers (i.e., one or more virtual
layers), and a total number of mapped layers is equal to a
rank value.
In other words, if a transmission rank is 3, a
total number of transmission layers is also set to 3.
Information mapped to each layer is precoded. In this case,
data information mapped to each layer is mapped to a physical
CA 02731210 2011-01-17
WO 2010/018969 15
PCT/KR2009/004468
layer through a precoding process (where, the term 'layer'
means a virtual layer as far as it especially designates a
physical layer). Information is transmitted to each antenna
via each physical layer.
Under the condition that no
specified explanation is shown in FIG. 3, the precoding is
carried out in a frequency domain, and an OFDM information
transmission scheme is used for information mapped to the
physical layer. The information mapped to the physical layer
is mapped to a specific frequency domain, and is then IFFT -
processed.
After that, a cyclic prefix (CP) is attached to
the IFFT result.
Thereafter, information is transmitted to
each antenna via a radio frequency (RF) chain.
[0047] The precoding process may be carried out by matrix
multiplication. In each of the matrices, the number of rows
is equal to the number of physical layers (i.e., the number
of antennas), and the number of columns is equal to a rank
value. The rank value is equal to the number of layers, so
that the number of columns is equal to the number of layers.
Referring to the following equation 2, information mapped to
a layer (i.e., a virtual layer) is xl and x2, each element pij
of a (4 x 2) matrix is a weight used for precoding. yi, y2,
y3, and y4 are information mapped to physical layers, and are
transmitted via respective antennas using individual OFDM
transmission schemes.
[0048] [Equation 2]
CA 02731210 2011-01-17
WO 2010/018969 16
PCT/KR2009/004468
_ _ - _
.Y1 Al P21 '
[
XI
Y2 ' P12 . P22
33 .. .P13 A3 :
[0049] In the following description, a virtual layer will
hereinafter be referred to as a layer so long as such use
will not lead to confusion.
An operation for mapping a
virtual layer signal to a physical layer will hereinafter be
considered to be an operation for directly mapping a layer to
an antenna.
[0050] The precoding method can be mainly classified into
two methods, i.e., a wideband precoding method and a subband
precoding method.
[0051] The wideband precoding method is as follows.
According to the wideband precoding method, when precoding is
carried out in a frequency domain, the same precoding matrix
is applied to all information transmitted to the frequency
domain.
[0052] FIG. 4 is a conceptual diagram illustrating a
method for precoding information of each layer and
transmitting the precoded information via an antenna.
[0053] Referring to FIG. 4, it can be recognized that
information corresponding to a plurality of layers is
precoded while being classified according to subcarriers of
each frequency domain, and the precoded information is
CA 02731210 2011-01-17
WO 2010/018969 17
PCT/KR2009/004468
transmitted via each antenna. All precoding matrices 'P' for
use in the wideband precoding method are equal to each other.
[0054] The subband precoding method is provided by the
extension of the wideband precoding method.
The subband
precoding method applies a variety of precoding matrices to
each subcarrier without applying the same precoding matrix to
all subcarriers.
In other words, according to the subband
precoding method, a precoding matrix 'P' is used in a
specific subcarrier, and another precoding matrix 'M' is used
in the remaining subcarriers other than the specific
subcarrier.
Herein, element values of the precoding matrix
'P' are different from those of the other precoding matrix
[0055] Uplink signal transmission is relatively sensitive
to PAPR or CM properties as compared to downlink signal
transmission.
The increase of filter costs caused by the
increase of PAPR or CM properties may generate more serious
problems in a user equipment (UE). Thus, the SC-FDMA scheme
is used for uplink signal transmission.
[0056] FIG. 5 is a conceptual diagram illustrating a
general SC-FDMA scheme. =
[0057] As shown in FIG. 5, the OFDM scheme and the SC-FDMA
scheme are considered to be identical with each other,
because they convert a serial signal into parallel signals,
map the parallel signals to subcarriers, perform an IDFT or
CA 02731210 2011-01-17
WO 2010/018969 18
PCT/KR2009/004468
IFFT process on the mapped signals, convert the IDFT- or
IFFT- processed signals into a serial signal, attach a cyclic
prefix (CP) to the resultant serial signal, and transmit the
CP resultant signal via a radio frequency (RF) module.
However, in contrast to the OFDM scheme, the SC-FDMA scheme
converts parallel signals into a serial signal, and performs
DFT spreading upon the serial signal, so that it reduces the
influence of a next IDFT or IFFT process and maintains a
single signal characteristic of more than a predetermined
level.
[0058] In the meantime, the reason why the CM value is
degraded when a MIMO scheme is applied to uplink signal
transmission is as follows. If a plurality of single-carrier
signals each having good CM properties is simultaneously
overlapped with each other, the overlapped signals may have
poor CM properties.
Therefore, if the SC-FDMA system
multiplexes output information of several layers using a
minimum number of single-carrier signals or one single-
carrier signal on a single physical antenna, a transmission
signal having a good CM can be generated.
[0059] A codeword-layer mapping process may be performed
before information to be transmitted is precoded. Since the
SC-FDMA scheme is generally used for one transmission mode
(1Tx), the number of layers is 1.
However, if the SC-FDMA
scheme supports a MIMO scheme, the number of layers is plural,
CA 02731210 2011-01-17
WO 2010/018969 19
PCT/KR2009/004468
and a codeword composed of a single transport block may be
mapped to a plurality of layers.
[0060] FIG. 6 is a conceptual diagram illustrating a
method for mapping a codeword to several layers.
[0061] Referring to FIG. 6, if the codeword-layer mapping
is carried out after a DFT process for the SC-FDMA scheme is
performed, a CM value may be increased. That is, because an
output signal of a DFT block undergoes other processes before
entering an IFFT module, i.e., because the output signal of
the DFT block is divided into two layers, a CM value may be
increased.
[0062] FIG. 7 is a conceptual diagram illustrating a
method for performing DFT upon each layer after performing
codeword-to-layer mapping (i.e., a codeword-layer mapping) so
as to prevent a CM value for each antenna from being
increased.
[0063] Therefore, if the number of DFT blocks is changed
while being classified according to layer numbers based on a
rank value, a low CM value can be maintained. That is, the
output signal of the DFT block is directly input to the IFFT
block without passing through other processes, so that a low
CM value can be maintained.
In the case of actual
implementation, a plurality of layers may share a single DFT
block.
CA 02731210 2011-01-17
WO 2010/018969 20
PCT/KR2009/004468
[0064] If a plurality of layer signals is transmitted via
a single antenna by applying the MIMO scheme to uplink signal
transmission, a PAPR or a CM property may be deteriorated.
In order to overcome the above-mentioned problem, the
following embodiments of the present invention will describe
a method for designing a codebook based on a precoding matrix
by which only one layer signal is transmitted via a single
antenna.
[0065] For convenience of description and better
understanding of the present invention, in a transmission
system, it is assumed that a set of signals transferred to a
precoding block is set to 'x', and a set of precoded signals
is set to 'y'. In this case, if the precoding matrix is 'P',
the following equation 3 is acquired.
[0066] [Equation 3]
Y = P = x
[0067] In Equation 3, a dimension of 'P' is NT X NL, a
dimension of 'x' is NL x 1, a dimension of 'y' is NT X 1.
In
this case, NT is the number of antennas, and NL is the number
of layers.
[0068] In the following description, a principle of
designing a codebook that is capable of being applied to
uplink signal transmission using a MIMO scheme by a UE will
be firstly described in chapter (I), and a detailed format of
the codebook will then be described in chapter (II).
CA 02731210 2011-01-17
WO 2010/018969 21
PCT/KR2009/004468
[0069] I. Principle of Codebook Design
[0070] <2Tx codebook>
[0071] A variety of embodiments according to a structure
of a precoding matrix contained in a codebook used in a 2Tx
mode will hereinafter be described.
[0072] The method according to embodiments of the present
invention includes: generating a plurality of streams by
mapping a codeword to a plurality of layers; and precoding
the generated streams, mapping the precoded streams to a
plurality of antennas, and transmitting the mapped result via
the antennas. In this case, the codebook may be configured
as follows. A precoding matrix used in Rank 1 and the other
precoding matrix used in Rank 2 will be described in
different ways.
[0073] 2Tx - Rank 1 Precoding Matrix
[0074] In case of 2Tx - Rank 1, Equation 3 can be
rewritten as the following equation 4 according to
embodiments of the present invention.
[0075] [Equation 4]
y. .p.x. [x i]=[1]
Y2 b bx,
[0076] In general, if it is assumed that a wideband
precoding scheme is used, a specific constant value is
CA 02731210 2011-01-17
WO 2010/018969 22
PCT/KR2009/004468
multiplied by a signal of each layer according to a Rank 1
precoding scheme, a PAPR and CM value of a signal transmitted
via each antenna in the 2Tx mode are equal to those in a 1Tx
mode. Thus, when using wideband precoding, the PAPR and the
CM are not affected by values of a 2Tx - Rank 1 precoding
matrix.
[0077] Precoding is a method for changing a channel so as
to acquire a constructive effect between signals transferred
via each channel.
Thus, transmission performance of each
signal is improved.
Accordingly, 'a' indicating a first
element of the precoding matrix P illustrated in Equation 4
is set to '1', and a second element 'b' of the precoding
matrix P may be set to an arbitrary value.
Signals
transferred via respective antennas have the same power, so
that all power amplifiers contained in each antenna can be
maximally used. For this purpose, the above-mentioned second
element 'b' may be a complex number having an absolute value
of 1.
In other words, P shown in Equation 4 may be
1
13= [0.61
represented by
[0078] There is a limitation in the number of precoding
matrices contained in a codebook used for the precoding,
because both a transmission end and a reception end must have
a codebook and information about a predetermined precoding
matrix is communicated between the transmission end and the
CA 02731210 2011-01-17
WO 2010/018969 23 PCTXR2009/004468
reception end.
As a result, the transmission end and the
reception end must use a limited number of precoding matrices.
For this operation, a complex number that has an absolute
value of 1 and a phase corresponding to any one of +0 , +45 ,
+90 , +135 , +180 , -135 , -90 , and -45 may be used as each
element of the precoding matrix.
That is, in the above -
H
.................................. [
1?::.!. D
mentioned expression .
, (9 may be represented by
{
0e 0, Ir Ir 375r, 57r 67r 77r
4 ' 2 ' 4 4 ' 4 ' 4 = In other words,
P may be
represented
by
. 1 ' L . - 1 .
pc 1+ j ' 1..- j -1-j ;:-. . : -1+ j`
, , , ______ , A. 1 __ A A : .. . : A
1 i4. ' Nti. . µ1 1:. '111. h ' -1 . : : :kJ. . :17 :1.:
:la 1: .
{
[ I ' 1; H . [ . 1 :; 1 : [ I . 1 : L ; 1 . : 1 ;
1 0 '
[0079] 2Tx - Rank 2 Precoding Matrix
[0080]
In case of a 2Tx - Rank 2, the equation 3 may be
rewritten as the following equation 5.
[0081] [Equation 5]
=
[Y1 = -rh x = P11 P12 , = X1 P11X1 4. P12X2
y I-- = =
Y2 P21 P22 x2 P21X1 + P23-x2
[0082]
In Equation 5, the signal yk transferred via each
antenna is composed of a combination of several input signals
xi, so that a CM value may be increased.
CA 02731210 2011-01-17
WO 2010/018969 24
PCT/KR2009/004468
[0083] In this case, if each of p12 and pn is set to zero
'0' or if each of pn and p22 is set to zero '0', only one
signal can be transmitted via each antenna. Thus, if it is
assumed that a CM value of a signal xi is considered to be
good, a CM value of the precoded signal also becomes good.
In association with FIG. 7, in the case where a codeword is
mapped to each layer, DFT spreading is applied to the
resultant signal mapped to each layer, and a precoding
process for allowing each antenna to transmit only one layer
signal is carried out, the same effect as in an IDFT or IFFT
process that is performed as soon as a DFT process was
performed can be acquired, and a PAPR or CM property can be
maintained at a good status. A detailed description of this
will hereinafter be explained in the following description.
[0084] In this case, if each of p12 and p21 is zero '0', a
signal corresponding to each layer is transmitted via each
antenna after being multiplied by a constant complex value.
As a result, although the above-mentioned constant complex
value is set to 1, performance is not affected by this
constant complex value of 1.
[0085] Therefore, Equation 5 can be represented by the
following equation 6.
[0086] [Equation 6]
[Y I [P I Lx. I 11 .................. [x
-x I Pin 1 0
y2 0: /222 ,,x2 0 1 x2 X2 0 1
CA 02731210 2011-01-17
WO 2010/018969 25
PCT/KR2009/004468
[0087] <4Tx codebook>
[0088] A variety of embodiments according to a structure
of a precoding matrix contained in a codebook used in a 4Tx
mode will hereinafter be described.
[0089] The method according to embodiments of the present
invention includes: generating a plurality of streams by
mapping a codeword to a plurality of layers; and precoding
the generated streams, mapping the precoded streams to a
plurality of antennas, and transmitting the mapped result via
the antennas. In this case, the codebook may be configured
as follows. Precoding matrices respectively used in Rank 1,
Rank 2, Rank 3, and Rank 4 will be described in different
ways.
[0090] 4Tx - Rank 1 Precoding Matrix
[0091] In case of 4Tx - Rank 1, Equation 3 can be
rewritten as the folloWing equation 7.
[0092] [Equation 7]
Yi a axi
y, b
y= =P = x= = rx, =
Y3 C=CX1
_Y4_
[0093] In case of using the wideband precoding scheme in
the same manner as in the 2Tx - Rank 1 codebook, a CM of a
signal transmitted via each antenna by a 4Tx - Rank 1
CA 02731210 2011-01-17
WO 2010/018969 26 PCT/KR2009/004468
precoding process is equal to a CM of a signal used in the
1Tx mode. Thus, all kinds of precoding matrices can be
freely applied to such a CM without any problems.
[0094] 4Tx - Rank 2 Precoding Matrix
[0095] In case of 4Tx - Rank 2, Equation 3 can be
rewritten as the following equation 8.
[0096] [Equation 8]
Yd... : ..: : :: _P I1 : P12 ..: i . . 12
[
: : Yi:: :. ,n: L:Pii'= ::P22: : "3.C;L:: : :
1:7jili''IP
y: ----.-.-:: ...... F.-,--:1--: :H ::.; :..F : : : .
:Y4:: " : .. :. ' :.: ::P,ii . 1:Pij ! :: qqi.;
i.Pii:NI:*:P3'X2 . .:
. .
_Y4-; _PCI . P42 .._P41.X,1lf P42:X3 L
[0097] In a 4Tx - Rank 2 codebook, in a similar way as in
the 2Tx - Rank 2 codebook, a specific element of a precoding
matrix is set to zero '0', so that the overlapping of signals
transmitted via respective antennas is minimized and thus a
CM can be maintained at a low value.
[0098] In Equation 8, if it is assumed that pki or pk2 in a
signal (pkixi + pk2x2) transmitted via each antenna is set to
zero '0', the signal transmitted via each antenna becomes
equal to a signal transmitted from a single layer, and
therefore a CM of the signal transmitted via each antenna can
be maintained at a low value.
CA 02731210 2011-01-17
WO 2010/018969 27
PCT/KR2009/004468
[0099] In one embodiment of the present invention, 'P'
1711 ID
1321 0
P=
0 pn
0 pu
included in Equation 8 may be represented by - -
.
Equation 8 may be rewritten as the following equation 9.
[00100] [Equation 9]
y1 Fyn 0 P11x1
Y2 Par 0 Xi P21x1
Y3 a P32 Lx2J P32;
5- a P42 _ p43 x2
-- - : : -
[00101] Referring to Equation 9, only one layer is mapped
to a signal transmitted via each antenna. From the viewpoint
of a single layer, it is considered that the 2Tx - Rank 1
precoding is applied to information transmitted via this
single layer. Thus, the 4Tx - Rank 2 precoding matrix can be
configured using a 2Tx - Rank 2 precoding matrix. In other
words, the 4Tx - Rank 2 precoding matrix may be a super
matrix of the 2Tx - Rank 1 precoding matrix.
[00102] For example, 'P' according to one embodiment of the
present invention can be represented by Equation 10.
[00103] [Equation 10]
- 'Q..7, 1
[0
X 0 X 0
0 [Y 0 1 0 1 1+ j 1-j
1 -1-j -1+j
0
X Y E 1, ¨, , j, , , , , , j, ,
Y , V2 42 V2 N/2
- ,
CA 02731210 2011-01-17
WO 2010/018969 28
PCT/KR2009/004468
[00104] The above-mentioned 2Tx-Rankl precoding matrix is
used for a method for transmitting information by applying
two antennas to a single layer signal.
However, if it is
assumed that there are four physical antennas, communication
performance may be changed according to which combination
composed of two antennas is used for data transmission.
In
this case, the selected combination of antennas may be
changed according to a value of the precoding matrix P.
[00105] For example, according to one embodiment of the
present invention, the precoding matrix P may be configured
in various formats.
Respective formats may indicate
different antenna combinations.
[00106] [Equation 11]
{ 1 0 1 0 1 0
AP 0 0 1 0 1
P E
0 1 X 0 0 Y
0 Y. 0 Y. X 0
[00107] In
Equation 11, if an appropriate value is selected
as a precoding matrix P, performance improvement due to
precoding can be enhanced.
If the precoding matrix is
configured as described above, a signal corresponding to each
layer uses two antennas among a total of four antennas,
channel estimation performances among respective layers
become similar to each other, and a CM value for each antenna
can be minimized.
CA 02731210 2011-01-17
WO 2010/018969 29
PCT/KR2009/004468
[00108] Generally, although a constant value is multiplied
by a specific column vector of an arbitrary precoding matrix,
characteristics of the precoding matrix are not changed.
