U.S. patent application number 14/960170 was filed with the patent office on 2017-06-08 for systems and methods for calculating log-likelihood ratios in a mimo detector.
This patent application is currently assigned to Uurmi Systems Pvt Ltd. The applicant listed for this patent is Uurmi Systems Pvt Ltd. Invention is credited to Nanda Kishore Chavali, Anvesh Reddy Yalla.
Application Number | 20170163319 14/960170 |
Document ID | / |
Family ID | 58708388 |
Filed Date | 2017-06-08 |
United States Patent
Application |
20170163319 |
Kind Code |
A1 |
Chavali; Nanda Kishore ; et
al. |
June 8, 2017 |
Systems and Methods for Calculating Log-Likelihood Ratios in a Mimo
Detector
Abstract
A method and a communication receiver have been described for
calculating log-likelihood ratios in a communication receiver. The
log-likelihood ratio is calculated for each bit of one or more
subsymbols of each of the one or more spatial streams by computing
effective noise on one or more spatial streams after considering
noise terms resulting from MIMO detection estimates of the
subsymbols on each spatial stream. Finally, signal to noise ratio
is determined for one or more spatial streams from the effective
noise and scaling bit log-likelihood ratios with the signal to
noise ratio.
Inventors: |
Chavali; Nanda Kishore;
(Telangana, IN) ; Yalla; Anvesh Reddy; (Telangana,
IN) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Uurmi Systems Pvt Ltd |
Telangana |
|
IN |
|
|
Assignee: |
Uurmi Systems Pvt Ltd
Telangana
IN
|
Family ID: |
58708388 |
Appl. No.: |
14/960170 |
Filed: |
December 4, 2015 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04L 1/0054 20130101;
H04B 7/0413 20130101; H04B 1/71057 20130101; H04L 25/067
20130101 |
International
Class: |
H04B 7/04 20060101
H04B007/04; H04L 5/00 20060101 H04L005/00 |
Claims
1. A method for decoding multiple input multiple output (MIMO)
signals by calculating log-likelihood ratios in a communication
receiver, the method comprising: receiving, by the communication
receiver, samples of one or more spatial streams; determining one
or more demodulated samples of the received samples by applying one
or more discrete Fourier transforms; determining one or more
subsymbols from the demodulated samples by utilizing a multiple
input multiple output (MIMO) detection method; calculating a
log-likelihood ratio for each bit of one or more subsymbols of each
of the one or more spatial streams by using
.GAMMA..sub.i(k).DELTA.(b.sub.i.sup.l(k)), wherein .GAMMA..sub.i(k)
is an effective signal to noise ratio in i.sup.th spatial stream
and k.sup.th subcarrier, b.sub.i.sup.l(k) is l.sup.th bit of
subsymbol x.sub.i(k) in the k.sup.th subcarrier and the i.sup.th
spatial stream, and .DELTA. indicates calculating the
log-likelihood ratio of b.sub.i.sup.l(k) without considering the
effective signal to noise ratio in the i.sup.th spatial stream and
the k.sup.th subcarrier, and wherein k, l, and i are integers, and
wherein calculating a log-likelihood ratio for each bit further
comprises: computing an effective noise on one or more spatial
streams after considering noise terms resulting from the multiple
input multiple output (MIMO) detection estimates of one or more
subsymbols on one or more spatial streams, wherein the effective
noise of the i.sup.th spatial stream and the k.sup.th subcarrier is
computed using z.sub.i(k)=.SIGMA.c.sub.ij(k){circumflex over
(n)}.sub.j(k) such that c.sub.ij(k) is a coefficient of
contribution corresponding to a contribution of a noise term in a
j.sup.th stream to overall noise in the i.sup.th spatial stream and
{circumflex over (n)}.sub.j(k) is Additive White Gaussian noise in
the j.sup.th stream, and wherein j is an integer; and determining
signal to noise ratio on one or more spatial streams from the
effective noise and scaling bit log-likelihood ratios with the
signal to noise ratio, thereby decoding the multiple input multiple
output (MIMO) signals using the bit log-likelihood ratios.
2. The method as claimed in claim 1, wherein the communication
receiver is a baseband multiple input multiple output
(MIMO)-orthogonal frequency division multiplexing (OFDM)
communication receiver.
3. (canceled)
4. (canceled)
5. The method as claimed in claim 1, wherein the MIMO detection
method is minimum mean squared error (MMSE) method.
6. The method as claimed in claim 5, wherein the coefficient of
contribution c.sub.ij(k) is equal to c ij ( k ) = { 1 r ii ( k ) -
.sigma. q 2 , ii ( k ) ; for i = j 0 ; for i > j - m = i + 1 j (
r im ( k ) - .sigma. q 2 , im ( k ) ) c mj ( k ) r ii ( k ) -
.sigma. q 2 , ii ( k ) ; for i < j ##EQU00021## where
r.sub.ii(k) are the elements of an upper triangular matrix for
i.sup.th spatial stream and k.sup.th subcarrier, .sigma. is a
standard deviation of a complex Additive white Gaussian noise, and
q2,ij(k) is an element of a matrix obtained by performing QR
decomposition on an augmented channel matrix.
7. The method as claimed in claim 1, wherein the MIMO detection
method is a zero forcing method.
8. The method as claimed in claim 7, wherein the coefficient of
contribution is c ij ( k ) = { 1 r ii ( k ) ; for i = j 0 ; for i
> j - m = i + 1 j r im ( k ) c mj ( k ) r ii ( k ) ; for i <
j ##EQU00022## where r.sub.ii(k) are elements of an upper
triangular matrix for the i.sup.th spatial stream and the k.sup.th
subcarrier, obtained by performing QR decomposition of a channel
matrix.
9. A communication receiver comprising a multiple input multiple
output (MIMO) detector for decoding MIMO signals by calculating
log-likelihood ratios, wherein the MIMO detector comprises: a QR
decomposition processor configured to compute QR decomposition of
an augmented multiple input multiple output (MIMO) channel matrix
corresponding to one or more spatial streams on one or more
subcarriers; a Backsolving processor configured to compute MIMO
detection estimates of one or more subsymbols corresponding to the
one or more spatial streams on the one or more subcarriers; a Soft
demapper processor configured to compute bit log-likelihood ratios
of each bit of the one or more subsymbols corresponding to the one
or more spatial streams on the one or more subcarriers, the Soft
demapper processor comprises: a Noise Coefficient Computation
processor configured for calculating an effective noise term
z.sub.i(k) in i.sup.th spatial stream on k.sup.th subcarrier,
wherein z.sub.i(k)=.SIGMA.c.sub.ij(k){circumflex over (n)}.sub.j(k)
such that coefficient c.sub.ij(k) corresponds to a contribution of
a noise term in a j.sup.th stream to overall noise in the i.sup.th
spatial stream on the k.sup.th subcarrier and {circumflex over
(n)}.sub.j(k) is Additive White Gaussian noise in the j.sup.th
stream, and wherein i, j, and k are integers; and an SNR scaling
processor configured for estimating signal to noise ratio from the
effective noise term z.sub.i(k) in the i.sup.th spatial stream and
computing the bit log likelihood ratio for each of the encoded bits
of the subsymbol in the i.sup.th spatial stream on the k.sup.th
subcarrier, wherein the bit log likelihood ratio is a function of
effective signal to noise ratio in the i.sup.th spatial stream
which is estimated after considering the noise terms resulting from
the MIMO detection estimation of the one or more subsymbols
corresponding to one or more spatial streams on one or more
subcarriers, thereby decoding the multiple input multiple output
(MIMO) signals using the bit log-likelihood ratios.
