The burden of designing multiple training sequences for systems having multiple transmit antennas, is drastically reduced by employing a single sequence from which the necessary multiple sequences are developed. The single sequence is selected to create sequences that have an impulse-like autocorrelation function and zero cross correlations. A sequence of any desired length nt can be realized for an arbitrary number of channel taps, L. The created sequences can be restricted to a standard constellation (that is used in transmitting information symbols) so that a common constellation mapper is used for both the information signals and the training sequence. In some applications a a training sequence may be selected so that it is encoded with the same encoder that is used for encoding information symbols. Both block and trellis coding is possible in embodiments that employ this approach.
|
1. A space-time diversity transmitter that includes (a) n transmitting antennas, where n is greater than one, (b) first encoder responsive to an applied information bit stream, said encoder developing n symbol streams, and (c) a constellation mapper responsive to said n symbol streams, for mapping symbols of each of said n symbol streams into a standard signal constellation to create a corresponding mapped stream, said constellation mapper applying each of the created n mapped streams to a different one of said n antennas in blocks of nt mapped symbols that are synchronized in time to each other, the improvement comprising:
a training generator for generating either n sequences of signals or a sequence of signals that contains n subsequences of signal;
a second encoder for creating n training symbol sequences from said signals created by said training generator, each containing nt symbols, where nt represents a training sequence length; and
said constellation mapper is configured to map said n training symbol sequences onto a standard constellation to develop n mapped training sequences, and to apply the n mapped training sequences to said n antennas at times when said constellation mapper stops applying said n mapped stream to said n antennas;
where said n training symbol sequences have an impulse-like autocorrelation function and zero cross correlation.
17. A space-time diversity transmitter that includes (a) n transmitting antennas, where n is greater than one, (b) first encoder responsive to an applied information bit stream, said encoder developing n symbol streams, and (c) a constellation mapper responsive to said n symbol streams, for mapping symbols of each of said n symbol streams into a standard signal constellation to create a corresponding mapped stream, said constellation mapper applying each of the created n mapped streams to a different one of said n antennas in blocks of nt mapped symbols that are synchronized in time to each other, the improvement comprising:
a training generator for generating either n sequences of signals or a sequence of signals that contains n subsequences of signal;
a second encoder for creating n training symbol sequences from said signals created by said training generator, each containing nt symbols, where nt represents a training sequence length; and
said constellation mapper is configured to map said n training sequences onto a standard constellation to develop n mapped training sequences, and to apply the n mapped training sequences to said n antennas at times when said constellation mapper stops applying said n mapped stream to said n antennas;
where said n training sequences have an impulse-like autocorrelation function and zero cross correlation, and where n2 n=2, said generator creates a sequence s, and said second encoder creates a first training sequence that is equal to −s concatenated with s, and a second training sequence that is equal to s concatenated with s.
2. The transmitter of
0. 3. The transmitter of
4. The transmitter of
5. The transmitter of
6. The transmitter of
7. The transmitter of
8. The transmitter of
9. The transmitter of
10. The transmitter of
0. 12. The transmitter of
(a) said n training sequences are selected to enable a receiver that receives said n training sequences to determine characteristics of a transmission medium between said transmitter and said receiver, and
(b) said generator develops said symbols sequence to comprise nt+(n−2)L+m symbols long, where L is the number of channel parameters that describe a channel within said transmission medium from one of said n transmitter antennas to a receiving antenna of said receiver, m is number of symbols stored in a memory of said trellis encoder, and nt is not less than 2L−1.
13. The transmitter of
14. The transmitter of
15. The transmitter of
16. The transmitter of
18. The transmitter of
|
This application
where S is a convolution matrix of dimension (Nt′−L′+1)×L′. Again, for optimality, the imposed requirement is that
and once the sequence s is found, the task is to create the subsequences s1 and s2 from the found sequence s. Preferably, the subsequences s1 and s2 can be algorithmically generated from sequence s. Conversely, one may find subsequences s1 and s2 that satisfy the requirements of equation (8) and be such that sequence s can be algorithmically generated. This permits the use of a single training signal generator that, through a predetermined algorithm (i.e., coding) develops the subsequences s1 and s2. Both approaches lead to embodiment depicted in
Actually, once we realized that the complexity of the training sequence determination problem can be reduced by focusing on the creation of a single sequence from which a plurality of sequences that meet the requirements of equation (8) can be generated, it became apparent that there is no requirement for s to be longer than s1 and s2.