Therefore, although a constant value is multiplied by a
specific column of the above-mentioned precoding matrix,
characteristics of the precoding matrix are not changed. As
a result, the above-mentioned operation for multiplying a
constant value by a specific column vector of the precoding
matrix does not depart from the scope of the present
invention.
[00109] In addition, if a predetermined scaling factor is
multiplied by the precoding matrix shown in Equation 11, the
multiplied result may be represented by the following
equation 12.
[00110] [Equation 12]
1 0 1 0 1 0
X 0 0 1 0 11+j ;1-j -1-j -11
PE 1 {k- ,k- ,k- X,YE{1, r- 5 j )
-N5 5 j5
0 X0 OY V2 V2 Vi
0 Y0 Y X 0
[00111] 4Tx - Rank 3 Precoding matrix (1)
[00112] In case of 4Tx - Rank 3, Equation 3 can be
rewritten as the following equation 13.
[00113] [Equation 13]
CA 02731210 2011-01-17
WO 2010/018969 30
PCTXR2009/004468
= Pi PL2 ,P13
= .7 '7 P11; P12X2 +.P13X3
" 1
P21 :F22 P23 P21X11
P22X2 P23)c3:
r =, 2 : "
= _).?:=== Pi : iP33,
-FP32 X3 t=T33.4.3:
. . .
176: Pit ;';:!'7
= = .. . . .
= = =
[00114] In a 4Tx - Rank 3 precoding matrix in a similar way
as in the 4Tx - Rank 2 precoding matrix, a specific element
of a precoding matrix is set to zero '0', so that the
overlapping of signals transmitted via respective antennas is
minimized and thus a CM can be maintained at a low value.
[00115] In Equation 13, if it is assumed that pu, pk2, or
Pu in a signal (puxi + pk2x2 + pux3) transmitted via each
antenna is set to zero '0', a CM of the signal transmitted
via each antenna can be maintained at a low value.
[00116] In one embodiment of the present invention, '2'
included in Equation 12 may be represented by
¨
õ
: .
o . :
:. p 2,
0 O. p33
_P41 Pu Pu_ Equation 13 may be rewritten as the
following equation 14.
[00117] [Equation 14]
.õ
. =
.=:"
= = -"Xµ = 0 p2= X, =
y3 .0 0 ?33 2
_y4 p41. P42 P43 _ P42?Xa -17 p43 x3.
[00118] In Rank 3, the number of layers to be transmitted
is 3, and the number of physical antennas is 4. In this case,
CA 02731210 2011-01-17
WO 2010/018969 31
PCT/KR2009/004468
each of the three antennas may be independently mapped to a
single layer. Herein, only a signal of a single layer may be
mapped to the remaining one antenna, or signals of at least
two layers may be mapped to the remaining one antenna.
If
only a signal of a specific single layer is mapped to the
remaining one antenna, a CM of the signal transmitted via
this antenna may have good characteristics, but communication
performance of the specific single layer may be different
from that of another layer. For example, in the case where
information of a first layer (Layer 1) is mapped to a first
antenna (Antenna 1) and a fourth antenna (Antenna 4),
information of a second layer (Layer 2) is mapped to a second
antenna (Antenna 2), and information of a third layer (Layer
3) is mapped to a third antenna (Antenna 3), communication
performance of the Layer 1 information may be different from
that of either the Layer 2 or the Layer 3.
[00119] In one embodiment of the present invention, in
order to minimize a CM value for each antenna in the
precoding process, the precoding matrix P may have any one of
the values of P1,P2, and P3 shown in the following equation
15.
[00120] [Equation 151
':.(1 A37. b- -1 0-
0 10 010 04 0
P- P-
1- P3=
0 0 1 .001 0 0 1
.2ir 00 0.1'0 00Z
CA 02731210 2011-01-17
WO 2010/018969 32
PCT/KR2009/004468
{
v v 7 i 1+; , 1-; -1-j 4 ¨ 1+ j
la ¨r-- v Jo r¨ I -1 ____________________ ,- . -
. ...................... 42 42 . .42 42
where,
[00121] In case of using the above-mentioned precoding
matrices P1, P2, and P3, numbers of antennas used for
individual layers are different from each other. However, if
it is assumed that the precoding matrices Pl, P2 and P3 are
evenly used to transmit certain information, instead of using
any one of the precoding matrices P
- 1, P2 and P3, numbers of
antennas used for individual layers may be normalized.
Although the precoding matrices pl, P2 and P3 can be
alternately used in a frequency domain, a single carrier
property of a signal composed of a single carrier is damaged,
so that a CM value is unavoidably increased. Therefore, if
the precoding matrices Pl, P2 and P3 are alternately applied
to each SC-FDMA symbol, no additional increase in CM is
achieved. In
case of transmitting data, information may be
decoded in units of one subframe.
Thus, if the precoding
matrices Pl, P2 and P3 are alternately applied to each SC-FDMA
symbol, each layer information of the whole information
transmitted via a single subframe can be transmitted via the
same number of antennas on average.
[00122]
In another embodiment of the present invention, the
position of an antenna used by each layer is changed so that
performance can be improved.
The changing of the antenna
position may be carried out with time.
In particular, the
CA 02731210 2011-01-17
WO 2010/018969 33
PCT/KR2009/004468
antenna position can be changed at each SC-FDMA symbol. A
detailed method for changing the antenna position will
hereinafter be described in detail.
[00123] For example, the position of a value other than '0'
in the precoding matrix is changed to another position in the
range of a row vector, so that the position of an antenna via
which each layer signal is transmitted can be changed to
another position. As another example, the above-mentioned
method may be implemented by a row/column permutation because
position permutation is carried out between rows or columns
of a given precoding matrix.
[00124] FIG. 8 is a conceptual diagram illustrating a
method for performing permutation on the position of a row or
column of a precoding matrix.
[00125] In more detail, FIG. 8(a) is a conceptual diagram
illustrating a method for performing permutation on the
position of a row, and FIG. 8(b) is a conceptual diagram
illustrating a method for performing permutation on the
position of a column.
[00126] In the precoding matrix shown in Equation 15, a
precoding matrix Pi can be row-permuted and/or column-permuted,
so that a precoding matrix P2 or P3 can be generated.
Therefore, in the structure such as the precoding matrix Pl,
P2 Or P31 a new unique precoding matrix can be acquired only
by row permutation.
CA 02731210 2011-01-17
WO 2010/018969 34
PCT/KR2009/004468
[00127] The order of rows changed by row permutation
available in the 4Tx mode can be represented by the following
expression.
{1, 2, 3, 4}, {1, 2, 4, 3}, {1, 3, 2, 4}, {1, 3, 4, 2},
{1, 4, 2, 31, {1, 4, 3, 2}, {2, 1, 3, 4}, {2, 1, 4, 31,
{2, 3, 1, 4}, (2, 3, 4, 1), {2, 4, 1, 3}, {2, 4, 3, 1},
{3, 2, 1, 4}, {3, 2, 4, 1}, {3, 1, 2, 41, {3, 1, 4, 2},
{3, 4, 2, 11, {3, 4, 1, 2}, {4, 2, 3, 1}, {4, 2, 1, 3},
{4, 3, 2, 1}, {4, 3, 1, 2}, {4, 1, 2, 31, {4, 1, 3, 2}
[00128] In the above-mentioned expression, {w, x, y, z}
means that row vectors 1, 2, 3 and 4 of the precoding matrix
are rearranged in the order of parenthesized numbers on the
condition that a given precoding matrix Pk exists.
[00129] By row permutation, signals corresponding to a
specific layer are mapped to different antennas. By column
permutation, the same effect as in the switching of
information of different layers can be acquired. If there is
no need to distinguish performance of each layer, and a
system for requesting similar performance from each layer
need not utilize the column permutation.
Thus, the effect
corresponding to antenna selection can be acquired using only
the row permutation.
[00130]
In the meantime, in the case where a given scaling
factor is multiplied by each precoding matrix shown in
CA 02731210 2011-01-17
WO 2010/018969 35
PCT/KR2009/004468
Equation 15, the result can be represented by the following
equation 16.
[00131] [Equation 16]
1 0 0 1 0 0 1 0 0
0 1 0 0 1 0 0 1 0
Pi=k= P =k-
0 0 1 2 0 0 1 0 0 1
X 0 0 0 Y 0 0 0 Z
=
________________ X, Y,Z E Ti - 1+j 1- j -1, j
[00132] 4Tx - Rank 3 Precoding Matrix (2)
[00133] In case of 4Tx - Rank 3, if each antenna transmits
information corresponding to only one layer, a CM value of a
signal transmitted via each antenna can be maintained at a
low value, however information of only one layer is
transmitted via only one antenna so that communication
performance can be deteriorated. Therefore, in case of 4Tx -
Rank 3, there is a need for a codebook to be designed in a
manner that a maximum of two layers are multiplexed and
transmitted via a single antenna, so that the increment of CM
can be minimized and at the same time communication
performance can be increased.
[00134] In accordance with one embodiment of the present
invention, when information corresponding to two layers is
transmitted via a single antenna, the precoding matrix P
CA 02731210 2011-01-17
WO 2010/018969 36 PCTXR2009/004468
shown in Equation 13 can be represented by P4 of Equation 17
or P5 of Equation 18.
[00135] [Equation 17]
1 0 1
p4=
0
=
4-- o 1: . n = , - = = =.1+ i
...1- j= ,.. .1-. j., ..-.. . -1+j
... 0 r 9 , x- 0 1
E 1, _______________________________________ : j - i ' __ -1 __
r- .. ,-- . . ,- WO = f--.
. .L. z 42 ' 42
" '
[00136] [Equation 18]
1 0 0
x1 0
Ps=
0 Y 1 j1 1+j .1-j -1-j . -1+j
_0 0 2._ If Nrj li .µi
,
[00137] In Equation 17, in order to satisfy Rank 3, 'X'
must be different from 'Z' in the precoding matrix P4.
[00138] A method for using the precoding matrix P4 or P5
has a disadvantage in that only a signal of a single layer is
transmitted via the other antenna whereas a signal of two
layers is multiplexed and transmitted via a specific antenna.
[00139] In one embodiment of the present invention, in
order to obviate the above-mentioned problem, the precoding
matrix P may have any one of values P8, P7 and P8 shown in the
following equation 19.
[00140] [Equation 19]
. ......... .- 1. o z-. , 1. 0. 4.7: ... .r. 0
27
:p.,_ x. 1. 0 , ,F, X .1 ..:171:i:!.. pi õ. X I
.:aj ..,
'''T o Y' 1. .. 1 . o k . 11 ' '81 - .o. -.1..fi' .:1.! i
[
. .A 0 C . 0 . B CA A Et Ø
CA 02731210 2011-01-17
WO 2010/018969 37
PCT/KR2009/004468
1+i 1
X,Y,ZEA,B,CE j, -j, -1 -
-
where,
[00141] In association with the precoding matrix P4, P5, P6,
P7 or Pg, row permutation and/or column permutation can be
carried out on the 4Tx - Rank 3 precoding matrix. Because
the row permutation and the column permutation are carried
out, an antenna selection function and layer permutation
function for enabling a signal of a specific layer to be
transmitted via an arbitrary antenna can be implemented by
the precoding.
[00142] In one embodiment of the present invention,
individual column vectors of the precoding matrix may be
configured to have orthogonality therebetween.
[00143] If individual column vectors of the precoding
matrix are configured to have orthogonality therebetween, the
precoding matrix is able to satisfy properties of a one side
unitary matrix. That is, the precoding matrix P can have a
characteristic denoted by the following equation 20.
[00144] [Equation 201
PHP = cr.I #13P11
[00145] In one embodiment of the present invention, the
precoding matrix of Rank 3 can be configured in the form of
the following equation 21.
The precoding matrix P for
satisfying the following equation 21 is able to satisfy the
relationship illustrated in Equation 20.
CA 02731210 2011-01-17
WO 2010/018969 38
PCT/KR2009/004468
[00146] [Equation 21]
,-..-1 , II.: : 1 ..7
X : = 0 :1==:.2f.
1?...*:.= = = :..:::: ,- ::.:i: .. .. . .: :
. .. :õ:.:... : .. ,:. . ., ..,. , ... :. ,.:
.
'0'..:1: ... )::'-7, -.)..141:2. ....1-r_.f..''.1.=-111 .':.11+ j
{
-:0 . T.:: - 0.r - = =!:,. .. '.1..va -' ..42:,
:':::,..a.:,'.. .'. 42 =
.. --.. .: ... . : .: ., .: .::: ' -:
.: .. =.:. ,:. . ..- .: :- . . :
[00147] In Equation 21, since the relationship denoted by
. ,:: -;;.. = ..:: ' , :.=: ,1;:.: ::1"). :
'..t
1, : 41-7'.; '.Ø.1
[ Y :Ii.:==:' :,:i.:...: ;1.7', ,......i.. iHt ;;IZ,,lc;(:,1:):
pH !! =. ,.: i:), ci:. :11.:,õ= .. ''-':.' :.. ' :
' =: = :T = = ' : 2 ..,='..li ,..:[.*:.I.I.'
- ' .:=:: : :: . :::.: . ''..' = -:' .: ':
=AIT,'.: ,1:1',',':::Q ' i= ..F..= == = ... . : . ii :..
.:!: .
;a: ;.X.:.:' 0 = .A:. : . .. . H. = . K. :.. ' Z:H.a ..:2L:H =
.: :. = .:' . ..:. . . : = : . . = , H- :
: 0:,= '.11.';: : ..':=:::. = ...::: . :'.=:,,,:= ..:=.:
is satisfied, it
can be recognized that the matrix P satisfies Equation 20.
[00148] 4Tx - Rank 4 Precoding Matrix (1)
[00149] In case of 4Tx - Rank 4, Equation 3 can be
rewritten as the following equation 22.
[00150] [Equation 22]
:".: . = . ::: . . = =.: " : , ,. : ..=
=.= .. .: . = " . :=.:-.:.=,,:", :, .,= .,:-
:: = " = . . =:== .= : ,. =,
!:::==== ' .X::. ':. :, , :.:: .:. :r,12,:ill
=:iF4i:: :Rii. :::.p,i,s.:,,...;c:i:!! :
,ip.aAi.:t!'p.1.2;.c.it::,,iiix.,:tipi-44':::
. .= Y?...... !. ,,...: : : . Pii- P2.2 p23 Pic.:
:1.,i:. =.::Pi1X1f 717::P22 'Xi'''F.P.23';C3 47 'P24" X 4 =
7:==-= . - 1:=:17...xf= i õ ' , = " ..
õ..'= ::!:g :=.:' :.:: ==. = , " ". . " "
==. = Y3:= P31. = :P.32 ...,P33 3:?:;'3k. =?. 3c3. , :
P311,47.,P32XV+=P.,333 4- =P'34X=4
.. .:_Y=:.4 _ . P. 41.. : =P42 P41: P44.,= .-.7,c4
: LP411.771-= P42,.,x2,E -4- P43.X3 7.!--P.4.4x.4 =
_ . = . = -- - . õ ,. . = . _.
[00151] In case of 4Tx - Rank 4, signals from four layers
are multiplexed and transmitted via respective antennas.
[00152] In one embodiment of the present invention, if a
precoding matrix is configured in the form of an identity
matrix, one antenna is able to transmit only a signal
corresponding to a single layer. In this case, Equation 22
can be rewritten as the following equation 23.
[00153] [Equation 23] .
CA 02731210 2011-01-17
WO 2010/018969 39 PCT/KR2009/004468
-1 _ 0 0 x
y2 0 1. 0 0 x2 x2
y= = .r.x=.- =
y3 0 0 1 0 x3 x3
_y4_ -0- 0 0 1_ _x4_
[00154] 4Tx - Rank 4 Precoding matrix (2)
[00155] In a 4Tx -
Rank 4 codebook, if the number of Rank-4
precoding matrices is increased, communication performance
can also be increased. As the number of precoding matrices
contained in a codebook increases, a precoding matrix closer
to an actual channel can be selected. Thus, the greater the
number of precoding matrices, the higher the performance.
However, the selection of a precoding matrix in a codebook
becomes complicated, so that it is preferable that an
appropriate number of precoding matrices should be included
in such a codebook.
However, in case of 4Tx - Rank 4, in
order to transmit only a signal corresponding to a single
layer via each antenna, the precoding matrix should be an
identity matrix, so that a signal corresponding to two or
more layers should sometimes be transmitted via a single
antenna in case of using several Rank 4 precoding matrices.
Therefore, in order to minimize a CM value and increase the
number of Rank 4 precoding matrices in a codebook, a specific
element of the precoding matrix may be set to zero '0'. In
Equation 22, if it is assumed that two values of Pklr Pk2r Pk3
and pk4 in the signal (Pkixi Pk2X2 Pk3X3
Pk4X4) transmitted
CA 02731210 2011-01-17
WO 2010/018969 40
PCT/KR2009/004468
via each antenna are respectively set to zero '0', a CM of
the signal transmitted via each antenna can be maintained at
a low value.
[00156] In one embodiment of the present invention, the
precoding matrix can be set to Pg of the following equation 24,
Pio of the following equation 25, or P11 of the following
equation 26.
[00157] [Equation 24]
.1 A: ..- 0 õCI
p _ ...p. 1: ,,,0 0
'9.7. [ 0::i0 .1.. ie
=. . 11.?.. ;(1 .:P. 1:. , lio ABCD ,
. 1+j
. . -1-j
A,B, C, D e 1, ,_ , ), ,_ -1, .,_ - __
[00158] [Equation 25]
. -1 0. 1: 0-
.. . .. .
.. '.A .a:. Ø .13..
?" 7 0101 .
- õ:p . 4... oz: D
_ . . . ._
-1+3 .
A,B, C; D e 1, '4-,-1 j '-7 -1, -',.--/ -
{ 1
where, A o C, B 0 D , =Nn ' ' V2 ' 42 ' ''- 42 :
[00159] [Equation 26]
:1' :(1: V: - Q !