10. A non-transient computer-readable medium comprising program
instructions for causing a programmable processor to decode
multiple input multiple output (MIMO) signals by calculating
log-likelihood ratios in a communication receiver by performing:
receiving, by a communication receiver, samples of one or more
spatial streams; determining one or more demodulated samples of the
received samples by applying one or more discrete Fourier
transforms; determining one or more subsymbols from the demodulated
samples by utilizing a multiple input multiple output (MIMO)
detection method; calculating a log-likelihood ratio for each bit
of one or more subsymbols of each of the one or more spatial
streams by using .GAMMA..sub.i(k).DELTA.(b.sub.i.sup.l(k)), wherein
.GAMMA..sub.i(k) is an effective signal to noise ratio in i.sup.th
spatial stream and k.sup.th subcarrier, b.sub.i.sup.l(k) is
l.sup.th bit of subsymbol x.sub.i(k) in the k.sup.th subcarrier and
the i.sup.th spatial stream, and .DELTA. indicates calculating the
log-likelihood ratio of b.sub.i.sup.l(k) without considering the
effective signal to noise ratio in the i.sup.th spatial stream and
the k.sup.th subcarrier, and wherein k, l, and i are integers, and
wherein calculating a log-likelihood ratio for each bit further
comprises: computing an effective noise on one or more spatial
streams after considering noise terms resulting from the multiple
input multiple output (MIMO) detection estimates of one or more
subsymbols on one or more spatial streams, wherein the effective
noise of the i.sup.th spatial stream and the k.sup.th subcarrier is
computed using z.sub.i(k)=.SIGMA.c.sub.ij(k){circumflex over
(n)}.sub.j(k) such that c.sub.ij(k) is a coefficient of
contribution corresponding to a contribution of a noise term in the
j.sup.th stream to overall noise in the i.sup.th spatial stream and
{circumflex over (n)}.sub.j(k) is Additive White Gaussian noise in
the j.sup.th stream, and wherein j is an integer; and determining
signal to noise ratio on the one or more spatial streams from the
effective noise and scaling bit log-likelihood ratios with the
signal to noise ratio, thereby decoding the multiple input multiple
output (MIMO) signals using the bit log-likelihood ratios.
11. The non-transient computer-readable medium of claim 10, wherein
the non-transient computer-readable medium is a field-programmable
gate array.
Description
FIELD OF THE DISCLOSURE
[0001] The presently disclosed embodiments are generally related to
wireless communication, and more particularly to soft demapping for
log-likelihood ratio (LLR) computation in coded MIMO-OFDM
systems.
BACKGROUND
[0002] The subject matter discussed in the background section
should not be assumed to be prior art merely as a result of its
mention in the background section. Similarly, a problem mentioned
in the background section or associated with the subject matter of
the background section should not be assumed to have been
previously recognized in the prior art. The subject matter in the
background section merely represents different approaches, which in
and of themselves may also correspond to implementations of the
claimed technology.
[0003] Orthogonal Frequency Division Multiplexing (OFDM) is a
modulation technique that is used in many wireless and
telecommunications standards. OFDM is a modulation technique in
which a high rate data stream is divided into many low rate
parallel data streams and each is modulated using a separate
narrowband close-spaced subcarrier thereby making the data stream
less sensitive to frequency selective fading. A multiple-input
multiple-output (MIMO) communication system employs multiple
transmitting antennas and multiple receiving antennas for data
transmission. A MIMO channel formed by the transmitting antennas
and receiving antennas may be decomposed into independent channels.
Each independent channel may also be referred to as a spatial
subchannel of the MIMO channel. The MIMO system provides improved
performance over that of a single-input single-output (SISO)
communication system if the additional dimensionalities created by
the multiple transmit and receive antennas are utilized.
[0004] There has been increased demand for throughput, spectral
efficiency, and improved link reliability in wireless communication
systems. To meet this demand, orthogonal frequency division
multiplexing (OFDM) is combined with multiple input multiple output
(MIMO) signal processing and employed in many wireless standards
including wireless LAN. In such techniques, OFDM is employed as a
multicarrier modulation scheme wherein data is transmitted on many
orthogonal subcarriers to combat the effects of multipath channel,
and MIMO signal processing is used to increase the throughput of
the radio link by increasing the number of transmit and receive
antennas. In a MIMO-OFDM system, modulated signals corresponding to
independent data streams are transmitted from the multiple
transmitting antennas of a transmitter that are then received by
the multiple receiving antennas at the receiver. The receiver upon
receiving the multiple signals, utilizes various MIMO detection
schemes such as maximum likelihood (ML), zero forcing (ZF) and
minimum mean squared error (MMSE) to detect the data symbols from
multiple signals corresponding to each subcarrier after OFDM
demodulation. A forward error correction (FEC) code is employed to
improve the bit error rate performance in a MIMO-OFDM system also
known as a coded MIMO-OFDM system. In coded MIMO-OFDM systems, the
MIMO detector may provide either hard bits (either 1's or -1's) or
soft bits (having the same sign as hard bits, but with magnitude
indicating the reliability of the decision) to the FEC decoder. The
performance of the FEC decoder may be improved by the soft bits,
whose soft information is in the form of log likelihood ratios
(LLR's). Therefore, it is desirable to employ a soft demapper in
the coded MIMO-OFDM system to provide the soft bits to the FEC
decoder.
[0005] A number of methods have been proposed which claim to
provide the above mentioned facilities. In one such method, a soft
MIMO demapper based on ZF and MMSE equalizers is proposed but its
complexity is exponential in number of transmitted streams thus
making its practical implementation difficult. In another such
method, a receiver includes an inner decoder structure having a
soft output M-algorithm (SOMA) based multiple-in multiple-out
(MIMO) joint demapper that uses a SOMA-based MIMO detection process
to perform joint inner demapping over each tone. In yet another
method, a MMSE MIMO detector by using QR decomposition (QRD) of the
augmented channel matrix is presented including a bias removal
technique to decode the data on each stream by permuting the
augmented channel matrix and computing its QRD on each stream.
However, such soft demapping techniques proposed by the existing
methods are cumbersome and are computationally intensive. Hence,
there exists a need for providing a method and a soft demapper
implementing the method for determining Log-Likelihood ratios that
is practically implementable and achieves performance close to the
near optimal-ML receiver even in delay spread channels.
BRIEF SUMMARY
[0006] It will be understood that this disclosure in not limited to
the particular systems, and methodologies described, as there can
be multiple possible embodiments of the present disclosure which
are not expressly illustrated in the present disclosure. It is also
to be understood that the terminology used in the description is
for the purpose of describing the particular versions or
embodiments only, and is not intended to limit the scope of the
present disclosure.
[0007] In an example embodiment, a method for calculating
log-likelihood ratios in a communication receiver is described. The
method comprises receiving, by the communication receiver, samples
of one or more spatial streams. One or more demodulated samples of
the received samples are determined by applying one or more
discrete Fourier transforms. One or more subsymbols are determined
from the demodulated samples by utilizing a MIMO detection method.
A log-likelihood ratio is then determined for each bit of one or
more subsymbols of each of the one or more spatial streams. The
method of calculating log-likelihood ratio for each bit comprises
computing effective noise on one or more spatial streams after
considering noise terms resulting from the MIMO detection estimates
of one of more subsymbols on one or more spatial streams by the one
or more processors. Finally, a signal to noise ratio is determined
by the one or more processors, on one or more spatial streams from
the effective noise and the bit log-likelihood ratios are scaled
with the signal to noise ratio.