In response to the training sequences transmitted by antennas 11 and 12, receiving antenna 21 develops the signal vector y (where the elements of the vector y are the signals received from antennas 11 and 12). Considering the received signal during the first Nt/2 time periods as y1 and during the last Nt/2 time periods as y2, and employing only the useful portion of the signal (that is, the portions not corrupted by signals that are not part of the training sequence) one gets
where S is a convolution matrix of dimension (Nt−L+1)×L. In accordance with the principles disclosed herein, the
If the sequence s is such that SHS=(Nt−L+1)IL, then
If the noise is white, then the linear processing at the receiver does not color it, and the channel transfer functions correspond to
with a minimum squared error, MSE, that achieves the lower bound expressed in equation (7); to wit,
The above result can be generalized to allow any matrix U to be used to encode the training sequence, s, so that
as long as UHU=2I for a two antennas case, and UHU=KI for an K antenna case.
Whereas
The
With a signal arrangement as shown in
where the matrices Di and {tilde over (D)}i (for i=1, 2) are convolution matrices for d1 and {tilde over (d)}1, respectively, of dimension (Nt−L+1)×L. Recalling from equation (8) that MMSE is achieved if and only if DHD has zeros off the diagonal; i.e.,
and identity matrices on the diagonal; i.e.,
Various arrangements that interrelate sequences d1 and d2 can be found that meet the above requirement. By way of example (and not by way of limitation), a number of simple choices satisfy these conditions follow.
(1) (D1*)TD1=(Nt−L+1)IL, {tilde over (D)}1=D1, and D2=D1. To show that equation (21) holds, one may note that {tilde over (D)}2T{tilde over (D)}2* (the first term in the equation) becomes D1TD1*, but if (D1*)TD1 is a diagonal matrix then so is {tilde over (D)}2T{tilde over (D)}2*. Thus, according to this training sequence embodiment, one needs to only identify a sequence d1 that is symmetric about its center, with an impulse-like autocorrelation function, and set d2 equal to d1. This is shown in FIG. 7.
(2) (D1*)TD1=(Nt−L+1)IL, and {tilde over (D)}2=D1. To show that equation (21) holds, one may note that the {tilde over (D)}2T{tilde over (D)}2* first term in the equation also becomes D1TD1*. Thus, according to this training sequence embodiment, one needs to only identify a sequence d1 with an impulse-like autocorrelation function, and set d2 equal to {tilde over (d)}1. This is shown in FIG. 8.
(3) (D1*)TD1=(Nt−L+1)IL, and {tilde over (D)}2*=D1. To show that equation (21) holds, one may note that the {tilde over (D)}2T{tilde over (D)}2* first term in the equation becomes (D1*)TD1. Thus, according to this training sequence embodiment, one needs to only identify a sequence d1 with an impulse-like autocorrelation function, and set d2 equal to {tilde over (d)}1*. This is shown in FIG. 9.
Training Sequences Employing Trellis Coding
Consider a trellis code with m memory elements and outputs from a constellation of size C, over a single channel with memory 2mC(L−1)−1. To perform joint equalization and decoding one needs a product trellis with 2mC(L−1) states. For a space-time trellis code with m memory elements, n transmit antennas and one receive antenna, over a channel with memory (L−1), one needs a product trellis with 2mCn(L−1).
The receiver can incorporate the space-time trellis code structure in the channel model to create an equivalent single-input, single output channel, heq, of length m+L. The trellis, in such a case, involves C(m+L−1) states. The approach disclosed herein uses a single training sequence at the input of the space-time trellis encoder to directly estimate heq used by the joint space-time equalizer/decoder. The channel heq that incorporates the space-time code structure typically has a longer memory than the channel h1 and h2 (in a system where there are two transmitting antennas and one receiving antenna).