. . 'A -ag ' ,7.4. '0. .
p11= = = . .. . : ' :
0 1: . 8: - ;I . . ' ., 1.44 ..1-
f . -1 -1-) .:,-1-4- j
. : . A, B e i, ,_ ,.._ '1;.: __ ,__. , --
j,
' . _0 B 0 - B . .42 -,12 ' 42 42
- f
CA 02731210 2011-01-17
WO 2010/018969 41
PCT/KR2009/004468
[00160] The precoding matrix P9, Pn or Pil is an example of
a precoding matrix for transmitting a signal corresponding to
a maximum of two layers via each antenna. As described above,
the row/column permutation is performed on the precoding
matrix P9, Pn or Pll, so that signals of different layers can
be transmitted via different antennas.
[00161] The precoding matrix Pll is a unitary matrix, so
that the advantages of the unitary precoding matrix can be
utilized.
[00162] 4Tx - Rank 4 Precoding Matrix (3)
[00163] In case of 4Tx - Rank 4, only one element among
elements of each row of a precoding matrix can be set to zero
'0'.
In case of using the above method, a signal
corresponding to three layers can be multiplexed and
transmitted via a single antenna, so that communication
performance can be improved. However, in the case of using
the above-mentioned method, a CM value further increases, but
the increased CM value may be lower than another CM value
acquired when all elements of the precoding matrix are each
set to any of other values except for zero '0'.
Thus, the
above-mentioned method can be effectively utilized under a
good SNR status in which a transmitter need not transmit data
or information at a maximum transmission power.
CA 02731210 2011-01-17
WO 2010/018969 42
PCT/KR2009/004468
[00164] In one embodiment of the present invention, the
precoding matrix P can be represented by Pn of the following
equation 27, Pn of the following equation 28, P14 of the
following equation 29, or Pn of the following equation 30.
[00165] [Equation 27]
1 mu in13 13 -
n 0 1 . /7223 ff224
1-12 = .
:iitt : . 02
i : 31 0 1 . 34
P24i- 14i
1: I? a ' 1
. ., .. = , _
,
. .1+,f .. 1 --,/ -1-:j .. .7-1
{ ................................ _+) . k 12
in,,e 1 ,__. ),--._-_,-1, , ,-7. j; r._ , 2,, ..--! ,
;3,4
--: ' 42 - -42
[00166] [Equation 28]
1 0 1: . 1 -
n . ?In. 0. :4223" =M24 ,
.1 -
= ............ n I" 4 #1.33:: 1./.?34
. . ..
. . _,0 .042 in 43:: .M.4:4:t :,
= . . . =
{
Z.:;k=, L2;3;11-,
- 42 N12
-
[00167] [Equation 29]
1 0 1 1
"31 :a ina3 In24
"33" j 1--.7': -1- :7- -14-
i
{
. = : : 1:4r1'3 1 ' . - .. 311. ;;IA - E .:1; f._ f
n,._ ' -1; r_ : .-:i r_ : t k =12 3 4
u 1 M43 =1444 jig" 42., ... ' ' .-42 ' " ' '
- :-.. .. 7
[00168] [Equation 30]
_
1 1 1 0
0 c -c c
P15 =
a 0 -a -a.. , ,:. itj . i=L'I ::.,= i-.j _ . -1+ j.
4koic: E 1, , J.. , . li, ,:: .1, r-
b - b 0 b: .. ' 1?-
,
CA 02731210 2011-01-17
WO 2010/018969 43
PCT/KR2009/004468
[00169] The precoding matrix P15 illustrated in Equation 30
is a unitary matrix, so that the advantages of the unitary
precoding matrix can be utilized.
[00170] A matrix acquired when a constant is multiplied by
a specific column of the precoding matrix or another matrix
acquired when row/column permutation is performed on the
above-mentioned precoding matrix may be used as a part of a
codebook.
[00171] Elements of the above-mentioned precoding matrices
are selected from a complex number that has an absolute value
of 1 and a phase corresponding to any one of +0 , +450, +90 ,
+135 , +180 , -135 , -90 , and -45 .
That is, elements of
the precoding matrix are selected
from
{
;;' = .i ...ir ir 3217:. . 50' 6* 19r
el ..... .8.Ã :0 7r ¨
,. : ,¨ p :- ; -ri : ; : - ; --'. 1
. : .4 2. .4. . 4 :4i :4:.: :. .
For example, the above-
mentioned selection has been disclosed only for illustrative
purposes, and the elements of the precoding matrix may be
selected from a set of complex numbers that have an absolute
value of 1 and different phases.
For example, each element
of the precoding matrix may be selected from
e', 0 E 0,¨ 7I" 71- 3.7r . 57r 6,1- 77r
a ¨ --, 9r, --, ¨ ¨
4' 2' 4 4' 4' 4
(where a is an arbitrary
constant).
[00172] Power Balancing
CA 02731210 2011-01-17
WO 2010/018969 44 PCT/KR2009/004468
[00173] In the meantime, transmission power balancing of
respective antennas and/or transmission power balancing of
respective layers may be considered to be an important matter
in codebook design. If transmission powers of
individual
antennas are not adjusted for maximal uniformity, there
arises a difference in performance between respective
transmission antennas.
Likewise, if transmission powers of
individual layers are not adjusted for maximal uniformity,
there arises a difference in performance between respective
codewords.
[00174] Therefore, one embodiment of the present invention
proposes a method for designing a precoding matrix in
consideration of antenna power balancing using norms of all
elements (i.e., all elements of a specific row of the
precoding matrix) corresponding to individual antennas of the
precoding matrix. In more detail, the precoding matrix shown
in the following equation 31 may be utilized in the form of a
antenna power balancing shown in the following equation 32.
[00175] [Equation 31]
= = = ANL
P = k = . . =P 21 = P2NL
_PATTI = = = PNTNL _
[00176] [Equation 32]
CA 02731210 2011-01-17
WO 2010/018969 45
PCT/KR2009/004468
P11 ANL
11/3121
P21. P2NL=
2
P =k. )1P221 ===1 2NL )1P221 ===P22NL
PNTi PNTNL
VPN2Ti '=*PN2TNL VioN2T1 ...pN2TNL
[00177] On the other hand, one embodiment of the present
invention provides a method for designing a precoding matrix
in consideration of layer power balancing using norms of all
elements (i.e., all elements of a specific column of the
precoding matrix) of individual layers. In more detail, the
precoding matrix shown in the following equation 33 may be
utilized in the form of layer power balancing shown in the
following equation 34.
[00178] [Equation 33]
Pit ===
P=k=
=P21 = P2NL
. .
_PNT1 === PNTNL _
[00179] [Equation 34]
P11 ANL
Vpi21 ...pN2TI p N2 TN!,
P21 P2NL
P=k= õ2 4_ 2 2
'''PNTIPNL = = 'Prirk
PNT1 Pnirk
Pn121 ...pN2Ti p 12NL p N2T
-
[00180] In this case, differently from a Rank 2 precoding
matrix, it can be inappropriate for the number of rows and
CA 02731210 2011-01-17
WO 2010/018969 46
PCT/KR2009/004468
the number of columns in a 4Tx - Rank 3 precoding matrix to
simultaneously perform the antenna power balancing and the
power balancing.
However, in a specific situation, for
example, in a system of using a layer shift that changes a
layer used for transmission to another layer according to a
specific pattern in a transmission mode, there occurs an
effect in which a difference in performance between layers is
dispersed, the layer power balancing may be relatively less
important than the antenna power balancing.
Therefore, one
embodiment of the present invention proposes the use of a
precoding matrix acquired when the antenna power balancing is
firstly carried out on the condition that it is impossible to
simultaneously perform the antenna power balancing and the
layer power balancing.
[00181] In the meantime, the following precoding matrices
among the above-mentioned 4Tx - Rank 3 precoding matrices
indicate that the antenna power balancing can be carried out
because two symbols are transmitted to each layer, as denoted
by the following equation 35.
[00182] [Equation 35]
- - _
1 A 1
- pH 0 0 - ¨,_ v ¨ 1 0 0 _,_
V2 V2 X 1
0 P22 0
Po' = k = 0 0 p33 P4 = k = 5 w' A .5 P5 = '' . Y 1
-i-
P41 P42 P43 0 1 0 V2 V2
_.sh- Z IS- 0 0 0 Z
Y -
_ 0 _ _
CA 02731210 2011-01-17
WO 2010/018969 47 PCT/KR2009/004468
[00183] Similarly, in case of the following precoding
matrices among the 4Tx - Rank 3 precoding matrices, because
only one symbol is transmitted to one antenna, only the layer
power balancing can be carried out as shown in the following
equation 36.
[00184] [Equation 36]
1
Z -
-1 - - 0 _ _
1 0 0 1 0 0
, 0 0
V2 1
0 , 0 0 1 0 X 1 0
0 1 0 . =N12
P; = k = P2 = k = 13' = k = 0 0 ¨,- P7 = k =
0 0 1 0 0 1 3 ,12 Y 1
o o
x r z=
¨ o , o o o o 13-
A n C
- v ¨,-
0 0
X 1 0 X 1
1_ 0
Nri ,fi
P, = k =
Y 1 1,= k =
9
Y 1
0 0
õ B C A B
u 0
V3 V3 1/3 V3
- - -
[00185] In the meantime, in accordance with another
embodiment of the present invention, from the viewpoint of
simultaneous execution of both the antenna power balancing
and the layer power balancing, the present invention proposes
the 4Tx - Rank 3 precoding matrix including the following
precoding matrices denoted by Equation 37.
[00186] [Equation 37]
CA 02731210 2011-01-17
WO 2010/018969 48
PCT/KR2009/004468
_
1 0 0 1 0 0 1 0 0 0 0 0
(0' = 0 1 0 ' 0 1 0 ' 0 0 0 ' 1 0 0
PO ' 0010p(2) = 0 0 0 o I:" = e =
0 1 0 0 1 0
0 0 0 0 0 1 0 0 1 0 0 1
_
_ _ _ _ _
[00187] In other words, Equation 37 shows precoding
matrices used as the 4Tx - Rank 3 precoding matrices, and
each precoding matrix of Equation 37 is established to
transmit no signal to a single specific antenna.
[00188] In the meantime, examples of the precoding matrix
acquired when the layer power balancing is carried out on the
4Tx - Rank 4 precoding matrix can be represented by the
following equation 38.
[00189] [Equation 38]
-
0
1 n 1 1 -
-
V2 2 2 A/3 2 2
m2' 0 MD MN m21 0 MD MN
P:3 = k = Nii 2 2
PI.4 -- k= Ah 2 2
0 1 1fl33 1fl34 m3I 0 M33 In34
2 2 -ii 2 2
0 m42 m43 m44 0 1 '31n n''4
Vi
- 2 2 - 2 2 _
-
[00190] <Codebook Pruning>
[00191] In a 4Tx system, precoding matrices corresponding
to Rank 1, Rank 2, Rank 3, and Rank 4 can be used as elements
of a codebook used in transmission and reception ends.
However, in the case of using all precoding matrices, the
size of a codebook is excessively increased, so that it is
necessary to reduce the number of precoding matrices
simultaneously while maintaining performance at an
CA 02731210 2011-01-17
WO 2010/018969 49
PCT/KR2009/004468
appropriate level.
Embodiments capable of reducing the
number of precoding matrices will hereinafter be described in
detail.
Methods for restricting the following precoding
matrix can be independently or simultaneously utilized.
[00192] Codebook Element Alphabet Restriction
[00193] Each element of the above-mentioned precoding
matrices is selected from a complex number that has an
absolute value of 1 and a phase corresponding to any one of
+0 , +45 , +900, +135 , +180 , -135 , -90 , and -45 .
[00194] In one embodiment of the present invention, in
order to reduce the number of precoding matrices, each
element of the precoding matrices may be selected from a
complex number that has an absolute value of 1 and a phase
corresponding to any one of +0 , +90 , +180 , and -90 . That
is, each element of the precoding matrix may be selected from
- .1)
[00195] Otherwise, each element of the precoding matrix may
be extracted from a subset composed of N alphabetical letters
among 8 alphabets which are spaced apart from each other by
an angle of 45 .
[00196] Restriction to Unitary Precoding Matrix
[00197] In the case where individual column vectors
contained in a precoding matrix have orthogonality
CA 02731210 2011-01-17
WO 2010/018969 50
PCT/KR2009/004468
therebetween, the precoding matrix may be a unitary matrix or
a partially unitary matrix. If the precoding matrix has the
above-mentioned characteristics, an additional gain can be
obtained.
[00198] Thus, in accordance with one embodiment of the
present invention, unitary matrices or partially unitary
matrices among all the aforementioned precoding matrices are
collected so that a codebook can be formed.
[00199] For example, the row/column permutation is carried
out on the precoding matrices shown in the following equation
39 and the precoding matrices shown in the following equation
40 so as to obtain a few matrices, and the obtained matrices
are combined, so that a codebook can be generated.
[00200] [Equation 39]
-1- . -1 0- 1 0 1 -
pM. a p(2)= a 0 a
b 0 1 1 t
r
c
-. 11 0: fi-
, - 1
- -.. -
1 0 1 0 . .1 [ 1. 0
a 0 -a 0 0 c -c c
Flo = FT) =
0 1 0 1 a 0 -a -a
0 b 0 - b b - b 0 b
r r
b .1- i -1-i
a c e , 1, ¨,_ i --,_ .,-1 ,_ ¨i ,
where,
[00201] Restriction to Nested Structure
CA 02731210 2011-01-17
WO 2010/018969 51
PCT/KR2009/004468
[00202] When constructing the precoding matrices of Rank 1,
Rank 2, Rank 3 and Rank 4, in the case where the precoding
matrix of Rank 2 or Rank 3 can be constructed with column
vectors of the Rank 4 precoding matrix, the constructed
precoding matrix is called a precoding matrix with a nested
structure. If a specific Rank 4 precoding matrix is used as
a part of a precoding codebook, the Rank 3 precoding matrix
should be configured with column vectors of the specific Rank
4 precoding matrix, such that there occurs a limitation in
the construction of the precoding matrix. Thus, the codebook
size can be limited according to the aforementioned norm or
standard.
[00203] In one embodiment of the present invention, the
precoding matrix of Rank 1, Rank 2, Rank 3, or Rank 4 may
have a nested structure.
[00204] For example, a codebook can be constructed with a
combination of matrices acquired by performing the row/column
permutation on the precoding matrices shown in the following
equation 40.
[00205] [Equation 40]
_
1 0 1 1 0 1 0
po) a p(2) a 0 a 0 -a
PP)= n(4) ..= a 0 -a 0
b 0 1 0 1 0 1-1
0 1 0 1
C 0 b 0 b 0 0 b 0 -b
- _
_ _ - _
, , , -
b {1+j .1-j -1-j .-1+j
a,,c E 1, --- , J, , - 1, ______ ,.:-.. A.
where,
CA 02731210 2011-01-17
WO 2010/018969 52
PCT/KR2009/004468
[00206] In addition to the matrices shown in the above-
mentioned equations, other applicable matrices may also exist.
It can be easily understood that the applicable matrices can
be obtained by performing the row permutation and/or the
column permutation on the above-mentioned matrices.
In the
present invention, because the precoding matrix has elements
each having a value of 0, a certain antenna may not be mapped
to a specific input stream. This operation may be recognized
as an antenna selection function.
[00207] II. Detailed Format of Codebook
[00208] Hereinafter, in the case where a codebook is
designed to satisfy the above-mentioned codebook design rule,
a method for deciding a precoding matrix for each rank
contained in the codebook in consideration of a chordal
distance will be described in detail.
[00209] FIG. 9 is a conceptual diagram illustrating a
chordal distance.
[00210] A chordal distance is well known as one of norms
(or standards) for comparing performances of various codebook
sets.
Herein, the term "chordal" indicates a straight line
between two points located at the circumference. Therefore,
given a two-dimensional (2D) case, a chordal distance
indicates a distance between two points located at the
CA 02731210 2011-01-17
WO 2010/018969 53
PCT/KR2009/004468
circumference of a circle (e.g., a unit circle) as shown in
FIG. 9.
[00211] There is a need for the 4Tx - codebook to consider
a four-dimensional chordal distance, so that the following
equation 41 can be used as a chordal distance for selecting a
codebook set.
[00212] [Equation 41]
H
d(.1 , Q) = 1 111 3H ¨ QQH
[00213] In Equation 41, P is 134171 V2
V/v] , and Q is
Q = [Ui U2 =
= = UN , where v, and u, ( i= 1,2,===N , N=4 in the
case of 4Tx antennas) are principal vectors of the matrices P
and Q, respectively.
In addition, AOF= 111Ern En lay 12 = Vtrace(4,4 H
1=1 j=1
is the Frobenius norm of the matrix.
The above-mentioned
chordal distance can also be measured by the following
equation 42.
[00214] [Equation 42]
1
c1,03,Q)=APPH ¨QQ1
2
= trace(AA HBBH)
where A and B are orthonormal generation magtrices for P and Q respectively
[00215] The above-mentioned codebook design for the 4Tx
system based on four transmission antennas will be described
CA 02731210 2011-01-17
WO 2010/018969 54
PCT/KR2009/004468
using the above-mentioned chordal distance concept.
For
convenience of description and better understanding of the
present invention, a factor related to the power balancing
will be omitted from the following expressions.
[00216] Rank 2
[00217] Firstly, it is assumed that the following codebooks
of three groups capable of maintaining good CM performance
about the 4Tx - Rank 2 system are used.