[0008] In another example embodiment, a communication receiver
comprising a MIMO detector for calculating log-likelihood ratios is
described. The MIMO detector comprises a QR decomposition unit, a
Backsolving unit, a Soft demapper unit, and an SNR scaling unit.
The QR decomposition unit is configured to compute QR decomposition
of an augmented MIMO channel matrix corresponding to one or more
spatial streams on one or more subcarriers. The Backsolving unit is
configured to compute MIMO detection estimates of one or more
subsymbols corresponding to the one or more spatial streams on the
one or more subcarriers. The Soft demapper unit is configured to
compute bit log-likelihood ratios of each bit of the one or more
subsymbols corresponding to the one or more spatial streams on the
one or more subcarriers. The Soft demapper unit comprises a Noise
Coefficient Computation Unit. The Noise Coefficient Computation
Unit is configured for calculating effective noise term z.sub.i(k)
in the i.sup.th spatial stream on k.sup.th subcarrier, wherein
z i ( k ) = j = i n T c ij ( k ) n ^ j ( k ) ##EQU00001##
such that coefficient c.sub.ij(k) corresponds to the contribution
of the noise term in the j.sup.th stream to the overall noise in
the i.sup.th stream on k.sup.th subcarrier. The SNR scaling unit of
the MIMO detector is configured for estimating signal to noise
ratio from the effective noise term z.sub.i(k) in the i.sup.th
stream and computing the bit log likelihood ratio for each of the
encoded bits of the subsymbol in the i.sup.th spatial stream on
k.sup.th subcarrier, wherein the bit log likelihood ratio is a
function of effective signal to noise ratio in the i.sup.th spatial
stream which is estimated after considering the noise terms
resulting from the MIMO detection estimation of one or more
subsymbols corresponding to one or more spatial streams on one or
more subcarriers.
[0009] Other systems, methods, features and advantages will be, or
will become, apparent to one with skill in the art upon examination
of the following figures and detailed description. It is intended
that all such additional systems, methods, features and advantages
be included within this description, be within the scope of the
embodiments, and be protected by the following claims and be
defined by the following claims. Further aspects and advantages are
discussed below in conjunction with the description.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The accompanying drawings illustrate various embodiments of
systems, methods, and embodiments of various other aspects of the
disclosure. Any person with ordinary skills in the art will
appreciate that the illustrated element boundaries (e.g. boxes,
groups of boxes, or other shapes) in the figures represent one
example of the boundaries. It may be that in some examples one
element may be designed as multiple elements or that multiple
elements may be designed as one element. In some examples, an
element shown as an internal component of one element may be
implemented as an external component in another, and vice versa.
Furthermore, elements may not be drawn to scale. Non-limiting and
non-exhaustive descriptions are described with reference to the
following drawings. The components in the figures are not
necessarily to scale, emphasis instead being placed upon
illustrating principles.
[0011] FIG. 1 illustrates a transmitter in a coded MIMO-OFDM
system, according to an example embodiment.
[0012] FIG. 2 illustrates a transmitter in a coded MIMO-OFDM
system, according to an example embodiment.
[0013] FIG. 3 illustrates a block diagram of a QR Based ZF/MMSE
MIMO detector with proposed soft demapper, according to an example
embodiment.
[0014] FIG. 4 is a flowchart illustrating a method for calculating
log-likelihood ratios, according to an example embodiment.
[0015] FIG. 5 illustrates a PER Performance in TGac channel model A
for 2.times.2 VHT frame with MCS7, according to an example
embodiment.
[0016] FIG. 6 illustrates a PER Performance in TGac channel model D
for 2.times.2 VHT frame with MCS7, according to an example
embodiment.
DETAILED DESCRIPTION
[0017] Some embodiments of this disclosure, illustrating all its
features, will now be discussed in detail. The words "comprising,"
"having," "containing," and "including," and other forms thereof,
are intended to be equivalent in meaning and be open ended in that
an item or items following any one of these words is not meant to
be an exhaustive listing of such item or items, or meant to be
limited to only the listed item or items.
[0018] It must also be noted that as used herein and in the
appended claims, the singular forms "a," "an," and "the" include
plural references unless the context clearly dictates otherwise.
Although any systems and methods similar or equivalent to those
described herein can be used in the practice or testing of
embodiments of the present disclosure, the preferred, systems and
methods are now described.
[0019] Embodiments of the present disclosure will be described more
fully hereinafter with reference to the accompanying drawings in
which like numerals represent like elements throughout the several
figures, and in which example embodiments are shown. Embodiments of
the claims may, however, be embodied in many different forms and
should not be construed as limited to the embodiments set forth
herein. The examples set forth herein are non-limiting examples and
are merely examples among other possible examples.
[0020] The existing techniques for soft demapper require matrix
inversions or multiple iterations of QR decomposition, which is
computationally intensive.
[0021] The method for MIMO soft demapping described herein computes
bit Log-Likelihood Ratios (LLRs) with minimal complexity while
achieving the performance close to that of the Maximum likelihood
(ML) receiver which is ideal in terms of packet error rate (PER)
performance. A general soft demapper in MIMO detector uses the
change in noise variance during the detection process on each
subcarrier to improve the system performance. However, the soft
demapping techniques in MIMO systems are cumbersome and are
computationally intensive.
[0022] According to the disclosed methods and systems the bit LLRs
are computed in a single iteration of QR decomposition and without
any matrix inversion which makes it practically implementable. The
described method while computing bit LLR's for PSK/QAM subsymbols
in a particular stream considers the interference from subsymbols
in other spatial streams. As a result, the performance of the MIMO
detector with the disclosed technique is close to that of the
Maximum likelihood (ML) receiver. Unlike in an ML receiver where
LLRs are calculated jointly for all layers, a per data stream LLR
calculation approach is followed in the MIMO detector with proposed
soft demapper, thereby reducing the complexity. Further, the
proposed technique can be easily extended to any higher order
MIMO-OFDM system with a linear increase in complexity. One skilled
in the art would appreciate the fact that in the proposed systems
and methods the noise variance on each subcarrier is computed by
considering the interference terms from the other spatial streams.
It intelligently combines equalization and soft demapping processes
to eliminate the interference terms effectively and achieves
performance close to the near optimal-ML receiver even in delay
spread channels.
[0023] FIG. 1 illustrates a transmitter block diagram with n.sub.T
antennas in coded MIMO-OFDM system of a disclosed apparatus,
according to an example embodiment. The transmitter 100 comprises
of n.sub.T antennas (122n.sub.1 to 122n.sub.T) in coded MIMO-OFDM.
An input bit sequence {tilde over (d)}(n) is initially fed to a
scrambler 102 which is an encrypting device wherein the data is
mixed up in a digital stream of information. The scrambler 102
transposes or inverts the data stream to make the data stream
unintelligible at a receiver that is not equipped with an
appropriately set descrambling unit. The scrambler 102 also
prevents long sequences of zeros or ones.
[0024] The scrambled data stream from the scrambler 102 is then
sent to a convolutional encoder 106 which is a type of
error-correcting coding for generation of parity symbols via a
sliding application of a Boolean polynomial function to the data
stream. In an aspect, the encoder 106 may be an FEC encoder that
encodes the data stream to enable error correction.
[0025] The encoded data stream is then sent to a stream parser 108
that parses the scrambled data stream into n.sub.T spatial streams
before interleaving, which is a process or methodology to make a
system more efficient, fast and reliable by arranging data in a non
contiguous manner. The stream parser 108 divides the output of the
convolutional encoder 106 into multiple spatial streams that are
then individually sent to different interleavers (110n.sub.1 to
110n.sub.T) and mappers (112n.sub.1 to 112n.sub.T).