To illustrate, assume an encoder 30 as depicted in
With such an arrangement, the received signal at time k can be expressed as
A block of received signals (corresponding to the useful portion of the training sequence block) can be expressed in matrix form by
y=Sheq+z (25)
where
and following the principles disclosed above, it can be realized that when the training sequence is properly selected so that SHS is a diagonal matrix, i.e., SHS=(Nt−L)IL+1, an estimate of heq that is, ĥeq, is obtained from
If the training sequence were to comprise only the even constellation symbols, ei2πk/8, k=0, 2, 4, 6, per equation (24), the elements of {tilde over (h)}eq would correspond to
If the training sequence were to comprise only the odd constellation symbols, ei2πk/8, k=1, 3, 5, 7, the elements of {tilde over (h)}eq would correspond to
If the training sequence were to comprise a segment of only even constellation symbols followed by only odd constellation symbols (or vice versa), then channel estimator 22 within receiver 20 can determine the heqeven coefficients from the segment that transmitted only the even constellation symbols, and can determine the heqodd coefficients from the segment that transmitted only the even constellation symbols. Once both heqeven and heqeven and heqodd are known, estimator 22 can obtain the coefficients of h1 from
and the coefficients of h2 from
What remains, then, is to create a single training sequence s of length Nt where one half of it (the seven portion) consists of only even constellation symbols (even sub-constellation), and another half of it (the sodd portion) consists of only odd constellation symbols (odd sub-constellation). The sequences s1 and s2 of length Nt are derived from the sequence s by means of the 8-PSK space-time trellis encoder. The sequences s1 and s2 must also meet the requirements of equation (8). Once seven is found, sodd can simply be
sodd=αseven, where α=eiπk/4 for any k=1,3,5, 7. (32)
Therefore, the search for sequence s is reduced from a search in the of 8N
For a training sequence of length Nt=26, with an 8-PSK space-time trellis encoder, we have identified the 12 training sequences specified in Table 1 below.
TABLE 1
sequence #
α
Se
1
exp(i5π/4)
−1 1 1 1 1 −1 −i −1 1 1 −1 1 1
2
exp(i3π/4)
1 1 −1 1 i i 1 −i i −1 −1 −1 1
3
exp(iπ/4)
1 −1 −1 −i i −i 1 1 1 −i −1 1 1
4
exp(iπ/4)
1 −1 −1 −i 1 −1 1 −i −i −i −1 1 1
5
exp(iπ/4)
1 i 1 1 i −1 −1 i 1 −1 1 i 1
6
exp(i3π/4)
1 i 1 i −1 −1 1 −1 −1 i 1 i 1
7
exp(i7π/4)
1 −i 1 1 −i −1 −1 −i 1 −1 1 −i 1
8
exp(i5π/4)
1 −i 1 −1 i 1 −1 −i 1 1 1 −i 1
9
exp(i3π/4)
−1 1 1 1 −1 −1 −i −1 −1 1 −1 1 1
10
exp(i7π/4)
−1 i −1 −i 1 −i i i 1 i 1−i 1
11
exp(iπ/4)
−1 −i −1 i 1 i −i −i 1−i 1 i 1
12
exp(i3π/4)
−1 −i −1 i −1 i −i −i −1 −i 1 i 1
Construction of Training Sequence
While the above-disclosed materials provide a very significant improvement over the prior art, there is still the requirement of selecting a sequence s1 with an impulse-like autocorrelation function. The following discloses one approach for identifying such a sequence without having to perform an exhaustive search.
A root-of-unity sequence with alphabet size N has complex roots of unity elements of the form
As indicated above, the prior art has shown that perfect roots-of-unity sequences (PRUS) can be found for any training sequence of length Nt, as long as no constraint is imposed on the value of N. As also indicated above, however, it is considered disadvantageous to not limit N to a power of 2. Table 2 presents the number of PRUSs that were found to exist (through exhaustive search) for different sequence lengths when the N is restricted to 2 (BPSK), 4 (QPSK), or 8 (8-PSK). Cell entries in Table 2 with “-” indicate that sequence does not exist, and blank cells indicate that an exhaustive search for a sequence was not performed.