[00218] [Equation 43]
1 0 1 0 1 0 1 0
x0 X 0 ¨X 0 ¨X 0
Group 1X,YE j __ 1
0 1 0 1 0 1 0 1 " '
0 Y 0 ¨Y 0 Y 0 ¨Y
\- _/
4- 1 0 1 0 1 0 1 0 -N
0 1 0 1 0 1 0 1
Group 2X ,Y e{1,1+ j
-5'}
0 Y 0 ¨Y 0 Y 0 ¨Y
1-
1 0 1 0 1 0 1 0
0 1 0 1
Group 3 0 1 0 1
XYeL ____________________________________________________________ j
0 Y' 0 ¨Y' 0 Y' 0 ¨Y V2 J
\_X 0 X 0 ¨X 0 ¨X 0
[00219] While the number of precoding matrices satisfying
the above-mentioned formats may be a considerably high number,
it is preferable that a codebook for including a
predetermined number of precoding matrices be designed
according to a reasonable norm.
The following description
CA 02731210 2011-01-17
WO 2010/018969 55
PCT/KR2009/004468
proposes a method for limiting the number of precoding
matrices for each rank to a predetermined number or less
using the following norms.
[00220] First Norm (Norm 1): Chordal distance
[00221] Second Norm (Norm 2): Reference indicating whether
the precoding matrices are uniformly selected from individual
groups. If the number of precoding matrices/vectors in a
codebook is not divided by the number of groups, the
precoding matrices are most uniformly selected in
consideration of the first norm (Norm 1).
[00222] The above-mentioned norm can be equally applied not
only to Rank 3 but also to Rank 4 to be described later.
[00223] In more detail, one embodiment of the present
invention proposes a method for selecting the set of
precoding matrices from a codebook about a specific rank
using the Norm 1. In a first step, a chordal distance about
all precoding matrix pairs contained in a single codebook is
calculated using Equation 42. For example, if four codebook
sets exist, four minimum chordal distance values can be
represented by the following expression.
[00224] [Expression]
[00225] 1
dc,min = 1 d2 = 0.56 d3
I CAin r c,min = 0.71 and dc4min =1
[00226] In the above expression, the higher the value of
c,min (where i is a codebook set number), the higher the
CA 02731210 2011-01-17
WO 2010/018969 56
PCT/KR2009/004468
system performance. Thus, it is preferable that first and
fourth codebooks go to a next selection step.
[00227] In a second step, in order to support a variety of
wireless channel environments, the present invention proposes
a method for most uniformly selecting the precoding matrices
for each group. For example, in accordance with the proposed
method of the present invention, if there are three codebook
groups and 16 precoding matrices are needed as the Rank-2
codebook, 5 precoding matrices are selected from two groups,
and 6 precoding matrices are selected from the remaining one
group. For example, in accordance with the proposed method of
the present invention, 5 precoding matrices are selected from
first two groups, and 6 precoding matrices are selected from
the last one group. One embodiment of the present invention
may consider a method for limiting alphabets of each
precoding matrix as described above, in which, for example,
an alphabet 'X' may be limited to X = 1, j, -1, or -j. The
following description illustrates exemplary 4Tx Rank-2
codebooks capable of being configured by the above steps.
[00228] [Table 1]
Rank-2 codebook set 1-1
CA 02731210 2011-01-17
WO 2010/018969 57 PCT/KR2009/004468
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 1 0 -1 0 0 1 0 1 0 1 0 1 0 1
0 j 0 1 0 1 -j 0 j 0 -j 0 -1 0 0 -1
1 0 0 j 0 -1_ 0 -1 _O 1_ 0 1_ Of-JO
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 1 0 JO -JO 0 1 0 1 0110
j 0 '0 1 '0 1 0 1 1 -1' 0 1'0 1
0 -1 0 -J 0 1 0 1 -J 0 -1 0 -1 0 0 -1
Rank-2 codebook set 2-1
- 1 0 1 0 1 0 1 0 1 0 1 0 -
1010
¨JO 0 1 ¨1 0 0 1 ¨j 0 0 1 0110
0 1 ' 0 0 j 0 ' 0 1'1 0'¨j 0'0
0 ¨1 ¨1 0 0 j 0¨f 0 1 0 1 0 JO]
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 1 0 0 1 0 1 ¨1 0 0 1
0 ¨j 0 j 1 0 0 1 0 ¨j 0 ¨1 0 1 ¨j 0
¨j 0 ¨j 0 0 ¨1 0 ¨j j 0 ¨1 0 0 ¨j 0 ¨j
Rank-2 codebook set 3-1
- 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
0-
-j 0 ¨1001 0 1 0 1 0 1 JO 0 1
0 1 0 1 1 0 ¨1 0 0 ¨j 0 j 0 1 ¨1
0 ¨1 0 j 0 ¨1 0 1 1 0 1001¨JO
- 1 0 1 0 1 oi 0 -1 0- -1 0- - 1 0
1 0
0 1 0 1 0 1 j 0 0 1 1 0 0 ¨j
¨1 0 0 1 ¨j 0 '0 1 j 0'0 0 ¨j' 0 1
0 ¨1_ 0 ¨ j_ _O ¨1 Of _0 j_ _¨ 1 0 _ 0 1
Rank-2 codebook set 4-1
CA 02731210 2011-01-17
WO 2010/018969 58 PCT/KR2009/004468
1 0 ¨ 1 0¨ 1 0 -1 0 1 0- -1 0 -
1 0- -1 0--
0 1 0 1 - j 0 0 1 0 1 0 1 0 1 0 1
-j 0 0 j' 0 1 '0 -j'O l'j 0 0
1 1 0'
0 -j -j 0 0 -j j 0 1 0 0 -j -1 0 0 j
1 0 1 0 1 0 0 0 1 0 0 1 0
0 1 j 0 j 0 0 1 0 0 1 -1 0 - j
0
0 -j'O 1 '0 1 '-1 0 '0 1 '-1 0'
1 0 1
-f 0 0 - j 0 j 0 -
1 0 1 0 1 0 -1 0 j
Rank-2 codebook set 5-1
1 0 1 0 1 0 1 0 1 0 1 0 1010
0 1 0 1 0 1 1 0 0 1 1 -j 0 0 1
j 0' 0 -1'-f 0' 0 1' 0 1' -j 0 1' 1 0 '
Of 1 0 0 1 0 1 -1 0 = 0 0 j 0 -1
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 -1 0 1 0 j 0 j 0 0 1 0 1 0 1
0 1' 0 1 0 1 0 1 0 1' 0 -j -1 0 j 0
1 0 0 -1 0 -1 0 -j _0 j _-j 0 _0 -1 _0 -j
Rank-2 codebook set 6-1
1 0 1 0 0 1 0 1 0 1 0 1 0 1
0 1 1 0 0 1 0 1 0 1 -1 0 0 1 -j
1 0' 0 1 -j' -j 0 0 f 0 1 -j 0' 0
0 1 0 fi 0 0 -j_ _-j 00
10 j_ _ 0 -j_
1 0 1 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 0 1 j 0 0 1 -1 0 0 1 0 1
j 0' 0 '1 0 '0 1 '0 j' 0 1' 0 -1' j 0
0 j -1 0 -1 0 -j_ _j 0 _0 -1 _-1 0 _ 0-f
Rank-2 codebook set 7-1
CA 02731210 2011-01-17
WO 2010/018969 59 PCT/KR2009/004468
- 1 0 1 0 1 0 1 0 1 0 1 0 1 0
1 0
¨Jo 0 1 0 1 0 1 1 0 j 0 0 1 0 1
0 1 '0 j'0 ¨j' 0 1 '0 1 '0 1 = 0' 1 0
0 ¨j j 0 1 0 ¨j 0 0 1 0 j 0 1 0 ¨j
- 1 0 101 0 1 0 1 0 1 0 1 0 1
0
j 0 0 1 0 1 1 0 ¨1 0 0 1 0 1 0 1
0 1 '-1 0' 0 ¨1'0 1 1 1 O'j 0 0 ¨j
0 ¨j 0 j ¨j 0 0 ¨1 0 1 0 j 0 ¨1 ¨1 0
Rank-2 codebook set 8-1
1 0 101010 1 0 1 0 1 0 1 0
j 0 010101 0 1 0 1 0 1 1 0
0 1 -1 0 j 0'0 -j' 0 j 0 '1 0 '0
0 -1 0 -1 0 j 1 0 -j 0 0 -j 0 -1 0 1
- 1 0 1 0 1 0 1 0 0 1 0 1 0 1
0
-JO 0 1 0 1 0 1 -1 0 -j 0 1 0 1
0 l'-j 0'1 0'0 0 1 '0 j' 0 -1
0 j_ _ 0 j_ 0 1_ _j 0 _ 0 1_ _ 0 - j_ _1 0_ 0 _
Rank-2 codebook set 9-1
n 1 O__1 0- -1 0 1 0- 1 0- 1 0 - -1 0- r-1 0
¨100101 0 1 0 1 0 1 0 1 JO
0 1 0 0 ¨1'¨1 0 0 1'¨1 0 0'O 15
0 ¨1 1 0 j 0 0 1 ¨1 0 0 ¨1 0 0 1
-1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
1 0 1 0 ¨j 0 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 0 ¨15 0 1 0 0 ¨j5 1
0 0 0 1 ¨j 0 0 0 1 1 0 0 ¨1
Rank-2 codebook set 10-1
CA 02731210 2011-01-17
WO 2010/018969 60 PCT/KR2009/004468
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 ¨JO 0 1 ¨1 0 0 1 0 1 0 1 0 1
1 0 1 '0 j' 0
1 '-1 0' 0 ¨j'¨1 0'0 ¨1'
0 ¨ j 0 -1 j 0 0 -j 0 1 -j 0 -1 1 0
1 0 1 0 1 0 0 1 0 1 0 1 0 1 0
1 0 0 1 1 0 j 0 0 1 0 1 0 1 0 1
0 1 0 0 1 0 1 0 -j 'O 1'-J 0 0
1
_0 j_ _0 j_ _0 - j_ _0 -1_ _j 0 _ _1 0_ _ 0 j_ _-1 0_
Rank-2 codebook set 11-1
1 0 1 010101 0 1 0 1 0 1 0
0 1 -j 0 1 0 0 1 -1 0 0 1 0 1 0 1
1 0' 0 1 '0 j 0' 1 -1 0 ' -1 0 0 -j
Of 0 -1 0 1 0 1 -j 0 j 0 -j -1 0
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 0 1 j 0 0 1 -1 0 0 1
0 0 -1'0 -j'j 0 '0 1 '0 -1'
0 0 1
JO-JO 1 0 0 -1 0 -1 j 0 0 j j 0_
Rank-2 codebook set 12-1
1 0 1 0 1 oi 0 - -1 0 - 1 0 - 1 0 -1 0-
0 1 1 0 jOO 1 0 1 0 1 -j 0 0 1
-1 0 1 0 1 -j 1 0 0 -1 0 1
0 j
0 0 -1 0 j = 0 0 -j -1 0 0 -j j 0
1 0 1 0 1 1 0 1 0 1 0 1
0 1 0
0 1 0 1 -1 1 0 0 1 0 1 0
1 j 0
0 1 ' -j 0' 0 1'O 1'J 0'0 -1'1 0'01
1 0 0 -1 0 1 0 1 0 -1 1 0 0 j 0 -j
[00229] The above-mentioned codebooks shown in Table I are
disclosed only for illustrative purposes, and row permutation
CA 02731210 2011-01-17
WO 2010/018969 61
PCT/KR2009/004468
and/or column permutation may be applied to all or some of
the precoding matrices.
[00230] If the 4Tx Rank-2 codebook includes 15 precoding
matrices, one precoding matrix may be removed from a group of
selecting the largest number of precoding matrices among
individual precoding matrix groups. The following
description shows exemplary 4Tx Rank-2 codebooks configured
by the above-mentioned schemes.
[00231] [Table 2]
Rank-2 codebook set 1-2
10101 0 1 0 1 0 1 0 1 0 1 0
0 1 1 0 -1 0 0 1 0 1 0 1 0 1 0 1
0 j 1 1 -j 0 j 0 -j 0 -1 0 0 -1
1 0 -j_ 0 -1 0 1 0 1 0 j_ 0 _
71 0 7 -1 0 - -1 0- - 1 0- - 1 0 - 1 oi 0 7
0 1 1 0 JO 0 1 0 1 01 j 0
j 0'0 1 '0 l' 0 1' -j' 0 ='0 1
0 -1 0 -j 0 1 -j 0 -1 0 -1 0 0 -1
Rank-2 codebook set 2-2
- 1 0 1 0 1 0 1 0 1 0 1 0 1 0
-j 0 0 1 -1 0 0 1 0 1 0 1 1 0
0 1 0 1' 0 1' j 0 1 O'-j 0' 0 1'
0 -1 -1 0 0j0-j01 0j0j
- 1 0 1 0 1 0 1 0 1 0 1 0 1
0_i 0
0 1 0 1 0 1 1 0 0 1 0 1 -1 0 0 1
0 -j ' 0 j' 1 0 ' 0 1 ' 0 0 -1' 0 1 ' -j
0
-J 0 _-j 0 0 -1 0 -j_ _j 0 _ _-1 0 _ _ 0 -JO -1
CA 02731210 2011-01-17
WO 2010/018969 62 PCT/KR2009/004468
Rank-2 codebook set 3-2
_
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
-j 0 -1001 0 1 0 1 0 1 JO 0 1
0 1 ' 0 1 ' 1 0 ' -1 0 ' 0 -j ' 0
j ' 0 1 ' 0 -1 '
0 -1 0 j 0 -1 0 1 1 0 1 0 0 1 - j 0
_
_
1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 j 0 0 1 1 0 0
-1 0 ' 0 1'-J ' 0 1 ' j
0'O l' 0 -j
0 -1 -j 0 0 -j 0 -1 0 j 0 j -1 0
-
Rank-2 codebook set 4-2
_
1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 0 1 0 1 0 1 0 1
-j 0 ' 0 j' 0 -J'0 1'J 0 ' 0 1' 1 0'
0 -j -j 0 j 0 1 0 0 -j -1 0 0 j
_
_
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 j 0 JO 0 1 1 0 0 1 -1 0 -j 0
0 -j ' 0 1 ' 0 1 ' -1 0 ' 0 1 -1 0 0 1 ' 0 1
-j 0 0 -j 0 j 0 -1 0 1 0 1 0 -1 0 j
_ _
Rank-2 codebook set 5-2
_
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 1 0 0 1 0 1 -j 0 0 1
J 0 0 -1 -j 0 0 1 0 1 0 -j ' 0 1 ' 1 0 '
0 j 1 0 0 1 0 1 -1 0 j 0 0 jO -1_
1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 -1 0 1 0 JO 0 1 0 1 0 1
0 1'O 1 0 1 ' 0 1 ' 0 -j -1 0 ' j 0
1 0 0 -1 0 -1 0 j -j 0 0 -1 0 -j
_ _
CA 02731210 2011-01-17
WO 2010/018969 63 PCT/KR2009/004468
Rank-2 codebook set 6-2
1 0 1 0 0 1 0 1 0 1 0 1 0 1 0
0 1 1 0 1 0 1 0 1 -1 0 0 1 -j 0
1 0 0 1 -j -j 0 0 j 0 1 -j 0 0 1
OlOji 0 0 - j_ j 0_ _ 0 1_ _ 0 j_ _ 0 - j
- 1 O1 0 -1 0-1-1 0--1 0--1 0 -1 0
0101 j 0 0 1 -1 0 0 1 0 1
0 '1 0'O 1 '0 j' 0 1 ' 0 -1' j 0
_-1 _0 -1__O - _j 0_ _ 0
-1_ _-1 0 _0 -j_
Rank-2 codebook set 7-2
- 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
-JO 0 1 0 1 0 1 1 0 j 0 0 1 0
1
0 1 '0 j'0 -j' 0 l'O l'O j 0'1 0
0 -j j 0 1 0 -j 0 0 1 0 j 0 1 0-f
1 0 1 0 1 0 1 0 1 0 1 0 1 0
J 0 0 1 0 1 -1 0 0 1 0 1 0 1
0 1 '-1 0' 0 -1' 0 l'l 0 j 0 0 -j
0-f Of-JO 0 1 0 j 0 -1 -1 0
Rank-2 codebook set 8-2
1 0 1 0 1 0 1 0 1 0 1 0 1 0
j 0 010101 0 1 0 1 1 0
0 1 '-1 0 j 0'0 -j' 0 1'1 0 0
0 -1 0 -1 0 j _1 0 _ 0_ _0 -1_ _0 1
-
1 0 1 0 1 0 1 01 0 1 0 1 0 1 0
-j 0 0 1 0 1 0 1 -1 0 -j 0 0 1 0 1
0 l'-j 0'1 0'0 1' 0 1' 0 1 '0 j' 0 -1
Of 0 j 0 1 JO 0 1 0 -j 1 0 -j
CA 02731210 2011-01-17
WO 2010/018969 64 PCT/KR2009/004468
Rank-2 codebook set 9-2
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
-100101 0 1 0 1 0 1 0 1 JO
0 1 0 j 0 -1 -1 0 0 1 -1 0 j 0 0 1
0 -11 0 j 0 0 1 -1 0 0 -1 0 j 0 1
_ _ _ _ _ _
-1 0 -1 0- - 1 0- 1 0 -1 0 -1 0 -1 0
1 0 1 0 -j 0 0 1 0 1 0 1 0
0 1 0 1 0 1 0 -1 j 0 0 -j 1 0
0 -j 0 j 0 1 -j 0 -j 1 00-1
Rank-2 codebook set 10-2
1 0 1 0 1 0 101 0 1 0 1 0
0 1 -JO -1 0 0 1 0 1 0 1 0 1
1 0 0 1 0 1 -1 0 0 -j -1 0 0 -1
0 -j 0 -1 0 -j_ _0 1_ _-j 0 _ _0 -1 _1 0_
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
1 0 0 1 1 0 j 0 0 1 0 1 0 1 0 1
0 j 0' 0 1 0 1 -J'0 l'-j 0' 0 1
_0 j_ _0 j_ _O -_0 -1J 0 _ _1 0_ _ 0 j_ _- 1 0_
Rank-2 codebook set 11-2
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 - j 0 1 0 0 1 -1 0 0 1 0 1 0 1
1 0 0 1 0 1'J 0 0 1 -1 0 -1 0 0 -j
0 j 0 -1 0 1 0 1 0 -j 0 j 0 -j -1 0
1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 0 1 j 0 0 1 -1 0 0 1
0 1 0 -j j 0 0 1 0 -1 0 1 0 1
J 0 1 0 0 -1 0 -1 j 0 0 j -j 0
CA 02731210 2011-01-17
WO 2010/018969 65
PCT/KR2009/004468
Rank-2 codebook set 12-2
1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 1 0 0 1 0. 1 0 1 -j 0 0 1
-1 ' 0 1 0 -j' 1 0 -1' 0 1 '0 j'
0 0 -1 j 0 0 -j -1 0 0 -j j 0
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
0 1 0 1 -1 0 1 0 0 1 0 1 0 1 j 0
0 l'-j 0' 0 l'O j 0'0 -1'1 0'0 1
1 0 0 -1 0 1 0 1 0 -1 1 0 0 j 0 -j
[00232] The codebooks shown in Table 2 are also disclosed
only for illustrative purposes, the row permutation and/or
column permutation may be carried out on all or some of
precoding matrices of the codebooks.