[0026] Each interleaver (110n.sub.1 to 110n.sub.T) permutes the
ordering of each spatial stream in a deterministic manner. The
subsequent interleaved bits from each interleaver (110n.sub.1 to
110n.sub.T) of the i.sup.th spatial stream are mapped to subsymbols
in the constellation set S (having size M) using complex PSK/QAM
constellation by a mapper (112n.sub.1 to 112n.sub.T). The mapper
(112n.sub.1 to 112n.sub.T) may map a group of log.sub.2(M) bits
onto a point in complex constellation. The mapping may be done
using some complex constellation, such as binary phase-shift keying
(BPSK), quadrature phase-shift keying (QPSK), 8 phase-shift keying
(8PSK), quadrature amplitude modulation (QAM), etc. The mapper
(112n.sub.1 to 112n.sub.T) may output N such symbols, each symbol
stream corresponding to one of the N orthogonal subcarriers of.
These N parallel symbols are represented in the frequency domain
and may be converted into time domain samples by an IFFT component
(118n.sub.1 to 118n.sub.T).
[0027] The preamble generation unit (114) generates a preamble
sequence prior to the transmission of data symbols that is added to
each spatial stream. The preamble sequence may consist of LSTF's
(legacy short training fields) and LLTF's (legacy long training
fields) by which the timing and frequency synchronizations are
facilitated. Additionally, the preamble sequence may contain signal
fields that transmit the PHY layer parameters such as modulation
and coding scheme, number of spatial streams, payload length, etc.
After transmission of the preamble, the data portion may be
transmitted.
[0028] The subsymbols are then mapped across data subcarriers of
OFDM symbol via multiplexer (116n.sub.1 to 116n.sub.T). OFDM
modulation is performed and time domain transmit signal is obtained
by using IFFT which is performed in an IFFT unit (118n.sub.1 to
118n.sub.T). A cyclic prefix (CP) of length L is prepended to the
time domain signal in a CP prepending unit (120n.sub.1 to
120n.sub.T) and transmitted on i.sup.th transmit antenna
(122n.sub.1 to 122n.sub.T) before analog processing. This signal
with all modified characteristics and modulation schemes is thus
transmitted wirelessly by transmitter and is receipted by a
receiver 200.
[0029] FIG. 2 illustrates a receiver 200 with antennas (202n.sub.1
to 202n.sub.R) in coded MIMO-OFDM system. The transmitted signals
from the transmitter are received wirelessly by receiver antenna
modules (202n.sub.1 to 202n.sub.R). Receiver antenna modules
(202n.sub.1 to 202n.sub.R) may down convert the received RF signal
to baseband and perform analog to digital conversion to provide
digital samples. It is followed by a receiver filter (204n.sub.1 to
204n.sub.R) to remove out of band thermal noise.
[0030] The time and frequency synchronization module 206 may
perform coarse time and frequency synchronizations during LSTF's
and fine time and frequency synchronizations during LLTF's. The
estimated frequency offset on each stream is then corrected by a
rotor module (208n.sub.1 to 208n.sub.R) while the timing offset is
corrected within the Remove GI (guard interval) module (210n.sub.1
to 210n.sub.R). Cyclic prefix which is prefixing of a symbol with a
repetition of the end for each OFDM symbol is also removed in the
Remove GI module (210n.sub.1 to 210n.sub.R). Then OFDM demodulation
is performed by a FFT module (212n.sub.1 to 212n.sub.R), to obtain
frequency domain subsymbols.
[0031] During LLTF's, the output of FFT module is used by the
channel and noise variance estimation module 214 to estimate legacy
channel and compute noise variance. This legacy channel estimate is
used to decode the signal fields (L-SIG and VHT-SIG A) which may
provide parameters required to decode data at the receiver. VHT
PPDU (Very High Throughput-Physical Layer Protocol Data Units) may
also contain VHT-LTF's for estimating MIMO channel in the frequency
domain, using which MIMO detector module 216, may estimate
transmitted data subsymbols.
[0032] If length of a cyclic prefix is larger than channel delay
spread, then timing and frequency synchronization may be performed
for calculating a received vector at output. Received vector at the
output of the FFT block in k.sup.th subcarrier may be given by
equation(1):
y(k)=H(k)x(k)+n(k); k=0,1,2 . . . N-1, (1)
wherein; y(k)=[y.sub.1(k),y.sub.2(k), . . . ,
y.sub.n.sub.R(k)].sup.T is a column vector of dimension
n.sub.R.times.1, x(k)=[x.sub.1(k),x.sub.2(k), . . . ,
x.sub.n.sub.T(k)].sup.T is a n.sub.T.times.1 vector of QAM
subsymbols transmitted in the k.sup.th subcarrier,
n(k)=[n.sub.1(k),n.sub.2(k), . . . ,n.sub.n.sub.R(k)].sup.T is a
n.sub.R.times.1 complex additive white Gaussian noise (AWGN) vector
with covariance matrix R.sub.nn=.sigma..sup.2I, and N is the total
number of data subcarriers. H(k) is a n.sub.R.times.n.sub.T MIMO
channel frequency response matrix in the k.sup.th sub-carrier and
may be given by
H ( k ) = [ h 11 ( k ) h 12 ( k ) h 1 n T ( k ) h 21 ( k ) h 22 ( k
) h 2 n T ( k ) h n R 1 ( k ) h n R 2 ( k ) h n R n T ( k ) ]
##EQU00002##
Elements of H(k) may be independent and identically distributed
(i.i.d) zero mean circularly symmetric complex Gaussian.
[0033] Now received vector y(k) and channel matrix H(k) may be used
to estimate transmitted subsymbols and bit LLR's for each of these
subsymbols may be computed by MIMO detector 216. The computed bit
LLR's are then passed through a deinterleaver (218n.sub.1 to
218n.sub.R). The deinterleaver (218n.sub.1 to 218n.sub.R) at the
receiver applies the inverse permutation to restore the sequence of
transmitted symbols to its original order. Then the data stream is
deparsed by the stream deparser 220 before giving to a Viterbi
decoder 222, which is a soft decision decoder. The decoded bits
from the soft decision decoder are finally passed through a
descrambler 224.
[0034] In an aspect, the MIMO detector 216 may perform either
Maximum Likelihood (ML) or zero forcing (ZF) or minimum mean
squared error (MMSE) detection. ML detector is optimal in terms of
PER performance but its complexity may grow exponentially with the
number of transmitted streams. Sub-optimal detectors employing
ZF/MMSE criteria have been proposed in the current state of art to
achieve near ML performance at reduced complexity, but these
detectors may require computation of matrix inversion. Direct
computation of matrix inversion may be numerically unstable, hence
alternate method such as QR decomposition (QRD) is proposed in
current disclosure. In QRD, channel matrix H(k) is decomposed into
a product of unitary matrix Q(k) and an upper triangular matrix
R(k). Using an upper triangular property of R(k) matrix, transmit
subsymbols on all the spatial streams may be estimated. The soft
demapper may then compute LLR for each coded bit b.sub.i.sup.l(k)
(where l=1,2 . . . log.sub.2M) of subsymbol x.sub.i(k) in the
i.sup.th spatial stream. The MIMO detector 216 is further described
in FIG. 3.
[0035] In an example embodiment, the transmitter 100 and the
receiver 200 may include one or more processors for executing
instructions stored in a memory. The processor may also be
configured to decode and execute any instructions received from
another transmitter or receiver. The processor may include one or
more general purpose processors (e.g., INTEL microprocessors)
and/or one or more special purpose processors (e.g., digital signal
processors). The processor may be configured to execute
computer-readable program instructions, such as program
instructions to carry out any of the functions described in this
description. In an aspect, the one or more processors may be a
field-programmable gate array (FPGA). The memory may include a
computer readable medium. A computer readable medium may include
volatile and/or non-volatile storage components, such as optical,
magnetic, organic or other memory or disc storage, which may be
integrated in whole or in part with the processor. Alternatively,
the entire computer readable medium may be remote from the
processor and coupled to the processor by connection mechanism
and/or network cable. Memory may be enabled to store various types
of data. For instance, memory may store one or more identifiers
related to the transmitter and the receiver and data received from
the transmitter and the receiver and computer-readable program
instructions executable the by processor.
[0036] FIG. 3 illustrates a block diagram of a QR Based ZF/MMSE
MIMO detector with proposed soft demapper, according to an example
embodiment. The MIMO detector includes basic working units such as
QRDU 302, BSU 304, and proposed soft demapper 306 that may include
NCCU 308, and SSU 310, to compute bit LLR. All mentioned units are
discussed later in this disclosure. MMSE Flag may differentiate
mode of detection used in detector. MMSE version of a MIMO detector
is generally termed as QR-MMSE MIMO detector and ZF version as
QR-ZF MIMO detector.
[0037] QR decomposition unit (QRDU) 302, may perform QR
decomposition (QRD) of channel matrix H(k) at each subcarrier into
Q(k) and R(k) matrices. The QRD of H(k) may be given by
equation(2):
H(k)=Q(k)R(k) (2)
For decomposing a channel matrix, any of the existing methods for
QR decomposition such as Householder transformations or Givens
rotation may be used. If MMSE mode is indicated by MMSE flag to
QRDU 302, channel augmentation may be additionally performed before
computing QRD of an augmented matrix. In an aspect, augmented
channel matrix H.sub.aug(k) may be constructed such that a weight
matrix for ZF equalization using an augmented channel H.sub.aug(k)
is equivalent to that of MMSE equalization for a channel H(k).
H aug ( k ) = [ H ( k ) .sigma. I n T ] = [ Q 1 ( k ) Q 2 ( k ) ] R
( k ) ( 3 ) ##EQU00003##
[0038] Again referring to FIG. 3, Back Solving Unit (BSU) 304, may
compute ZF/MMSE estimates of subsymbol vector x(k) from the
received vectory(k) using Q(k) and R(k) matrices. If MMSE mode is
indicated by MMSE flag, BSU 304 may compute MMSE estimate else it
may compute a ZF estimate. Upper triangular property of a R(k)
matrix may be used by BSU 304, to estimate the transmit subsymbols
one by one, beginning with last spatial stream and ending in first
spatial stream. While estimating a transmit subsymbol x.sub.i(k) in
i.sup.th spatial stream, interference of subsymbols
x.sub.i+1(k),x.sub.i+2(k) . . . x.sub.n.sub.T(k) may be removed
using their estimates. The procedure to compute ZF and MMSE
estimate of subsymbols is detailed in following example
embodiments.
[0039] To compute ZF and MMSE estimate of subsymbols for BSU 304,
in QR-ZF MIMO detector with MMSE Flag=0; equation (2) may be used
in MIMO system model given by equation (1) to get equation (4):
y(k)=Q(k)R(k)x(k)+n(k) (5)
Equation (5) may be pre-multiplied by Q.sup..dagger.(k) to get
equation (6):
y ^ ( k ) = R ( k ) x ( k ) + n ^ ( k ) [ y ^ 1 ( k ) y ^ 2 ( k ) y
^ n T ( k ) ] = [ r 11 ( k ) r 12 ( k ) r 1 n T ( k ) 0 r 22 ( k )
r 2 n T ( k ) 0 0 0 r n T n T ( k ) ] [ x 1 ( k ) x 2 ( k ) x n T (
k ) ] + [ n ^ 1 ( k ) n ^ 2 ( k ) n ^ n T ( k ) ] ( 6 )
##EQU00004##
wherein, y(k)=Q.sup..dagger.(k)y(k) may represent a modified
receive vector and {circumflex over (n)}(k)=Q.sup..dagger.(k)n(k)
may represent modified noise factor. For a unitary matrix Q(k), a
covariance matrix of {circumflex over (n)}(k) may remain same on
pre-multiplication with Q.sup..dagger.(k).
[0040] From a set of equations in (6), first ZF estimate of
subsymbol x.sub.n.sub.T(k) may be computed in a n.sub.T.sup.th
spatial stream. After estimating x.sub.n.sub.T(k), ZF estimate of
x.sub.n.sub.T.sub.-1(k) may be computed by removing interference of
x.sub.n.sub.T(k) from y.sub.n.sub.T.sub.-1(k) using its estimate.
In general, ZF estimate {circumflex over (x)}.sub.i(k) of subsymbol
x.sub.i(k) in an i.sup.th spatial stream after removing an
interference of data subsymbols x.sub.i+1(k),x.sub.i+2(k) . . .
x.sub.n.sub.T(k) from a modified received symbol y.sub.i(k), may be
represented by equation (7):
x ^ i ( k ) = y ^ i ( k ) - j = i + 1 n T r ij ( k ) x ^ j ( k ) r
ii ( k ) ( 7 ) ##EQU00005##
[0041] Now to compute ZF and MMSE estimate of subsymbols for BSU
304, in QR-ZF MIMO detector with MMSE Flag=1; BSU 304 in QR-MMSE
MIMO detector may estimate soft symbols using Q(k) and R(k)
matrices obtained via QRD of an augmented matrix H.sub.aug(k).
Equation (1) may be pre-multiplied by Q.sub.1.sup..dagger.(k) and
equation (3) may be used to get a modified receive vector as
equation (8):
y ^ ( k ) = Q 1 .dagger. ( k ) H ( k ) x ( k ) + Q 1 .dagger. ( k )
n ( k ) = ( R ( k ) - .sigma. Q 2 .dagger. ( k ) ) x ( k ) + Q 1
.dagger. ( k ) n ( k ) ( 8 ) ##EQU00006##
[0042] By following an approach for solving a system of equations
as in BSU for QR-ZF detector, MMSE estimate {circumflex over
(x)}.sub.i(k) of subsymbol x.sub.i(k) in an i.sup.th spatial stream
using QR-MMSE BSU may be expressed by equation (9) as follows:
x i ( k ) = y i ( k ) - j = i + 1 n T ( r ij ( k ) - .sigma. q 2 ,
ij ( k ) ) x j ( k ) r ii ( k ) - .sigma. q 2 , ii ( k ) ( 9 )
##EQU00007##
where Q.sub.2(k)=[q.sub.2,ij].sub.n.sub.T.sub..times.n.sub.T and
y(k)=[y.sub.1(k)y.sub.2(k) . . . y.sub.n.sub.T(k)].sup.T. From
equations (7) and (9), it may be observed that MMSE estimates may
reduce to ZF estimates when noise variance .sigma..sup.2 is
zero.