[00233] Rank 3 - First Embodiment
[00234] In order to design the 4Tx Rank-3 codebook so as to
maintain good CM properties, it is assumed that the following
three precoding matrix groups are used. For convenience of
description, a factor related to power balancing will be
omitted herein.
[00235] [Equation 44]
Group 1
(-1 0 1 - -1 0 1 ¨0 1 0 - -0 1 0
X 0 -X X 0 -X 0 X 0 0 -X 0
1+j 1--j
0 1 0 '0 1 0 '1 0 1'1 0 1
0 Y 0 0 -Y 0 Y 0 -Y Y 0 -Y
CA 02731210 2011-01-17
WO 2010/018969 66
PCT/KR2009/004468
Group 2
(- 1 0 1 --1 0 1 0 1 0 0 1 0 -\
0 1 0 0 1 0 1 0 1 1 0 1
1-1
0 -X' 0 X 0 ' 0 -X 0 x,yE
',j,
0 Y 00 -Y 0 Y 0 -Y Y 0 -Y
_ _ -1
Group 3
1-1 0 1-1 0 1 To 1 0 0 1
0 1 0 0 1 0 1 0 1 1 0 1
0 Y 0 0 -Y 0 Y 0 Y 0 -Y x,YE
X 0 -X X 0 -X 0 X 0 0 -X 0
[00236] In case of Rank 3, the present invention proposes a
method for constructing a codebook according to the above-
mentioned Norm 1 and Norm 2 in the same manner as in Rank 2.
In more detail, a chordal distance about all precoding matrix
combinations available in a codebook is calculated using
Equation 42, and then a minimum number of sets each having a
maximum chordal distance can be selected. In addition, the
present invention proposes a method for most uniformly
selecting the precoding matrix from each group (Group 1, 2,
or 3). If the letter denoted by a precoding matrix component
of each group is restricted to (1, j, -1, -j), the following
codebook capable of satisfying a minimum chordal distance
dc, =0.707 can be obtained.
[00237] [Table 3]
Rank-3 codebook set 1-1
CA 02731210 2011-01-17
WO 2010/018969 67
PCT/KR2009/004468
1 0 1 1 0 1 0 1 0 1 0 1 1 0 1 1 0 1 1 0 1 0 1 0
0 1 0 0 1 0 1 0 1 0 1 0 0 1 0 JO-jO 1 0 1 0 1
j 0 -j'O j 0'0 j 0' 0 -j 0 ' 0 1 0 ' 0 1 0 '
= 0 -j'O 1 0 '
Of 0 _j 0 -j _1 0 -1_1 0 -1 _1 0 -1õ0 1 0 -j 0 _ _j 0 -j
_
1 0 1 1 0 1 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1
0 1 0 j 0 - j 1 0 1 Of 0 1 0 1 0 -j 0 0-10
1 0
1 0 -1 0 1 0 ' 1 0 -1 ' 1 0 1 0 -f 0 ' 1 0 1 ' 1
0 1 ' j 0 -f
0 1 0 0 -1 0 0 1 0 j 0 -j 1 0 -1 1 0 -1 j 0 -j 0 j 0
_ _ _ _
_
Rank-3 codebook set 2-1
_O 1 0 - -0 1 0 - -0 1 0 - -1 0 1 - -1 0 1- 1 o
1 - -0 1 0- -0 1 o -
0-10 Of o 1 o 1 o 1 0010 = o ¨j 1 o 1 1 o 1
1 o 1'1 o 1 ' j o ¨j'o ¨1 o '1 o ¨1' 0 1 o '1 o ¨1'j o
¨j'
1 o
¨1_ _j o ¨JO j o _ _j o -j_0 j 0 _ -j 0 _ _0 1 0 _ -j 0 _
_
0 1 0 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 1 0 ii 0 1
1 0 1 1 0 1 1 0 1 j 0 - j 1 0 1 0 1
0 0 1 0 j 0 - j
0 -1 0 ' 0 1 0 ' 0 j 0 ' 0 1 0 ' 1 0 -1' 1 0 -rj 0 -j' 0 1 0
_j 0 - j_ _1 0 -1_ _j 0 -JO -1 0 _ _0 -1 0 _ 0 -1 0 _0 -J 0 _ _0 1 0 _
Rank-3 codebook set 3-1
-1 0 1-1-0 1 0- -0 1 o - -0 1 o -0 1 o -0 1
010 1 - -1 o 1 -
010 1 o 1 1 o 1 1 o 1 0-10 10101001 o
j o ¨j ' o 1 o ' j o ¨j ' o ¨j o ' 1 o 1 ' 1 o ¨1' o 1 o ' j o ¨j '
_o 1 o o ¨JO 1 o = o ¨ j _j o ¨ j _o j o o
¨1_ _o ¨1 o _
o 1 o 1 o 1 o 1 o o 1 01 o ii o ii o 1 1 o 1
1 o 1 j o ¨j o ¨j o 1 o 1 o 1 0 j o ¨j 1 o ¨1 o 1 o
O j o ' o 1 o ' 1 o 1 ' o ¨1 0 ' o ¨j 0
' o 1 o ' o 1 o ' o ¨1 o
1 o ¨1 o ¨j o 1 o ¨1 j o ¨j 1 o ¨1 o 1 o o ¨1 o 1 o ¨1
_ _ _ _ _ _
Rank-3 codebook set 4-1
CA 02731210 2011-01-17
WO 2010/018969 68
PCT/KR2009/004468
_
_ _ _
1 0 1 1 0 1 0 1 0 0 1 0 0 1 0 1 0 1 1 0 1 0 1
1 0 -1 0 1 0 0 1 0 0 -1 0 1 0 1 0 1 0 1 0 -1 1 0
0 1 0 ' 1 0 -1 ' 1 0 1 ' 1 0 1 '0 -j 0 '0 1 0
'0 1 0 '0 -1 '
0 -1 0 _0 j 0 _j 0 -j _j 0 -j_ _j 0 -j_j 0 -j_0 -j 0__1 0 -1_
_
_ _ _
1 0 10 1 0 1 0 1 0 1 0 1 0 1 1 0 11 0 10 1 0
10-1 101010 Of 0 0 1 0 0 1 0 0 1 0 0 1
0 1 0 ' 0 -1 ' 0 -1 0 1 0 1 ' 1 0 -1 ' 0 -j 0 '
0 -1 ' ' 0 -j
0 1 0 0 1 0 1 0 -
1_ _j 0 -j_ _0 -1 0_ _j 0 -j_ _0 1 0 0 -j 0
_ _
Rank-3 codebook set 5-1
1 0 1 1 0 1 1 0 1 1 0 1 0 1 0 0 1 0 1 0 1 1 0 1
0 1 0 0 1 0 0 1 0 1 0 -1 0 -1 0 1 0 1 1 0 -1 0 1 0
1 0 -1 ' 0 -j 0 '0 1 0 '0 1 0 ' 1 0 1 '0 -j 0 '0 1
0 '0 j 0 '
0 1 0 1 0 -1 1 0 -1 0 -1 0 1 0 -1 1 0 -1 0 1 0 j 0 -j
_ _ _ _ _
1 0 ii 0 1 0 1 0 0 1 0 0 1 0 0 1 0 1 0 ii 0 1
0 1 0 0 1 0 1 0 1 1 0 1 1 0 1 0 -j 0 0 1 1
0 -1
0 -1 0 '1 0 -1'1 0 -1'j 0 -j'0 j 0 ' 1 0 1 'j 0 -f'0 1 0
_j 0 -.L.!) j 0_0 Of 0_0 -1 0 _ _1 0 -1_ j 0 -j_ _0 -1 0-
f 0
Rank-3 codebook set 6-1
_
1 0 1 -1 0 1 - -0 1 o -0 1 0 ¨0 1 0 --0 1 0 ¨1 o 1 - To 1 o
-
0 1 00100-10 0-f 0101 1 o 1 01010 1
i o -i'l o -ri o 1'1 o 1 '0 1 o'f o -i'i o -.Co -f o '
o -1 o o 1 o 1 o -1 1 0 -1 1 0 -1 0 i o o i o i 0 -f
_ _ _ _
o 1 00 1 o 1 o ii o 1 1 o 1 o 1 o 1 o 1 o 1 o
o f o 101010 o -f o 1 o 1 o 1 1 o -1 1 o 1
1 0 1 ' f 0 '0 i 0' 1 o '0 1 0 'f 0 -f'o 1 o'f 0 -f
1 o -1 o -1 1 o -1 1 o 1 o -1 o 1 o o f o o -1 o
_ _ -- _ _
Rank-3 codebook set 7-1
0 1 0 1 0 1 0 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1
0-10 0 1 0 1 0 1 0 1 0 1 0 -1 j 0 -j 1 0 -1 0 1 0
1 o 1'1 0 -i'f 0 -f'o f o'o 1 o'o 1 o 'o
1 o'o 1 o'
_f o --f_ _o 1 o _ _o 1 01 o -1_ _o f o_ _o -f o _ _o -1 o _ _f o -
_O I o - -0 1 o - -0 I o -0 1 o -1 -0 1 0-' -0 1
0 - -1 0 1 - -1 0 1 -
1 0 1 1 0 1 1 0 1 0 -j 0 1 0 1 1 0 1 0 1 0 0 1 0
0 i 0 ' 0 -j 0 ' 0 1 0 ' 1 0 1 ' 0 -1 0 ' i 0 -i ' 0 -j 0 ' i 0 -f
j 0 - j_. j 0 - j_ j 0 - j_ j 0 -I_ ...1 0 -1_ ...0 -1 0 _ j 0 -f _0 -
f 0 _
CA 02731210 2011-01-17
WO 2010/018969 69
PCT/KR2009/004468
Rank-3 codebook set 8-1
-0 1 0 0 1 0 ¨1 0 1 -1 0 1 -
0 1 0 ¨1 0 1 -1 0 1 1 0 1
010 10 1 10 -1010 Oj 0 01 0 01 0 j0-j
1 0 1 j 0 -j 'O 1
0 0 -j 'l 0 1 0 -1 0'l 0 -1'O 1 0
_j 0 -j_ _0 -1 0 _ _0 -j 0_0 1 0 _ = 0 -j_1 0 -1_ -
j 0_0 j 0_
- - - - - - -
0 1 0 0 1 0 1 0 1 0 1 0 1 0 1101101101
1 0 1 1 0 1 j 0 -j 0 -1 0 0 1 0 0 1 0 0 1 0 0 1 0
1 0 -1 ' 0 -j 0'O 1 0 1 0 1 1 0 -1 ' 0 -j
0'! 0 -1 ' 0 j 0
0j0 j 0 -j0-10 JO -j0-10 10 -10j0 j0-j
- - - - - - - - - - -
[00238] It should be noted that the row permutation and/or
the column permutation may be carried out on all or some of
precoding matrices of the above codebooks shown in Table 3.
[00239] If only 15 precoding matrices are included in the
Rank-3 codebook, one precoding matrix of a group for
selecting the largest number of precoding matrices among
individual groups is removed from the codebooks shown in
Table 3, so that the removed result may be configured as
shown in the following Table 4.
[00240] [Table 4]
Rank-3 codebook set 1-2
1 0 1 1 0 1 0 1 0 1 0 1 1 0 1 1 0 1 1 0 1
01 0 01 0 1o1 0 1 0 010 =0-j 0 1 0
j 0 -j 0 j 0'0 j 0'0 -j 0'0 1 0'0 1 0 j 0 -j
_0 j 0 j 0 -f 1 0 -1 1 0 -1 1 0 -1 0 1 0 0 -j 0
-1 0 1 -1 0 1 -0 i 0 - 70 1 0 -0 1
0 0 1 0 -0 1 00 1 0
0 1 0 j 0 -j 0
1 0 j 0 1 0 1 0 - j 0 0-101 0 1
1 0 -1 0 1 0 1 0 -1' 1 0 1 0 -j 0 1 0 1
1 0 'j 0 -j
0 1 0 0-1 0 0 1 Of _j 0 -j_ _1 0 -I __i 0
_j 0 -j_ Of 0_
Rank-3 codebook set 2-2
CA 02731210 2011-01-17
WO 2010/018969 70
PCT/KR2009/004468
0 1 0 0 1 0 0 1 0 1 0 1 1 0 1 1 0 1 0 1 0 0 1 0
0 -1 0 0 j 0 1 0 1 0 1 0 0 1 0 j
0 - j 1 0 1 1 0 1
1 0 1 1 0 1 ' j 0 -j ' 0 -1 0 ' 1 0 -1 ' 0 1
0 ' 1 0 -1 ' j 0 -j '
1 0 -1 j 0 -j 0 j 0 ' 0 -j 0 j
0 0 -j 0 0 1 0 -j 0
_ _ _ _ _ _ _
0 1 0 0 1 0 1 0 1 0 1 0 1 0 1i 0 ii 0
1 0 1 1 0 1 j 0 - j 1 0 1 0 1 0 1 0 j 0 - j
0-i 0 0 1 0 ' 0 1 0 ' 1 0 -1' 1 0 -1'j 0
-j' 0 1 0
j 0 -j 1 0 -1 0 -1 0 0 -1 0 0 -1 0 1 0 0 -j
_
Rank -3 codebook set 3-2
1 0 1 0 1 0 0 1 0 0 1 0 0 1 0 1 0 1 1 0 1
0 1 0 1 0 1 1 0 1 0-10 1 0 1 0 1 0 0 1 0
j 0 -j ' j 0 -j ' 0 -j 0 ' 1 0 1 ' 1 0 -1 ' 0 1 0 ' j 0 -j '
0 1 0 _ _0 1 0
_' 0 -j _j 0 -j ..0 j 01 0-10 -1 0 _
_
0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 1 0 1 1 0 1 1 0 1
1 0 1 j 0 -j 0 -j 0 1 0 1 0 1 0 j 0 -j 1 0 -1 0 1 0
0 j 0 ' 0 1 0 ' 1 0 1 ' 0 -1 0 ' 0 -J 0
' 0 1 0 ' 0 1 0 ' 0 -1 0
1 0 -1 0 -j 0 1 0 -1_ _j 0 -j_1 0 -1 0 1 0 0 -1 0 1 0 -1
_ _ _ _ _ _ _ _
Rank-3 codebook set 4-2
1 0 1 10 1 01 0 0 1 O1 0 1 1 0 1 0 1 0
1 0 -1 0 1 0 0 1 0 1 0 0 1 0 1 0 -1 1 0 1
0 1 0 ' 1 0 -1' 1 0 1 '0 -j 0 '0 1 0'0 1 0'0 -
1 0'
0 -1 0 0 j 0 j 0 -j j 0 -j j 0 -j 0 -j 0 1 0 -1
_
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 1 1 0 1 0 1 0
1 0 -1 0 1 0 1 0 0 j 0 0 1 0 0 1 0 01 0 1 0 1
0 1 0 ' 1 0 -1' 0 -1 0 1 0 1 ' 1 0 -1' 0 -j 0 ' 1 0 -1' j 0 -j
0 1 0 0 1 0 1 0 -1 j 0 -j 0 -1 0 j 0 -j 0 1 0 0 -j 0
_
_
Rank-3 codebook set 5-2
CA 02731210 2011-01-17
WO 2010/018969 71 PCT/KR2009/004468
- - - - - - - - - -
1 0 1 1 0 1 1 0 1 1 0 1 0 1 0 0 1 0 1 0 1 1 0 1
0 1 0 0 1 0 0 1 0 1 0 -1 0 -1 0 1 0 1 1 0 -1 0 1 0
1 0 -1'O -j 0 0 1 0 0 1 0 1 0 1 0 -j 0 0 1 0 0 j 0
0 1 0 1 0 -1 1 0 -1 0 -1 0 1 0 -1 1 0 -1 0 1 0 j 0 -j
- - - - - - - -
1 0 1 1 0 1 0 1 0 0 1 0 0 1 0 1 0 1 1 0
0 1 0 0 1 0 1 0 1 1 0 1 0 -j 0 0 1 0 1 0 -1
0 -1 0 '1 0 -1'1 0 -1'0 j 0 '1 0 1 'j 0 0 1
j 0 -j_ _0 j 0 _ _0 j 0 _1 0 -1_ _j 0 -j_ _0 -1 0 _ _0 -j
Rank-3 codebook set 6-2
- - - - - -
1 0 1 1 0 1 0 1 0 0 1 0 0 1 0 0 1 0 1 0 1
0 1 0 0 1 0 0 -1 0 0-f 0 1 0 1 1 0 1 0 1 0
j 0 -j' 1 0 -1' 1 0 1 1 0 1 0 1 0 j 0 -j' j 0 -j'
0 -1 0 0 1 0 1 0 -1 1 0 -1 1 0 -1 0 j 0 Of 0
_ _ _ _
- - - - - - - -
0 1 0 1 0 1 0 1 101 1 0 1 0 1 0 1 0 1 0 1 0
Of 0 0 1 0 1 0 j 0 -j 0 1 0 1 0 1 1 0 -1 1 0 1
l' j 0'0 j 0'01 0'01 O'j 0 -j'0 1 O'j 0 -j
1 0 -1 0 -1 1 0 -1 0 1 0 1 0 -1 0 1 0 Of 0 0 -1 0
- - - - - - - -
5
Rank-3 codebook set 7-2
0 1 0 -1 0 1- -0 1 0 1 0 1 1 0 1 -1 0
1 -1 0 1- 1 0 1
0 -1 0 0 1 0 1 0 1 0 1 0 1 0 -1 j 0 -j 1 0 -1 0 1 0
1 0 1'10 -1'j 0 -j'0 j 0'01 0'0 1 0'0 1 0'01 0'
j 0 -j _0 1 0 0 1 0 1 0 -1 _0 j 0 0-f 0 0-10 _j 0 -j_
_ _
- - - - - -
0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 1 0 1 1 0 1
1 0 1 101 0-f 0 1 0 1 1 0 1 0 1 0 0 1 0
Of 0'01 0'1 0 l'O -1 O'j 0 -j'0-j O'j 0 -j
_j 0 -j_ _= 0 -j j 0 -j1 0 -1O -1 0 _j 0 -j _0 -j 0_
Rank-3 codebook set 8-2
CA 02731210 2011-01-17
WO 2010/018969 72
PCT/KR2009/004468
0 1 0 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 1 0 1 1 0 1
0 1 0 1 0 1 1 0 -1 0 1 0 0 j 0 0 1 0 0 1 0 j 0 -j
1 0 1 'j 0 -.CO 1 O'j 0 -j' 1 0 1 -1 0
1 0 -1' 0 1 0
j 0 -j 0 -1 0 0 -j 0 0 1 0 j
-j 1 0 -1 0 -j 0 0J0
0 1 0 0 1 0 1 0 1 1 0 1 1 0 1 0 1 1 0 1
1 0 1 1 0 1 j 0 -j 0 1 0 0 1 0 1 0 0 1 0
1 0 -1 0 -j 1 1 0 -1' -j 1 0 -1' 0 j
0j0j0-j0-1 0 0 -1 0 ,1 0 -10j0j0 -j
[00241] It should be noted that the row permutation and/or
the column permutation may be carried out on all or some of
the above precoding matrices shown in Table 4.