[0043] Further referring to FIG. 3 in which a soft demapper module
306 is proposed, that may calculate bit LLR's separately for each
stream. Let b.sub.i.sup.l(k) denote a l.sup.th bit of a subsymbol
x.sub.i(k) in a k.sup.th subcarrier and i.sup.th stream, where
l=1,2 . . . log.sub.2M. Following equations (2) and (3), a bit LLR
metric for b.sub.i.sup.l(k) may be represented by equation
(10):
L(b.sub.i.sup.l(k))=.GAMMA..sub.i(k).DELTA.(b.sub.i.sup.l(k))
(10)
wherein, .GAMMA..sub.i(k) is an effective SNR in an i.sup.th stream
that may be computed and
.DELTA.(b.sub.i.sup.l(k))=min.sub.{tilde over
(x)}.sub.i.sub.(k).epsilon.S.sub.0.sub.l|{circumflex over
(x)}.sub.i(k)-{tilde over (x)}.sub.i(k)|.sup.2-min.sub.{tilde over
(x)}.sub.i.sub.(k).epsilon.S.sub.1.sub.l|{circumflex over
(x)}.sub.i(k)-{tilde over (x)}.sub.i(k)|.sup.2
is the bit LLR for b.sub.i.sup.l(k) as computed in equation (5) for
various modulation schemes. Moreover, {circumflex over
(x)}.sub.i(k) may be an estimated subsymbol from BSU, {circumflex
over (x)}.sub.i(k) may be any QAM constellation symbol that may
belong to a constellation set S. S.sub.0.sup.l may represent a set
of QAM symbols having a bit 0 in l.sup.th position of {tilde over
(x)}.sub.i(k) and S.sub.1.sup.l may represent a set of QAM symbols
having a bit 1 in l.sup.th position of {tilde over
(x)}.sub.i(k).
[0044] In an example embodiment, SNR on each stream from noise
coefficients may be computed in NCCU 308. Then soft bits
.DELTA.(b.sub.i.sup.l(k)) may be scaled in SSU 310 using SNR
estimate. Novelty lies in computation of effective noise variance
in a particular stream after considering interference from other
streams. One skilled in the art would appreciate that complexity of
a proposed method is also significantly less compared to the
existing methods as bit LLRs may be computed in a single iteration
of QRD and without matrix inversion. Procedure for estimating SNR
is different for QR-ZF and QR-MMSE detectors and a type of detector
may again be selected by MMSE flag.
[0045] In FIG. 3 information from Q(k) and R(k) matrices may be
utilized by a Noise Coefficient Computation Unit (NCCU) 308 to
compute noise coefficients which may aid in estimating
SNR,.GAMMA..sub.i(k), of received signal in each stream. ZF/MMSE
estimate {circumflex over (x)}.sub.i(k) at the output of BSU 304
for i.sup.th spatial stream may be expressed as:
{circumflex over (x)}.sub.i(k)=x.sub.i(k)+z.sub.i(k) (11)
[0046] In equation (11), x.sub.i(k) is a transmitted subsymbol in
an i.sup.th stream and z.sub.i(k) is an effective noise in i.sup.th
stream. From equations (7) and (9) as well, it is clear that
ZF/MMSE estimate in an i.sup.th spatial stream may depend on
subsymbol estimates over i+1,i+2 . . . n.sub.T streams which may
make effective noise z.sub.i(k) in an i.sup.th stream a
superposition of noise terms in all these streams. Hence, this
effective noise term z.sub.i(k) in an i.sup.th stream may be
expressed as a linear combination of these noise terms and may be
represented as in equation (12):
z i ( k ) = j = i n T c ij ( k ) n ^ j ( k ) ( 12 )
##EQU00008##
wherein, coefficient c.sub.ij(k) may correspond to a contribution
of noise term in a j.sup.th stream to overall noise in an i.sup.th
stream. Noise coefficients c.sub.ij(k) in equation (12) may be
computed by NCCU unit 308. Noise coefficients for ZF and MMSE may
be computed differently and hence, MMSE flag may indicate
generation of discussed coefficients.
[0047] When MMSE Flag=0, then NCCU 308 in QR-ZF MIMO detector may
compute noise coefficients in each stream considering ZF estimates
from BSU 304. Generalized computation of noise coefficients for
n.sub.T.sup.th and (n.sub.T-1).sup.th streams is disclosed in an
example embodiment.
[0048] When MMSE Flag=0, then NCCU 308 in QR-ZF MIMO detector may
compute noise coefficients in each stream considering ZF estimates
from BSU 304. Generalized computation of noise coefficients for
n.sub.T.sup.th and (n.sub.T-1).sup.th streams is disclosed in an
example embodiment.
[0049] ZF estimate {circumflex over (x)}.sub.n.sub.T(k) in a
n.sub.T.sup.th spatial stream from equations (6) and (7) for a
QR-ZF MIMO detector may be expressed as:
x ^ n T ( k ) = y ^ n T ( k ) r n T n T ( k ) = x n T ( k ) + 1 r n
T n T ( k ) n ^ n T ( k ) ( 13 ) ##EQU00009##
[0050] Equations (11), (12) and (13) may be compared to get a noise
coefficient as
c n T n T ( k ) = 1 r n T n T ( k ) ##EQU00010##
Since, R(k) matrix is an upper triangular, the coefficients
c.sub.n.sub.T.sub.j(k) may be zero for j<n.sub.T.
[0051] Now, x.sub.n.sub.T.sub.-1(k) may be solved by substituting
an estimate of {circumflex over (x)}.sub.n.sub.T(k) in equation
(7), ZF estimate {circumflex over (x)}.sub.n.sub.T.sub.-1(k) in
(n.sub.T-1).sup.th spatial stream may be as follows:
x ^ n t - 1 ( k ) = y ^ n T - 1 ( k ) - r n T - 1 n T ( k ) x ^ n T
( k ) r n T - 1 n T - 1 ( k ) = y ^ n T - 1 ( k ) r n T - 1 n T - 1
( k ) - r n T - 1 n T ( k ) r n T - 1 n T - 1 ( k ) x n T ( k ) - r
n T - 1 n T ( k ) r n T - 1 n T - 1 ( k ) r n T n T ( k ) n ^ n T (
k ) ( 14 ) ##EQU00011##
Substituting for y.sub.n.sub.T.sub.-1(k) using equation (6),
equation (14) may be reduced to
x ^ n T - 1 ( k ) = x n T - 1 ( k ) + 1 r n T - 1 n T - 1 ( k ) n ^
n T - 1 ( k ) - r n T - 1 n T ( k ) r n T - 1 n T - 1 ( k ) r n T n
T ( k ) n ^ n T ( k ) ( 15 ) ##EQU00012##
[0052] Finally, noise coefficients for effective noise
z.sub.n.sub.T.sub.-1(k) may be retrieved from equations (12) and
(15) as:
c n T - 1 n T - 1 ( k ) = 1 r n T - 1 n T - 1 ( k ) and
##EQU00013## c n T - 1 n T ( k ) = - r n T - 1 n T ( k ) c n T n T
( k ) r n T - 1 n T - 1 ( k ) . ##EQU00013.2##
Similarly, computation of noise coefficients c.sub.ij(k) may be
extended to a general case as:
c ij ( k ) = { 1 r ii ( k ) ; for i = j 0 ; for i > j - m = i +
1 j r im ( k ) c mj ( k ) r ii ( k ) ; for i < j ( 16 )
##EQU00014##
wherein, noise coefficient c.sub.ij(k) is a contribution of noise
term in j.sup.th stream to an overall noise z.sub.i(k) in an
i.sup.th stream.