[00242] Rank 3 - Second Embodiment
[00243] In one embodiment of the present invention, a
method for constructing a codebook using 6 precoding matrix
groups capable of maintaining good CM properties will
hereinafter be described. The six 4Tx Rank-3 precoding
matrix groups for maintaining good CM properties can be
represented by the following equation 45.
[00244] [Equation 45]
-1 0 0 0 1 0 0 0 1 1 0 0
Group 1 0 1 0 , Group 1 0 0 , Group 3 o 1 0 , Group 4 0 o 1
0 0 1 0 0 1 1 0 0 0 1 0
X 0 0 X 0 0 X 0 0 X 0 0
1 0 0 -1 0 0
Group 5 x 0 0 , Group 6 0 1 0
0 0 1 X 0 0
0 1 0 0 0 1
1+j 1-j -1-j -11
A' _________________
where E{1, j " j
' " -5
CA 02731210 2011-01-17
WO 2010/018969 73
PCT/KR2009/004468
[00245] An example of the Rank-3 codebook including 24
precoding matrixes from 6 groups shown in Equation 45 is
shown in the following table 5.
In order to reduce
complexity, in the example shown in Table 5, letters denoted
by precoding matrix elements are restricted to 1, -j, -1, and
-j.
[00246] [Table 5]
1 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 1 0
0 1 0 0 1 0 0 1 0 0 1 0 1 0 0 1 0 0 1 0 0 1 0 0
0 0 l' 0 0 l'O 0 l' 0 0 1 0 0 l' 0 0 l'O 0 l' 0 0 1
_1 0 0 -1 0 Of _j 0 0 -j 0 0 1 0 0 -1 0 Of j 0 0 _-j 0 0
0 0 1 0 0 1 0 0 1 0 0 1 1 0 0 1 0 0 1 0 0 1 0 0
0 1 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 1
1 0 0' 1 0 0'1 0 0' 1 0 0 0 1 0' 0 1 0'0 1 0' 0 1 0
1 0 0 -1 0 0 j 0 0 -j 0 0 1 0 0 -1 0 0 j 0 0 -j 0 0
- - -- - - - - - - - - -
1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0
1 0 0 -1 0 0 j 0 0 -j 0 0 0 1 0 0 1 0 0 1 0 0 1 0
0 0 l' 0 0 l'O 0 l' 0 0 1 1 0 0'-1 0 O'j 0 0'-j 0 0
0 1 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 1
[00247] For another example, the present invention proposes
a method for utilizing the remaining groups other than a
fourth group (Group 4) generated by applying column
permutation to a first group (Group 1) among all groups shown
in Equation 45.
Generally, if three column vectors are
represented by [cl, c2, c3], 5 column permutation matrices
such as [cl, c3, c2], [c2, cl, c3], [c2, c3, cl], [c3, c2,
cl], and [c3, cl, c2] can be generated, thus 6 matrices can
be achieved.
CA 02731210 2011-01-17
WO 2010/018969 74
PCT/KR2009/004468
[00248] The reason why the specific vector permutation
matrix is not used as described above is that an encoded
sequence is mapped to a specific column vector (or a specific
layer) of the precoding matrix.
It is assumed that two
independently-encoded codewords in the above-mentioned
precoding matrix groups are mapped to different layers as
described below.
[00249] (1) A first codeword is mapped to a first layer.
[00250] (2) A second codeword is evenly distributed and
mapped to second and third layers.
[00251] On the assumption that the above codeword-layer
mapping is used, a specific column permutation does not
generate a difference in average SINR between different
codewords. For example, permutation from a column vector [cl,
c2, c3] to another column vector [cl, c3, c2] may indicate
that only a layer of a second codeword is swapped. In this
way, the swapping between two layers to which the same second
codeword is evenly distributed and mapped does not cause a
variation in performance. For systems utilizing SIC receivers,
correct decoding of a codeword given transmission of
plurality of codewords leads to performance enhancements.
This is because once a codeword is correctly decoded. So the
correctly decoded codeword information can be used to cancel
out spatial layer interference. In the case that transmission
power of multiple antennas is uniformly normalized, some
CA 02731210 2011-01-17
WO 2010/018969 75
PCT/KR2009/004468
column vectors of the precoding matrix may have larger
transmission power. In the case there is no layer
shifting/permutation between all transmission layers, a
specific layer corresponding to the column vector of the
precoding matrix which column vectors has larger transmission
power may have better performance. In case there is no layer
shifting/permutation across all transmitted layers, in order
to fully utilized SIC receivers the first layer, which the
first codeword is solely mapped to is mapped to the precoding
matrix column vector which has larger transmission power, and
the second codeword which is mapped to second and third layer
is mapped to precoding vector columns which has relatively
smaller transmission power compared to the first layer. In
the case above codeword-layer mapping is used, Precoding
matrices as shown in [Equation 46] can be used to further
enhance performance in case Successive Interference
Cancellation (SIC) receiver algorithm is used.
[00252] [Equation 46]
1 0 0 0 1 0 0 0 1
0 1 0 1 0 0 0 1 0
Group 1 , Group 2 , Group 3 , Group 4
0 0 1 0 0 1 1 0 0
X 0 0 X 0 0 X 0 0
1 0 0- 1 0 0-
X 0 0
, Group 5 0 1 0
0 0 1 X 0 0
0 1 0 0 0 1
CA 02731210 2011-01-17
WO 2010/018969 76
PCT/KR2009/004468
1+j 1-j -1-j -11
AT ___
where E{1, j __ ,-1, __ j
V2' 'V2
[00253] The following codebooks are exemplary 4Tx Rank-3
codebooks, each of which restricts letters contained in each
of the above precoding matrices groups to 1, j, -1, and -j,
and includes 20 precoding matrices.
[00254] [Table 6]
1 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 1 0
0 1 0 0 1 0 0 1 0 0 1 0 1 0 0 1 0 0 1 0 0 1 0 0
0 0 l' 0 0 l'O 0 l' 0 0 1 0 0 l' 0 0 l'O 0 l' 0 0 1
1 0 0 -1 0 0 j 0 0 -j 0 0 1 0 0 -1 0 0 j 0 0 -j 0 0
-0 0 1 0 0 1 0 0 1 0 0 1 1 0 0 1 0 0 1 0 0 1 0 0-
0 1 0 0 1 0 0 1 0 0 1 0 1 0 0 -1 0 0 j 0 0 -j 0 0
1 0 0' 1 0 0'1 0 0' 1 0 0 0 0 l' 0 0 l'O 0 l' 0 0 1
1 0 0 -1 0 0 j 0 0 -j 0 0 0 1 0 0 1 0 0 1 0 0 1 0
- - - - - -
1 0 0 1 0 0 1 0 0 1 0 0
010010010010
1 o o'¨1 o o'i o o'¨j o o
ool ool ool o ol
[00255] In the meantime, in accordance with another
embodiment of the present invention, the number of precoding
matrices required for acquiring optimum performance from a
high rank is less than the number of precoding matrices
required for acquiring optimum performance from a low rank,
so that the present invention can restrict the Rank-3
codebook to have below 24 precoding matrices. In this case,
CA 02731210 2011-01-17
WO 2010/018969 77
PCT/KR2009/004468
the present invention may evenly select the precoding
matrices from 6 precoding matrix groups using the Norm 2.
[00256] [Table 7]
1 0 0 1 0 0 1 0 0 1 0 0
0 1 0 0 1 0 0 1 0 0 1 0
0 0 1 0 0 1 0 0 1 0 0 1
1 0 0 -1 0 0 j 0 0 -j 0 0
_
1 0 0 1 0 0 1 0 0 1 0 0
0 1 0 0 1 0 0 1 0 0 1 0
0 0 1 0 0 1 0 0 1 0 0 1
0 1 0 0 -1 0 0 j 0 0 -j 0
1 0 0 1 0 0 1 0 0 1 0 0
0 1 0 0 1 0 0 1 0 0 1 0
0 0 1 0 0 1 0 0 1 0 0 1
0 0 1 0 0 -1 0 0 j 0 0 -j
j-
[00257] As can be seen from the example of Table 7, if e
is multiplied by a specific column vector, column permutation
in a precoding matrix has no influence upon improvement of
performance, so that the number of precoding matrices
contained in a codebook is limited to 12.
Meanwhile, in
accordance with one embodiment of the present invention,
antenna permutation may be carried out to obtain antenna
selection gain.
This antenna permutation may also be
implemented by row permutation of a precoding matrix
contained in the above-mentioned codebook.
[00258] Rank 3 - Third Embodiment
CA 02731210 2011-01-17
WO 2010/018969 78
PCT/KR2009/004468
[00259] In the third embodiment of the present invention,
it is assumed that the following 6 precoding matrix groups
are considered as precoding matrices capable of maintaining
good CM performance.
[00260] [Equation 47]
1 0 0 0 1 0 0 0 1
Group
0 1 0 1 0 0 0 1 0
1
0 0 1 0 0 1 1 0 0
X 0 0 0 X 0 0 0 X
0 1 0 1 0 0 0 1 0
Group 2 1 0 0 0 1 0 0 0 1
0 0 1 0 0 1 1 0 0
X 0 00 X 0 0 0 X
_ _
- 0 0 1 0 0 1 1 0 0 0 1 0
0 0 1
Group 3
0 1 0 1 0 0 0 1 0 Group 4 0 0 1 1 0 0
1 0 0 0 1 0 0 0 1 1 0 0 0
1 0
X 0 0 0 X 0 0 0 X 0 X 0 0
0 X
- 1 0 0 0 1 0 0 0 1
Group 5 X 0 0 0 X 0 0 0 X
0 0 1 0 0 1 1 0 0
0 1 0 1 0 0 0 1 0
- 1 0 0 0 1 0 0 0 1
Group 60 1 0 1 0 0 0 1 0
X 0 00X0 0 0 X
0 0 1 0 0 1 1 0 0
1+] 1-f -1-i
_______________________________________ if __
where X E {1, 3., 1,
[00261] In case of a first group (Group 1) in Equation 47,
it can be recognized that three permutation matrices are
CA 02731210 2011-01-17
WO 2010/018969 79
PCT/KR2009/004468
selected from [c1, c3, c2], [c2, cl, c3], [c2, c3, cl], [c3,
c2, cl], and [c3, cl, c2]. In case of a fourth group (Group
4), it can be recognized that one constituent precoding
matrix is excluded, because the excluded precoding matrix has
already been included in a first group (Group 1). It
is
preferable that the third embodiment be utilized when a layer
shift operation is not carried out. The third embodiment can
implement a layer shift using a codebook including a
precoding matrix set upon which column permutation is carried
out.
Thus, an information sequence may be mapped to all
layers, so that an SINR difference between layers can be
normalized.
[00262] The third embodiment can select a precoding matrix
using the first norm (Norm 1) and the second norm (Norm 2).
[00263] Rank 3 ¨ Fourth Embodiment
[00264] The fourth embodiment considers the following three
groups as precoding matrix groups for maintaining good CM
properties.
[00265] [Equation 48]
1 0 a -1 0 a
1 0
G1 =
X 0 b , G2 = G3 =
0 1 b' 0 1 b"
,
0 1 c X 0 c' 0 Y c"
0 Y d 0 Y d' X 0 d"
1+j 1-j -1-j ;-1+1
/- __ 9 )
where A',17Ã{1,771-' -5,
1/2
CA 02731210 2011-01-17
WO 2010/018969 80
PCT/KR2009/004468
_ada
[00266] The last vectors b
b"among precoding matrix
C C C
d d' d"
groups shown in Equation 48 may be different precoding
matrices such as DFT - based precoding vectors/matrices or
household - based precoding vectors/matrices.
For example,
an example of the last vectors may be a Rank-1 codebook of
the 3GPP LTE system (Release 8 system). Preferably, in order
to maintain orthogonal/partial unitary characteristics of the
1 0 a
I-]
LX b] [
LY di
matrix X 0 b , the matrices 1 a 1 c and
must satisfy
0 1 c
0 Y d
unitary characteristics.
Similarly, matrices [1 a]and
X c'
1 0 a'
[1 1 0 1 b' [1 a]
[1 bl
of the matrix and matrices and
LY d' X 0 c' X d" Y
0 Y d'
1 0 a" 0 1 b"-
of the matrix must satisfy unitary characteristics.
0 Y c"
X 0 d"
This means that parameters must satisfy the following
relationship.
[00267] [Equation 49]
In Group 1: a = 1 , b = -X , and c =-d =Y*
In Group 2: a' = 1 , b' = -X , and c' =-d' =Y*
CA 02731210 2011-01-17
WO 2010/018969 81
PCT/KR2009/004468
In Group 3: a" =1 , b" = ¨X , and c"=¨dn=Y*
[00268]
In this case, although a certain complex constant
is multiplied by each column vector of a specific precoding
matrix, this means that the multiplied results indicate the
same precoding matrix, so that it is assumed that a, a' , or
a is set to 1.
[00269] Preferably, the fourth embodiment may be applied to
a case when layer permutation is executed.
The layer
permutation operation indicates that a specific information
sequence is cyclically mapped and transmitted to all layers
so that SINR performance differences of individual layers are
normalized. If the same power is used in different layers, a
data sequence of the last layer corresponding to the last
column having no value of 0 has the highest power from the
viewpoint of a precoding output signal.
[00270] In case layer permutation is not used and enhanced
SIC receiver algorithm is used, layer which the first
codeword is mapped to should preferably correspond to the
precoding vector column which the transmission power is
relatively larger than other precoding vector columns. In
case of [Equation 48] the third column may have larger
transmission power than others. For cases which the first
column is mapped to the first layer, the second column is
mapped to the second layer, and the third column is mapped to
the third layer, [Equation 48a] may be used instead of
CA 02731210 2011-01-17
WO 2010/018969 82
PCT/KR2009/004468
[Equation 48]. This precoding matrix structure will allow
enhanced performance in case no layer permutation is used and
SIC receiver is used, due to the increased correct decoding
probability of a entire codeword given a plurality of
codeword transmission.
[00271] [Equation 48a]
_
a 0 I -a' 0 1-
an 0 1_
G1 = G2 = G3 =
b 0 X b' 1 0 b" 1 0
, ,
c 1 0 c' 0 X c" 0 X
d Y 0 d' Y 0 d" Y 0
where X,Ye Li+j j 1-i 1,-1-j j -1+]
Vi' Vi' 1)2
[00272] Rank 3 - Fifth Embodiment
[00273] In the fifth embodiment, it is assumed that the
following groups shown in Equation 50 are used as precoding
matrix groups for maintaining good CM performance.