[0053] Now, when MMSE Flag=1, then NCCU 308 in QR-ZF MIMO detector
may compute noise coefficients in each stream by repeating above
procedure, and noise coefficient c.sub.ij(k) of noise term in a
j.sup.th stream contributing to resultant noise z.sub.i(k) in an
i.sup.th stream may be obtained as
c ij ( k ) = { 1 r ii ( k ) - .sigma. q 2 , ii ( k ) ; for i = j 0
; for i > j - m = i + 1 j ( r im ( k ) - .sigma. q 2 , im ( k )
) c mj ( k ) r ii ( k ) - .sigma. q 2 , ii ( k ) ; for i < j (
17 ) ##EQU00015##
[0054] The soft demapper 306 may also include a SNR scaling unit
(SSU) 310 may be employed to estimate an effective SNR
.GAMMA..sub.i(k) in an i.sup.th stream and scale bit LLR,
.DELTA.(b.sub.i.sup.l(k)), following equation (10). For data
subsymbols x.sub.i(k) assumed to be of unit variance, effective SNR
.GAMMA..sub.i(k) in an i.sup.th stream may be represented as:
.GAMMA..sub.i(k)=[.sigma..sub.i.sup.2(k)].sup.-1 (18)
wherein, .sigma..sub.i.sup.2(k) is a variance of effective noise
z.sub.i(k) in an i.sup.th stream.
[0055] Referring equation (18), effective noise variance may need
to be computed in each stream for SNR estimation. However, noise
variance computation may differ in QR-ZF and QR-MMSE MIMO detectors
and is detailed separately for the two detectors in following
example embodiments.
[0056] When MMSE Flag=0, in a QR-ZF MIMO detector, variance
.sigma..sub.i.sup.2(k) of effective noise z.sub.i(k) for an
i.sup.th stream may be computed directly as:
.sigma. i 2 ( k ) = E { z i 2 ( k ) } = j = i n T c ij ( k ) 2
.sigma. 2 ( 19 ) ##EQU00016##
Therefore, bit LLR for a bit b.sub.i.sup.l(k) from equation (10)
after SNR scaling (ignoring the common term .sigma..sup.2 which may
not affect the performance) may be expressed as:
L(b.sub.i.sup.l(k)).apprxeq.g.sub.i(k).DELTA.(b.sub.i.sup.l(k))
(20)
wherein, SNR scale factor g.sub.i(k) may be:
g.sub.i(k)=[.SIGMA..sub.j=i.sup.n.sup.T|c.sub.ij(k)|.sup.2].sup.-1
(21)
[0057] Now when MMSE Flag=1, in QR-MMSE MIMO detector, Q.sub.1(k)
obtained by QRD of augmented matrix H.sub.aug(k) may not be a
unitary matrix and covariance matrix of {circumflex over (n)}(k)
may be different from n(k). Hence, z.sub.i(k) in terms of original
noise vector n(k) may be expressed as:
z i ( k ) = j = i n T c ij ( k ) m = 1 n T q 1 , mj * ( k ) n m ( k
) ( 22 ) ##EQU00017##
Therefore, variance of the effective noise z.sub.i(k) on an
i.sup.th stream may get computed to be:
.sigma..sub.i.sup.2(k)=E{z.sub.i.sup.2(k)}=.SIGMA..sub.m=1.sup.n.sup.T|E-
.sub.j=i.sup.n.sup.Tc.sub.ij(k)q.sub.1,mj*(k)|.sup.2.sigma..sup.2
(23)
[0058] Bit LLR for b.sub.i.sup.l(k) in a QR based MMSE detector
(after ignoring the common term .sigma..sup.2 which may not affect
the performance) may be represented as given in equation (20), with
SNR scale factor g.sub.i(k) as:
g.sub.i(k)=[.SIGMA..sub.m=1.sup.n.sup.T|E.sub.j=i.sup.n.sup.Tc.sub.ij(k)-
q.sub.1,mj*(k)|.sup.2].sup.-1 (24)
It is also clearly observed that at higher SNRs (.gtoreq.10 dB),
scale factors derived from a QR based MMSE detector may reduce to
those given for its ZF counterpart.
[0059] Usefulness of a proposed soft demapper in an example
embodiment, may be easily shown by performing simulations for
MIMO-OFDM system. In simulations, 802.11ac WLAN system with 2
transmit antennas and 2 receive antennas may be considered. All
simulations may be performed for VHT frame format with two spatial
streams operating on 40 MHz channel bandwidth. The transmitted
signal may be passed through analog front end (AFE) section with
analog filters, DAC/ADC, TG-ac channel model and RF impairments.
MIMO channel may introduce path-loss and shadow fading to a
transmitted output signal. 1000 different realizations of channel
and noise may be considered. On receiver side, timing and frequency
synchronizations may be performed before estimating MIMO channel.
Then, bit LLR's using proposed method may be computed. The computed
bit LLR's may be passed through a deinterleaver, stream deparser
and decoded using a soft input Viterbi decoder.
[0060] In the present simulations, bit LLR's in the proposed method
may be computed using equation (20) with scale factors for QR based
ZF detector given in equation (21). In existing QR-MMSE method, as
the computation of bit LLR's from resultant signal model is not
given, bit LLR's in proposed methods and apparatuses may be
computed by calculating the SNR from a signal model and
substituting it in equation (10) and may also be computed using
near-optimum ML bit metrics.
[0061] FIG. 4 is a flowchart illustrating a method for calculating
log-likelihood ratios, according to an example embodiment. At step
402, samples of one or more spatial streams are received by the
communication receiver. At step 404, one or more demodulated
samples of the received samples are determined by applying one or
more discrete Fourier transforms. At step 406, one or more
subsymbols from the demodulated samples are determined by utilizing
a MIMO detection method. In one aspect, the MIMO detection method
is minimum mean squared error (MMSE) method. In another aspect, the
MIMO detection method is zero forcing (ZF) method.
[0062] At step 408, calculating a log-likelihood ratio for each bit
of one or more subsymbols of each of the one or more spatial
streams. The step of calculating log-likelihood ratio for each bit
comprises computing effective noise on one or more spatial streams
after considering noise terms resulting from the MIMO detection
estimates of one of more subsymbols on one or more spatial streams.
The log likelihood ratio for a bit is equal to
.GAMMA..sub.i(k).DELTA.(b.sub.i.sup.l(k)), wherein .GAMMA..sub.i(k)
is the effective signal to noise ratio in i.sup.th spatial stream
and k.sup.th subcarrier, and b.sub.i.sup.l(k) is the l.sup.th bit
of the subsymbol x.sub.i(k) in the k.sup.th subcarrier and i.sup.th
stream.
[0063] Finally at step 410, signal to noise ratio is determined on
one or more spatial streams from the effective noise and scaling
bit log-likelihood ratios with the signal to noise ratio. The
effective signal to noise ratio in i.sup.th spatial stream and
k.sup.th subcarrier is a function of effective noise term
z.sub.i(k) in the i.sup.th spatial stream, wherein
z i ( k ) = j = i n T c ij ( k ) n ^ j ( k ) ##EQU00018##
such that c.sub.ij(k) is the coefficient of contribution
corresponding to the contribution of the noise term in the j.sup.th
stream to the overall noise in the i.sup.th stream.
[0064] In case, the MIMO detection method is minimum mean squared
error (MMSE) method then the coefficient of contribution
c.sub.ij(k) as calculated above is equal to
c ij ( k ) = { 1 r ii ( k ) - .sigma. q 2 , ii ( k ) ; for i = j 0
; for i > j - m = i + 1 j ( r im ( k ) - .sigma. q 2 , im ( k )
) c mj ( k ) r ii ( k ) - .sigma. q 2 , ii ( k ) ; for i < j
##EQU00019##
where r.sub.ii(k) are the elements of an upper triangular matrix
for i.sup.th spatial stream and k.sup.th subcarrier, .sigma..sup.2
is the noise variance of complex Additive white Gaussian noise,
q.sub.2ij(k) is an element of a matrix obtained by performing QR
decomposition on an augmented channel matrix.