[00274] [Equation 50]
4-1 0a01aa0 1-\ r-1 0 a' 0 1 a' 0 1-\
G1 =
X Ob0 X bb0 X G2 = 0 1 b' 1 0 b' b' 1 0
0 1 c 1 0 c c 1 0 X 0 c' 0 X c' c' 0 X
0 Y dY 0 ddY 0 0 Y d' Y 0 d' d' Y 0
-1
1 0 a"-- 0 1 a" a" 0 1
0 1 b" 1
G3=
0 Y c" Y 0 c" c" Y 0
X 0 d" 0 X d" d" 0 X1
1- Li 1-j -1-j .-11
where X,YE{1,
N/2 v2
CA 02731210 2011-01-17
WO 2010/018969 83
PCT/KR2009/004468
[00275] The precoding matrix groups shown in Equation 50
are composed of a plurality of precoding matrices acquired
when row permutation or column permutation is carried out on
the structure of the fourth embodiment. The column vectors
a a a
b b"
in the precoding matrix groups shown in Equation 50
C' c"
d d' d"
may be different precoding matrices such as DFT - based
precoding vectors/matrices or household - based precoding
vectors/matrices. For example, an example of the above
column vectors may be a Rank-1 codebook of the 3GPP LTE
system (Release 8 system).
[00276] Similar to the fourth embodiment, in the fifth
embodiment, it is preferable that precoding matrix vectors be
orthogonal to each other and elements other than a first
value of 0 in all column vectors of each precoding matrix
group be set to 1.
[00277] A codebook according to the fifth embodiment
includes a precoding matrix generated when column permutation
is carried out on the precoding matrices of the fourth
embodiment. As described above, the precoding matrix having
a column vector [cl, c2, c3] may have 6 column permutation
precoding matrices such as [cl, c3, c2], [c2, cl, c3], [c2,
c3, c1], [c3, c2, cl], [c3, cl, c2] and [c3, cl, c2].
CA 02731210 2011-01-17
WO 2010/018969 84
PCT/KR2009/004468
[00278] The reason why a specific column permutation is not
included is that second and third column permutations of the
precoding matrix in a system in which a first codeword is
mapped to a first layer and a second codeword is distributed
and mapped to second and third layers do not cause a
difference in performance.
[00279] Rank 3 - Sixth Embodiment
[00280] A precoding matrix according to the sixth
embodiment is configured in a format acquired when row
permutation is carried out on a precoding matrix of the
codebook shown in the fourth embodiment, because the
precoding matrix of the sixth embodiment can be acquired by
antenna switching.
[00281] The precoding matrix groups according to the sixth
embodiment can be represented by the following equation 51.
[00282] [Equation 51]
1 0aX0b01c0Ydl0al0a10
Gi= X0 b 1 OaX0bX0b01cOY dX0b
0 1 c 0 1 c 1 0 a 0 1cX0b0 1 c 0 Yd
OYd 0 Yd 0 Yd 1 0 a 0 YdX 0 b 0 1 c
-/
(-1 0 d¨O 1 Y¨X 0 e¨O Y 0 d-1 0 d-1 0 a'-'\
G2 = 0 1Y1 0d0 1Y0 1YX0 e 0Yd0 1 b'
X 0 eX0 e 1 0dX0 e 0 1YX0 e OYd
OYd'OYd'OYd1 0dOYd01YX0 e
CA 02731210 2011-01-17
WO 2010/018969 85
PCT/KR2009/004468
(-
1 0 a"¨ 0 1 b"-- 0 Y c"-- X 0 cl"- -1 0 a"¨ 1 0 a"¨ 1 0 a"-
\
G3 = 0 1 b" 1 0 a" 0 1 if 0 1 b" 0 Y c" X 0 d" 0 1 if
0 Y c" 0 Y c" 1 0 a" 0 Y c" 0 1 If 0 Y c" X 0 d"
X 0 d" X 0 d" X 0 d" 1 0 a" X 0 d" 0 1 b" 0 Y c"1
XE1,1+ j ¨1+j
L" -5 ' ' ' -5
[00283] [Equation 51]
a a a
b b"
[00284] The column vectors
or their row
d d' d"
permutation formats may be different precoding matrices such
as DFT - based precoding vectors/matrices or household -
based precoding vectors/matrices. For example, an example of
the above column vectors may be a Rank-1 codebook of the 3GPP
LTE system (Release 8 system).
[00285] Similar to the fourth embodiment, in the sixth
embodiment, it is preferable that column vectors of the
precoding matrix be orthogonal to each other and elements a,
a , or a/I are set to 1. An example of the codebook according
to the sixth embodiment can be represented by the following
equation 52.
[00286] [Equation 52]
CA 02731210 2011-01-17
WO 2010/018969 86
PCT/KR2009/004468
- - - - - -
1 0 1 X 0 - X 0 1 c 0 Y d 1 0 1 1 0 1
XO-X 10 1 X 0-X XO-X 01 c OY d
Gl'=
0 1 c 0 1 c 1 0 1 0 1 c X 0 - X 0 1 c
OY d OY d OY d 10 1 OY d XO-X
1 0 1
X 0 -x
0 Y d
0 1 c
1 0 1 ¨ 0 1 b' X 0 - X- 0 Y - 1 0 1 - - 1 0 1 -
G2'=0 1 b' 1 0 1 0 1 b' 0 1 b' X 0 - X 0 Y
X 0 -X X 0 -X 1 0 1 X 0 -X 0 1 b' X 0 -X
0 Y d' 0 Y d' 0 Y d' 1 0 1 0 Y d' 0 1 b'
1 0 1
0 1 b'
0 Y d'
X 0 -X
1 0 1 - 0 1 b" ¨ 0 Y c" X 0 - X- 1 0 1 ¨ 1 0 1 -
G3'=0 1 b" 1 0 1 0 1 If 0 1 b" 0 Y c" X 0 - X
0 Y c" 0 Y c" 1 0 1 0 Y c" 0 1 1," 0 Y
X 0 -X X 0 -X X 0 -X 1 0 1 X 0 -X 0 1 b"
- 1 0 1
0 1 If
X 0 -x
0 Y c"
where X,YE{1 = 1-j 1 ___ ' = - -fi14-i
-
[00287] Rank 3 - Seventh Embodiment
[00288] A codebook according to the seventh embodiment is
configured in a format of row permutation of the codebook
shown in the fifth embodiment. An example of the codebook
CA 02731210 2011-01-17
WO 2010/018969 87
PCT/KR2009/004468
according to the seventh embodiment can be represented by the
following equation 53.
[00289] [Equation 53]
(-
10a-X Ob 01 c--0Yd-10a-10a-1 0
X0b 10a X0b X0b 0 1 c OY d X0b
0 1 c'0 1 c'l 0 a'0 1 c'XOb'Olc'OYd
0 Yd OYd OYd 10a OYd X0b 01 c
- - - - - - - - - -
- - - - - - - - - - - -
0 1 a 0 X b 1 OcY0dOlaOla01 a
Gl= OXb 01a OXb 0Xb10c Y Od OXb
1 0 c' 1 0 c' 0 1 a' 1 0 c' 0 X b' 1 0 c' Y 0 d
YOd YOd YOd 0 1a Y Od OXb 10c
- - - - - - - - - - -
- - - - - - - - - - - -
a01 bOX c10 dY0 a01 a01 a01
bOX a01 bOX bOX c 1 0 dY 0 bOX
c 1 0 c 1 O'a 0 l'c 1 O'b 0 X'c 1 O'd Y 0
dY0dY0dY0a01dY0b0Xc10
- - - - - -
1 0 a' 0 1 b' X 0 c' 0 Y 1 0
a' 1 0 a' - 1 0 a' \
01 b' 1 0 a' 01 b' 01 b' X 0 c' 0 Y d' 01 b'
X 0 c' X 0 c' 1 0 a' X 0 c' 01 b' X 0 c' 0 Y d'
0 Y d' 0 Y d' 0 Y d' 1 0 a' 0 Y d' 01 b' X 0 c'
01 1 0 b' 0 X c'- Y 0 d' 0 1 a'- 01 a'-- 01 a'-
G2= 10 b' 01 a' 1 0 b' 1 0 b' 0 X c' Y 0 d' 1 0 b1
0 X c' 0 X c' 01 a' 0 X c' 1 0 b' 0 X c' Y 0 d'
Y 0 d' Y 0 d' Y 0 d' 01 a' Y 0 d' 1 0 b' 0 X c'
- a' 01¨b' 1 0 c' 0 X¨ d' Y 0--a' 01-a'01¨a'01
b' 1 0 a' 01 b' 1 0 b' 1 0 c' 0 X d' Y 0 b' 1 0
c' 0 X c' 0 X a' 01 c' 0 X b' 1 0 c' 0 X d' Y 0
Y 0 __ds Y 0 _d' Y __a' 0 1 __d' Y
0 __b' 1 0 __c' 0 X_J
CA 02731210 2011-01-17
WO 2010/018969 88
PCT/KR2009/004468
( 1 0 a" 0 1 b"-0 Y c" X 0 d"¨ 1 0 a"- 1 0 a"- 1 0 a"-.
0 1 b" 1 0 a" 0 1 b" 0 1 b" 0 Y c" X 0 d" 0 1 b"
0 Y c" 0 Y c" 1 0 a" 0 Y c" 0 1 b" 0 Y c" X 0 d"
X 0 d" X 0 d" X 0 d" 1 0 a" __X 0 d" 0 1 b" 0 Y c"
0 1 a" - 1 0 b" Y 0 c"- 0 X d" 0 1 a"¨ 0 1 a" 0 1 a"
G3= 1 0 b" 0 1 a" 1 0 b" 1 0 b" Y 0 c" 0 X d" 1 0 b"
Y 0 c" Y 0 c" 0 1 a" Y 0 c" 1 0 b" Y 0 c" 0 X d"
0 X d" 0 X d" 0 X d" 0 1 a" 0 X d" 1 0 b" Y 0 c"
- a" 0 1- b" 1 0- c" Y 0 d" 0 X a" 0 1 -a" 0 1¨a" 0 1
If 1 0 a" 0 1 If 1 0 b" 1 0 c" Y 0 d" 0 X b" 1 0
c" Y 0 c" Y 0 a" 0 1 c" Y 0 If 1 0 c" Y 0 d" 0 X
d"0Xd"0Xd"0 Xa" 01 d"OXb" 10c"V 0
where X,YE ___________ j 1-j j -1+]
a a a
bb' "
[00290] The column vectors
b or their row
d d' d"
permutation formats may be different precoding matrices such
as DFT - based precoding vectors/matrices or household -
based precoding vectors/matrices. For example, an example of
the above column vectors may be a Rank-1 codebook of the 3GPP
LTE system (Release 8 system) .
[00291] Similar to the fourth embodiment, in the seventh
embodiment, it is preferable that column vectors of the
precoding matrix be orthogonal to each other and elements a,
a, , or at, are set to 1.
It is preferable that the codebook
according to this embodiment be used when antenna permutation
is not carried out, because the antenna permutation effect
can be achieved by the precoding matrix to which row
CA 02731210 2011-01-17
WO 2010/018969 89
PCT/KR2009/004468
permutation is carried out when using the codebook of the
seventh embodiment.
[00292] An example of the codebook according to the seventh
embodiment can be represented by the following equation 54.
[00293] [Equation 54]
Gl=
(-
1 0 1 X 0 -X0 1 c 0 Y d 1 0 1 1 0 1 1 0
1 -\
X 0 -X 1 0 1X0 -XXO-X 0 1c0YdX0 -X
0 101 0 1 0 1cX0 -X 0 1c0Yd
0Yd0Yd0Yd10 1 OYdX0 -X0 1 c
0 1 1 OX-X 1 0cY0d0 1 1 0 1 101 1
OX-X 0 1 1 OX-X OX-X 1 0cY0d0X -X
1 OclOc0 1 1 1 0c0X-X 1 0cY0 d
Y 0 d _LY 0 d 0 di_O 1
1 Y o d_LO X -X1 0 c _
1 0 1-X 0Xc 1 0dY0 1 0 1 1 0 1 1 0 1
-X0X1 0 1 -X0X-X0Xc 1 0 dY0 -XOX
c 1 0 c 1 0 1 0 1 c 1 0 -X0Xc 1 OdY0
dY0dY0dY0 1 0 1dY0 -X 0Xcl0
\_
CA 02731210 2011-01-17
WO 2010/018969 90
PCT/KR2009/004468
G2=
(-1 0 1 ¨0 1 ¨ X 0 -X 0 Y d' 1 0 1 - 1 0 1 -
1 0 1
0 1 b' 1 0 1 0 1 b' 0 1
b' X 0 - X 0 Y d' 0 1 b'
X 0 -X X 0 -X 1 0 1 X 0 -X 0 1 b' X 0 -X 0 Y
0 Y d' 0 Y d' 0 Y d' 1 0 1 0 Y d' 0 1 b' X 0 -X
0 1 1 ¨ 1 0 b' 0 X - X¨ Y 0 d' 0 1 1 - 0 1 1
¨0 1 1 -
Y 0 d' Y 0 d' Y 0 d' 0 1 1 Y 0 d' 1 0 b' 0 X -X
b' 1 0 1 0 1 b' 1 0 b' 1 0 -
X 0 X d' V0 b' 1 0
-X 0 X -X 0 X 1 0 1 -X 0 X b' 1 0 -X 0 X dl Y0
Y 0 Y 0 Y 0
1 0 1 d' Y L b' 1 0 - X 0 X
G3=
(-1 0 1 0 1 b" ¨ 0 Y c" ¨ X 0 - X¨ 1 0 1 ¨ 1 0 1 - 1 0
1
0 1 b" 1 0 1 0 1 b" 0 1 b" 0 Y c" X 0 - X 0 1 b"
0 Y c" 0 Y c" 1 0 1 0 Y c" 0 1 b" 0 Y c" X 0 - X
X 0 -X X 0 -X X 0 -X 1 0 1 X 0 - X 0 1 b" 0 Y
0 1 1 1 0 b" V 0 c" 0 X d"- 0 1 1 - 0 1 1 0 1
1
1 0 b" 0 1 1 1 0 b" 1 0 b" Y 0 c" X - X
1 0 b"
V0 c" Y 0 c" 0 1
1 Y 0 c" 1 0 b" Y 0 c" 0 X - X
0 X -X ;0 X -X0 X -X0 1 1 0 X -Xl 0 b" Y 0 c"
b" 1 0 1
0 1 b" 1 0 b" 1 0 Y 0 - X 0 X b" 1 0
c" Y0 c" V0 1 0 1 c" V0 If 1 0 c" Y 0-X0X
-X 0 X -X 0 X -X 0 X 1 0 1-X 0 X b" 1 0 c" Y 0
_-1
1-
where X,Ye{1 -1-
, ______________________ j 1
V2 V2 V2 1/2
[00294] Reference for Selecting Additional Precoding Matrix
CA 02731210 2011-01-17
WO 2010/018969 91
PCT/KR2009/004468
[00295] In addition to the Norm 1 and the Norm 2, this
embodiment is designed to consider another norm.
In this
norm, elements denoted by letters contained in each precoding
matrix group are not selected from among eight values, but
are limited to 1, j, -1 and -j, thus reducing the number of
precoding matrices contained in a codebook.
[00296] In accordance with this embodiment, a codebook set
including 16 precoding matrices is considered. For example,
Rank 1 DFT vectors about 4Tx antennas can be represented as
follows.
[00297] N x N DFT matrix (or Fourier Matrix) FN based on a
given component, such as FN =e2'
normalized to 1NN can be
represented by the following equation 55.
[00298] [Equation 55]
-1 1 1 1
F2F N -1
. . .
FN= =
"
1 AT -1) F 2.(N-1) F (N -1).(N -1)
= = =
[00299] Rank 1 DFT vectors about the 4Tx antennas composed
of 16 column vectors located at the first four rows of
Equation 55.
[00300] [Table 8]
CA 02731210 2011-01-17
WO 2010/018969 92
PCT/KR2009/004468
_
_
_ _ _ _ _ _
1 1 1 1 1 3 1 1 1
1
1
1 - 1-
9 1 5 II 1
1
3 7
-.1-g -J-1r -f-Jr-j-g --j - Tr -.I-7r -.1-71" -.1-ff
-.1-11. -f¨Jr
1 e s e 4 8 e 4 e 8 e-Er e 8 e 4
e 8
1 1 3 5 3 7 1 5 11
-J-It -.I-7r -.1-1r -I-4g -f-Jr -J-it -f-It -
f-IT -.1---g
1 e 4 e 2 e 4 e--fir e e2 e 1 e 4 e
2 e 4
3 3 9 3 15 9 5 11 7 1
1 _jg -- --Yr -- -fi'T-- -J-fr 7qg -fig _e-' _-1-"
-fig
-f-It
__eie e e e e e-
_e-8e e8 _
_
_
1-
1 1 1
3 13 7 15
-f-IT -J¨ff -f-Jr -.1¨g
e 2e 8 e 4 e 8
5 3 7
-I-4g -J-21r -f-Jr4
e
e-fir e e
I 7 5 13
-J-2g
e-l-8g e-l-4g -Pi- g
e e
_ _
[00301] Next, 4Tx Rank 1 house hold vector (HH vector) may
be represented by the following Table 9.
5 [00302] [Table 9]
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1
j ¨1 1--./ -1-J -1+J 1,-
1 ¨ ./
i li Ji Ali li 1 ¨ j ¨1 j 1
1 ¨1 ¨1
1 ¨1 1 ¨1 ¨j j ¨ j j ¨1 1 ¨1 1 1
¨1 1 ¨1
-1-J 1-j 1+j -1+f1 j ¨1 ¨j ,- ¨1 ¨j 1 j ¨1 1 1 ¨1
[00303] Codebook Size Restriction
[00304] At least one of the first to third norms (Norm 1,
Norm 2 and Norm 3) may be used to limit the number of
precoding matrices contained in a codebook. In this
embodiment, codebook size restriction for each rank,
especially, size restriction in a Rank 1 codebook, will be
described in detail.
[00305] Presently, a downlink 4Tx codebook for the 3GPP LTE
system has prescribed that respective ranks have the same
number of vectors/matrices (i.e., 16 vectors/matrices).