[0065] In another case, the MIMO detection method is zero forcing
(ZF) method then the coefficient of contribution c.sub.ij(k) as
calculated above is equal to
c ij ( k ) = { 1 r ii ( k ) ; for i = j 0 ; for i > j - m = i +
1 j r im ( k ) c mj ( k ) r ii ( k ) ; for i < j ##EQU00020##
[0066] where r.sub.ii(k) are the elements of an upper triangular
matrix for i.sup.th spatial stream and k.sup.th subcarrier.
[0067] FIG. 5 illustrates a PER Performance in TGac channel model A
for 2.times.2 VHT frame with MCS7, according to an example
embodiment. Further, FIG. 6 illustrates a PER Performance in TGac
channel model D for 2.times.2 VHT frame with MCS7, according to an
example embodiment. FIGS. 5 and 6 which show PER performance of a
2.times.2 WLAN system with a proposed method, near-optimum ML and
existing QR-MMSE technique for VHT frame format and MCS 7 (64-QAM
modulation and code rate 5/6) in channel model A (flat fading
channel) and channel model D (frequency selective channel)
respectively.
[0068] From FIG. 5 and FIG. 6, it may be clearly noticed that
proposed methods and apparatuses, and QR-MMSE method may perform
similarly in both channel models and may remain inferior to that of
near-optimal ML by approximately 3 dB. However, the computational
complexity of proposed methods and apparatuses is significantly
less compared to both near-optimal ML method and existing QR-MMSE
method. As scale factors derived for a QR based MMSE detector
reduce to those given for its ZF counterpart, for SNR's considered
herein, PER performance of a proposed MMSE detector is same as that
of the proposed ZF detector.
[0069] Finally, for a square matrix of size n, the complexity
analysis of a QR-MMSE MIMO detector with proposed method is
compared with that of the existing QR-MMSE method in Table 1.
TABLE-US-00001 TABLE 1 Complexity analysis of the proposed method
with the existing QR MMSE method Proposed QR Existing QR Operations
MMSE Method MMSE Method Multiplications 2n.sup.3 + 5n.sup.2 - 7n +
3 2n.sup.4 + 2n.sup.3 - 4n.sup.2 + 4n Divisions 2n - 1 n(n - 1)
Square Root n - 1 n(n - 1) Cordic units for n - 1 n(n - 1) phase
estimation
[0070] In the complexity analysis, Householder Reflection may be
employed for QR decomposition of the square matrix with complex
elements. Also, each complex multiplication may be considered to
have been implemented using 3 real multipliers.
[0071] The techniques described herein may be used for various
wireless communication networks such as Orthogonal Frequency
Division Multiplexing (OFDM) networks, Time Division Multiple
Access (TDMA) networks, Frequency Division Multiple Access (FDMA)
networks, Orthogonal FDMA (OFDMA) networks, Single-Carrier FDMA
(SC-FDMA) networks, Code Division Multiple Access (CDMA) networks,
etc. The terms "networks" and "systems" are often used
interchangeably. A CDMA network may implement a radio technology
such as Universal Terrestrial Radio Access (UTRA), CDMA2000, etc.
UTRA includes Wideband-CDMA (W-CDMA) and Low Chip Rate (LCR).
CDMA2000 covers IS-2000, IS-95 and IS-856 standards. A TDMA network
may implement a radio technology such as Global System for Mobile
Communications (GSM). An OFDMA network may implement a radio
technology such as Evolved UTRA (E-UTRA), IEEE 802.11, IEEE 802.16
(e.g., WiMAX (Worldwide Interoperability for Microwave Access)),
IEEE 802.20, Flash-OFDM.RTM., etc. UTRA, E-UTRA, and GSM are part
of Universal Mobile Telecommunication System (UMTS). Long Term
Evolution (LTE) and Long Term Evolution Advanced (LTE-A) are
upcoming releases of UMTS that use E-UTRA. UTRA, E-UTRA, GSM, UMTS
and LTE are described in documents from an organization named "3rd
Generation Partnership Project" (3GPP). CDMA2000 is described in
documents from an organization named "3rd Generation Partnership
Project 2" (3GPP2). CDMA2000 is described in documents from an
organization named "3rd Generation Partnership Project 2" (3GPP2).
These various radio technologies and standards are known in the
art. For clarity, certain aspects of the techniques are described
below for LTE and LTE-A. The teachings herein may be incorporated
into (e.g., implemented within or performed by) a variety of wired
or wireless apparatuses (e.g., nodes). In some aspects a node
comprises a wireless node. Such wireless node may provide, for
example, connectivity for or to a network (e.g., a wide area
network such as the Internet or a cellular network) via a wired or
wireless communication link. In some aspects, a wireless node
implemented in accordance with the teachings herein may comprise an
access point or an access terminal.
[0072] In an example embodiment, a non-transient computer-readable
medium may be provided comprising instructions for causing a
programmable processor to receive, by the communication receiver,
samples of one or more spatial streams. Then, determining one or
more demodulated samples of the received samples by applying one or
more discrete Fourier transforms. Thereafter, determining one or
more subsymbols from the demodulated samples by utilizing a MIMO
detection method. A log-likelihood ratio is calculated for each bit
of one or more subsymbols of each of the one or more spatial
streams, wherein calculating log-likelihood ratio for each bit
comprises of computing effective noise on one or more spatial
streams after considering noise terms resulting from the MIMO
detection estimates of one of more subsymbols on one or more
spatial streams. Finally, the processor may enable determining
signal to noise ratio on one or more spatial streams from the
effective noise and scaling bit log-likelihood ratios with the
signal to noise ratio.
[0073] Though the example embodiments described herein are related
to MIMO-OFDM system, however, the methods, apparatuses, and systems
described herein may also be used for various broadband wireless
communication systems including Orthogonal Frequency Division
Multiple Access (OFDMA) systems, Single-Carrier Frequency Division
Multiple Access (SC-FDMA) systems, and the like without deviating
from the essence of the disclosure.
[0074] Embodiments of the present disclosure may be provided as a
computer program product, which may include a computer-readable
medium tangibly embodying thereon instructions, which may be used
to program a computer (or other electronic devices) to perform a
process. The computer-readable medium may include, but is not
limited to, fixed (hard) drives, magnetic tape, floppy diskettes,
optical disks, compact disc read-only memories (CD-ROMs), and
magneto-optical disks, semiconductor memories, such as ROMs, random
access memories (RAMs), programmable read-only memories (PROMs),
erasable PROMs (EPROMs), electrically erasable PROMs (EEPROMs),
flash memory, magnetic or optical cards, or other type of
media/machine-readable medium suitable for storing electronic
instructions (e.g., computer programming code, such as software or
firmware). Moreover, embodiments of the present disclosure may also
be downloaded as one or more computer program products, wherein the
program may be transferred from a remote computer to a requesting
computer by way of data signals embodied in a carrier wave or other
propagation medium via a communication link (e. g., a modem or
network connection).
[0075] Moreover, although the present disclosure and its advantages
have been described in detail, it should be understood that various
changes, substitutions and alterations can be made herein without
departing from the disclosure as defined by the appended claims.
Moreover, the scope of the present application is not intended to
be limited to the particular embodiments of the process, machine,
manufacture, composition of matter, means, methods and steps
described in the specification. As one will readily appreciate from
the disclosure, processes, machines, manufacture, compositions of
matter, means, methods, or steps, presently existing or later to be
developed that perform substantially the same function or achieve
substantially the same result as the corresponding embodiments
described herein may be utilized. Accordingly, the appended claims
are intended to include within their scope such processes,
machines, manufacture, compositions of matter, means, methods, or
steps.
* * * * *