However, it is well known in the art that the number of
precoding matrices required to acquire optimum performance
CA 02731210 2011-01-17
WO 2010/018969 93
PCT/KR2009/004468
from a high rank is less than the number of precoding
matrices required to acquire optimum performance from a low
rank. For this purpose, this embodiment of the present
invention proposes a new codebook format in which the number
of precoding matrices of a low rank is higher than that of a
high rank so that individual ranks have different numbers of
precoding matrices.
[00306] In the meantime, a mobile communication system can
support a variety of transmission modes. It is assumed that
an X-th transmission mode is effectively used for a UE
located at a cell edge so that the UE can support a closed
loop operation using a Rank 1 Precoding Matrix Indicator
(PMI). In this case, a Rank 1 PMI vector may be selected
from the Rank 1 precoding matrices contained in an overall
codebook composed of a plurality of precoding matrices of all
ranks supporting a Y-th transmission mode such as an open
loop MIMO or closed loop MIMO. In this case, it is assumed
that the X-th transmission mode is different from the Y-th
transmission mode. For the Y-th transmission mode, the size
of the Rank 1 codebook need not be configured as a power of 2.
In addition, although the Rank 1 codebook size is configured
as a power of 2, only the codebook size can be increased
without higher performance improvement.
Thus, this
embodiment proposes a method for rationally restricting the
codebook size simultaneously while having appropriate
CA 02731210 2011-01-17
WO 2010/018969 94
PCT/KR2009/004468
performance so that the codebook can be expressed with a
smaller amount of feedback information.
[00307] Firstly, it is assumed that numbers of precoding
matrices of individual ranks supporting the Y-th transmission
mode are set to A- Rank 1, B- Rank 2, C- Rank 3, and D- Rank
4 (where D C B A).
In this case, the size of an
overall codebook is equal to the sum of A, B, C, and D. In
order to support the above codebook size, m-bit signaling for
satisfying the following condition shown in Equation 56 may
be needed.
[00308] [Equation 56]
A+B+C+D 2m
[00309] If a UE is configured to use the X-th transmission
mode, a UE is able to use Rank 1 PMI information.
It is
preferable that 2' Rank 1 PMIs (where n < m) be newly defined
to reduce the number of bits required for signaling.
A
variety of methods (1), (2), (3), (4), (5) and (6) may be
used to reduce the number of signaling bits.
[00310] (1) Method 1
[00311] If possible, an even-th index is selected.
[00312] (2) Method 2
[00313] If possible, an odd-th index is selected.
[00314] (3) Method 3
[00315] Initial 2' indexes are selected.
[00316] (4) Method 4
CA 02731210 2011-01-17
WO 2010/018969 95
PCT/KR2009/004468
[00317] Last 2' indexes are selected.
[00318] (5) Method 5
[00319] Indexes are arbitrarily selected.
[00320] (6) Method 6
[00321] Construction is achieved by signaling.
[00322] For example, for the Y-th transmission mode, 33
precoding matrices may be given for Rank 1, 15 precoding
matrices may be given for Rank 2, 15 precoding matrices may
be given for Rank 3, and 4 precoding matrices may be given
for Rank 4.
[00323] In this case, a variety of methods (1), (2), (3),
(4), (5) and (6) for constructing the Rank 1 codebook used
for indicating only 16 precoding matrices can be used.
[00324] (1) Method 1
[00325] If possible, an even-th index is selected.
[00326] (2) Method 2
[00327] If possible, an odd-th index is selected.
[00328] (3) Method 3
[00329] Initial 16 indexes are selected.
[00330] (4) Method 4
[00331] Last 16 indexes are selected.
[00332] (5) Method 5
[00333] Indexes are arbitrarily selected.
[00334] (6) Method 6
[00335] Construction is achieved by signaling.
CA 02731210 2011-01-17
WO 2010/018969 96
PCT/KR2009/004468
[00336] In the meantime, a variety of methods (1), (2), (3)
and (4) for constructing the Rank 1 codebook used for
indicating only 32 precoding matrices can be used.
[00337] (1) Method 1
[00338] Initial 32 indexes are selected.
[00339] (2) Method 2
[00340] Last 32 indexes are selected.
[00341] (3) Method 3
[00342] Indexes are arbitrarily selected.
[00343] (4) Method 4
[00344] Construction is achieved by signaling.
[00345] If 16 downlink Rank 1 vectors are contained in the
Rank 1 codebook including 32 precoding matrices, the
following restriction methods (I) and (II) can be used.
[00346] The restriction method (I) corresponds to a case
for constructing the 16-sized Rank 1 codebook, and a detailed
description thereof will hereinafter be described in detail.
[00347] A) 16 downlink Rank 1 vectors are selected.
[00348] B) The 16-sized Rank 1 codebook is selected
regardless of downlink Rank 1 vectors.
[00349] (1) Initial 16 indexes are selected.
[00350] (2) Last 16 indexes are selected.
[00351] (3) Indexes are arbitrarily selected.
[00352] (4) Construction is achieved by signaling.
CA 02731210 2011-01-17
WO 2010/018969 97
PCT/KR2009/004468
[00353] The other restriction method (II) corresponds to
another case for constructing the 32-sized Rank 1 codebook,
and a detailed description thereof will hereinafter be
described in detail.
[00354] A) Selection of 16 downlink Rank 1 vectors +
additional vectors.
[00355] (1) Initial 16 indexes are selected.
[00356] (2) Last 16 indexes are selected.
[00357] (3) Indexes are arbitrarily selected.
[00358] (4) Construction is achieved by signaling.
[00359] B) Selection of 32-sized Rank 1 codebook regardless
of downlink Rank 1 vectors.
[00360] (1) Initial 32 indexes are selected.
[00361] (2) Last 32 indexes are selected.
[00362] (3) Indexes are arbitrarily selected.
[00363] (4) Construction is achieved by signaling.
[00364] The number of codebooks for each rank can be
effectively constructed according to the above-mentioned
schemes.
[00365] III. Apparatus Configuration
[00366] Chapter III will hereinafter disclose an improved
structure to be contained in a UE, wherein the improved
structure can maintain good PAPR or CM properties
CA 02731210 2011-01-17
WO 2010/018969 98
PCT/KR2009/004468
simultaneously while applying the MIMO scheme to uplink
signal transmission.
[00367]
FIG. 10 is a block diagram illustrating a general
base station (BS) and a general user equipment (UE).
[00368] Referring to FIG. 10, a base station (BS) 10
includes a processor 11, a memory 12, and a Radio Frequency
(RF) module 13.
The RF module 13 is used as a
transmission/reception module for receiving an uplink signal
and transmitting a downlink signal.
The processor 11 may
control downlink signal transmission using downlink signal
transmission information (for example, a specific precoding
matrix contained in a codebook for downlink signal
transmission) stored in the memory 12.
Otherwise, as an
inverse process of the precoding process, the processor 11
may control a signal reception process by multiplying uplink
signal reception information (e.g., an uplink signal) stored
in the memory 12 by a Hermitian matrix of the same precoding
matrix as a precoding matrix used in the UE 20.
[00369] The UE 20 may include a processor 21, a memory 22,
and an RF module 23 used as a transmission/reception module
for transmitting an uplink signal and receiving a downlink
signal.
The processor 21 may control uplink signal
transmission using uplink signal transmission information
(for example, a specific precoding matrix contained in the
above-mentioned codebook for uplink signal transmission)
CA 02731210 2011-01-17
WO 2010/018969 99
PCT/KR2009/004468
stored in the memory 22. Otherwise, as an inverse process of
the precoding process, the processor 21 may control a signal
reception process by multiplying downlink signal reception
information (e.g., a downlink signal) stored in the memory 22
by a Hermitian matrix of the same precoding matrix as a
precoding matrix used in the UE 20.
[00370] In the meantime, a detailed description about a
processor of the UE 20 (or the BS 10), particularly, a
structure for transmitting a signal using the SC-FDMA scheme,
will hereinafter be described. A processor for transmitting
a signal based on the SC-FDMA scheme in the 3GPP LTE system
and a processor for transmitting a signal based on an OFDM
scheme in the 3GPP LTE system will hereinafter be described,
and a processor for enabling a UE to transmit an uplink
signal using the SC-FDMA scheme as well as the MIMO scheme
will then be described below.
[00371] FIGS. 11 and 12 illustrate an SC-FDMA scheme for
transmitting an uplink signal in the 3GPP LTE system and an
OFDMA scheme for transmitting a downlink signal in the 3GPP
LTE system.
[00372] Referring to FIG. 11, not only a UE for
transmitting an uplink signal but also a base station (BS)
for transmitting a downlink signal includes a Serial-to-
Parallel converter 401, a subcarrier mapper 403, an M-point
IDFT module 404, a Parallel-to-Serial converter 405, and the
CA 02731210 2011-01-17
WO 2010/018969 100
PCT/KR2009/004468
like. However, a UE for transmitting a signal using the SC-
FDMA scheme further includes an N-point DFT module 402, and
compensates for a predetermined part of the IDFT processing
influence of the M-point IDFT module 404 so that a
transmission signal can have single carrier characteristics.
[00373] FIG. 12 shows the relationship between a block
diagram for an uplink signal process prescribed in TS 36.211
including the 3GPP LTE system specification and a processor
for transmitting a signal using the SC-FDMA scheme.
In
accordance with the TS 36.211, each UE scrambles a
transmission signal using a specific scrambling sequence so
as to transmit an uplink signal, and the scrambled signal is
modulated so that complex symbols are generated. After that,
transform precoding for performing a DFT spreading process on
complex symbols is carried out.
That is, a transform
precoder prescribed in TS 36.211 may correspond to an N-point
DFT module. Thereafter, the DFT-spread signal may be mapped
to a specific resource element according to a resource block
(RB) - based mapping rule by a resource element mapper, and
it can be recognized that this operation corresponds to the
subcarrier mapper shown in FIG. 11. The signal mapped to the
resource element is M-point IDFT or IFFT - processed by the
SC-FDMA signal generator, parallel-to-serial conversion is
performed on the IDFT or IFFT processed result, and then a
cyclic prefix (CP) is added to the P/S conversion result.
CA 02731210 2011-01-17
WO 2010/018969 101
PCT/KR2009/004468
[00374]
In the meantime, FIG. 12 further shows a processor
of a base station (BS) that is used to receive a signal which
has been received in the base station through the above-
mentioned processes.
[00375] In this way, the processor for SC-FDMA transmission
in the 3GPP LTE system does not include a structure for
utilizing the MIMO scheme.
Therefore, the BS processor for
MIMO transmission in the 3GPP LTE system will be described
first, and a processor for transmitting an uplink signal by
combining the SC-FDMA scheme with the MIMO scheme using the
above BS processor will then be described.
[00376]
FIG. 13 is a block diagram illustrating a processor
for enabling the base station (BS) to transmit a downlink
signal using the MIMO scheme in the 3GPP LTE system.
[00377] A base station (BS) in the 3GPP LTE system can
transmit one or more codewords via a downlink.
Therefore,
one or more codewords may be processed as complex symbols by
the scrambling module 301 and the modulation mapper 302 in
the same manner as in the uplink operation shown in FIG. 12.
Thereafter, the complex symbols are mapped to a plurality of
layers by the layer mapper 303, and each layer is multiplied
by a predetermined precoding matrix selected according to the
channel status and is then allocated to each transmission
antenna by the precoding module 304.
The processed
transmission signals of individual antennas are mapped to
CA 02731210 2011-01-17
WO 2010/018969 102
PCT/KR2009/004468
time-frequency resource elements to be used for data
transmission by the resource element mapper 305. Thereafter,
the mapped result may be transmitted via each antenna after
passing through the OFDMA signal generator 306.
[00378] However, if a downlink signal scheme shown in FIG.
13 is used in the 3GPP LTE system, PAPR or CM properties may
be degraded. Thus, it is necessary for a UE to effectively
combine the SC-FDMA scheme for maintaining good PAPR and CM
properties described in FIGS. 11 and 12 with the MIMO scheme
shown in FIG. 13, and a UE for performing precoding using the
precoding matrix capable of maintaining good PAPR and CM
properties described in the above embodiment must be
constructed.
[00379] In accordance with one embodiment of the present
invention, it is assumed that a UE for transmitting an uplink
signal via multiple antennas (multi-antenna) includes
multiple antennas (not shown) for transmitting and receiving
signals. Referring to FIG. 10, the UE 20 includes a memory
22 for storing a codebook, and a processor 21 that are
connected to multiple antennas (not shown) and the memory 22
so as to process uplink signal transmission.
In this case,
the codebook stored in the memory 22 includes precoding
matrices established in a manner that a single layer signal
is transmitted to each of the multiple antennas.
The
CA 02731210 2011-01-17
WO 2010/018969 103
PCT/KR2009/004468
processor 21 of the UE configured as described above will
hereinafter be described in detail.
[00380]
FIG. 14 illustrates a processor of the UE according
to one embodiment of the present invention.
[00381] Referring to FIG. 14, the processor of the UE 20
according to one embodiment of the present invention includes
a codeword to layer mapper 1401 for mapping uplink signals to
a predetermined number of layers corresponding to a specific
rank, a predetermined number of DFT modules 1402 for
performing Discrete Fourier Transform (DFT) spreading on each
of the predetermined number of layer signals, and a precoder
1403. The precoder 1403 selects a specific precoding matrix
established in a manner that one layer signal is transmitted
to each antenna 1405 so as to precode a DFT-spread resultant
layer signal received from the DFT module 1402. Particularly,
in this embodiment of the present invention, each DFT module
1402 performs spreading of each layer signal, this DFT module
1402 for spreading each layer signal is located just before
the precoder 1403. When the precoder 1403 performs precoding,
the precoder 1403 is configured such that each layer signal
is mapped to one antenna and then transmitted via the mapped
antenna, so that single carrier characteristics of each layer
signal are maintained and good PAPR and CM properties are
also maintained. In the meantime, the UE 20 further includes
a transmission module.
The transmission module performs a
CA 02731210 2013-10-15
74420-482
104
process (including LEFT 1404 operation) constructing an SC-FDMA
symbol upon the precoded signal, and transmits the resultant
precoded signal to the base station (BS) via multiple antennas
1405.
[00382] In the meantime, the precoder 1403 selects a
precoding matrix to be used for signal transmission from among
the codebook stored in the memory 22, and performs precoding on
the selected precoding matrix. Preferably, these precoding
matrices may be precoding matrices established for equalizing
transmission powers of multiple antennas and/or transmission
powers of respective layers.
[00383] The number of multiple antennas 1405 may be 2 or 4.
The processor of the UE according to one embodiment of the
present invention may further perform not only a layer shift
function for periodically or aperiodically changing a layer
mapped to a specific codeword but also an antenna shift
function for periodically or aperiodically changing an antenna
via which a specific layer signal is transmitted. The layer
shift function may be performed by the layer mapper 1401
separately from the precoding of the precoder 1403, or may also
be performed through column permutation of the precoding matrix
when the precoder 1403 performs precoding. In addition, the
antenna shift function may also be carried out separately from
the precoding of the precoder 1403, or may also be performed
through row permutation of the precoding matrix.
CA 02731210 2011-01-17
WO 2010/018969 105
PCT/KR2009/004468
[00384] The exemplary embodiments described hereinabove are
combinations of elements and features of the present
invention.
The elements or features may be considered
selective unless otherwise mentioned.
Each element or
feature may be practiced without being combined with other
elements or features.
Further, the embodiments of the
present invention may be constructed by combining parts of
the elements and/or features. Operation orders described in
the embodiments of the present invention may be rearranged.
Some constructions or characteristics of any one embodiment
may be included in another embodiment and may be replaced
with corresponding constructions or characteristics of
another embodiment.
It is apparent that the present
invention may be embodied by a combination of claims which
do not have an explicit cited relation in the appended
claims or may include new claims by amendment after
application.
[00385] The embodiments of the present invention may be
achieved by various means, for example, hardware, firmware,
software, or a combination thereof. In
a hardware
configuration, the embodiments of the present invention may
be implemented by one or more application specific integrated
circuits (ASICs), digital signal processors (DSPs), digital
signal processing devices (DSPDs), programmable logic devices
CA 02731210 2013-10-15
74420-482
106
PLDs), field programmable gate arrays (FPGAs), processors,
controllers, microcontrollers, microprocessors, etc.
[00386] In a firmware or software configuration, the
embodiments of the present invention may be achieved by a
module, a procedure, a function, etc. performing the above-
described functions or operations. Software code may be stored
in a memory unit and driven by a processor. The memory unit
may be located at the interior or exterior of the processor and
may transmit data to and receive data from the processor via
various known means.
[00387] It will be apparent to those skilled in the art that
various modifications and variations can be made in the present
invention without departing from the scope of the invention.
Therefore, the above-mentioned detailed description must be
considered only for illustrative purposes instead of
restrictive purposes. The scope of the present invention must
be decided by a rational analysis of claims, and all
modifications within equivalent ranges of the present invention
are contained in the scope of the present invention. It is
apparent that the present invention may be embodied by a
combination of claims which do not have an explicit cited
relation in the appended claims or may include new claims by
amendment after application.
[00388] As apparent from the above description, the present
invention can maintain PAPR or CM properties while transmitting
uplink signals using a MIMO scheme.
[00389] In addition, the present invention uniformly controls
or adjusts antenna/layer transmission power, minimizes an
CA 02731210 2013-10-15
= 7,4420-482
107
amount of signaling overhead required for precoding matrix
information, and acquires a maximum diversity gain.
[00390] The present invention is applicable to a wideband
wireless mobile communication system.
[00391] It will be apparent to those skilled in the art that
various modifications and variations can be made in the present
invention without departing from the scope of the inventions.
Thus, it is intended that the present invention covers the
modifications and variations of this invention provided they
come within the scope of the appended claims and their
equivalents. The scope of the claims should not be limited by
the examples herein, but should be given the broadest
interpretation consistent with the claims as a whole.