TABLE 2
Nt =
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
BPSK
—
—
8
—
—
—
—
—
—
—
—
—
—
—
—
—
—
QPSK
8
—
32
—
—
—
128
—
—
—
—
—
—
—
6144
—
—
8-PSK
16
—
128
—
—
—
512
—
—
—
—
—
A sequence s of length Nt is called L-perfect if the corresponding training matrix S of dimension (Nt−L+1)×L satisfies equation (8). Thus, an L-perfect sequence of length Nt is optimal for a channel with L taps. It can be shown that the length Nt of an L-perfect sequence from a 2p-alphabet can only be equal to
which is a necessary, but not sufficient, condition for L-perfect sequences of length Nt. Table 3 shows the minimum necessary Nt for L=2, 3, . . . 10, the size of the corresponding matrix S, and the results of an exhaustive search for L-perfect sequences (indicating the number of such sequences that were found). Cell entries marked “x” indicate that sequences exist, but number of such sequences it is not known.
TABLE 3
L
2
3
4
5
6
7
8
9
10
Nt
3
6
7
10
11
14
15
18
19
S
2 × 2
4 × 3
4 × 4
6 × 5
6 × 6
8 × 7
8 × 8
10 × 9
10 × 10
BPSK
4
8
8
—
—
—
—
—
—
QPSK
16
64
64
—
—
128
x
8-PSK
64
512
512
x
x
It is known that with a PRUS of a given length, NPRUS, one can estimate up to L=NPRUS unknowns. It can be shown that a training sequence of length Nt is also an L-perfect training sequence if
Nt=kNPRUS+L−1 and k≧1. (34)
Accordingly, an L-perfect sequence of length kNPRUS+L−1 can be constructed by selecting an NPRUSsequence, repeating it k times, and circularly extending it by L−1 symbols. Restated and amplified somewhat, for a selected PRUS of a given NPRUS, i.e.,
the L-perfect sequence of length kNPRUS+L−1 is created from a concatenation of k sp(NPRUS) sequences followed by the first L−1 symbols of sp(NPRUS), or from a concatenation of the last L−1 symbols of sp(NPRUS) followed by k sp(NPRUS) sequences.
To illustrate, assume that the number of channel “taps” that need to be estimated, L, is 5, and that a QPSK alphabet is desired to be used. From the above it is known that NPRUS must be equal to or greater than 5, and from Table 2 it is known that the smallest NPRUS that can be found for QSPK that is larger than 5 is NPRUS=8. Employing equation (34) yields
While an L-perfect training sequence cannot be constructed from PRUS sequences for values of Nt other than values derived by operation of equation (34), it is known that, nevertheless, L-perfect sequences may exist. The only problem is that it may be prohibitively difficult to find them. However, in accordance with the approach disclosed below, sub-optimal solutions are possible to create quite easily.
If it is given that the training sequence is Nt long, one can express this length by
Nt=kNPRUS+L−1+M, where M>0 (37)
In accord with our approach, select a value of NPRUS≧L that minimizes M, create a sequence of length kNPRUS+L−1 as disclosed above, and then extend that sequence by adding M symbols. The M added symbols can be found by selecting, through an exhaustive search, the symbols that lead to the lowest estimation MSE. Alternatively, select a value of NPRUS≧L that minimizes M′ in the equation,
Nt=kNPRUS+L−1−M′, where M′>0 (38)
create a sequence of length kNPRUS+L−1 as disclosed above, and then drop the last (or first) M′ symbols.
The receiver shown in
Al-Dhahir, Naofal, Fragouli, Christine, Turin, William
Patent | Priority | Assignee | Title |
Patent | Priority | Assignee | Title |
5914982, | Jun 13 1997 | PCTEL, Inc | Method and apparatus for training linear equalizers in a PCM modem |
5949819, | Jun 13 1997 | Synaptics Incorporated | Method and apparatus for training linear equalizers in a PCM modem |
6424679, | Oct 07 1998 | Intel Corporation | Space time block coded transmit antenna diversity for WCDMA |
6449314, | Oct 07 1998 | Intel Corporation | Space time block coded transmit antenna diversity for WCDMA |
6643338, | Oct 07 1998 | Intel Corporation | Space time block coded transmit antenna diversity for WCDMA |
6788661, | Nov 12 1999 | Nikia Networks Oy | Adaptive beam-time coding method and apparatus |
6865237, | Feb 22 2000 | UNWIRED BROADBAND, INC | Method and system for digital signal transmission |
6891897, | Jul 23 1999 | Apple Inc | Space-time coding and channel estimation scheme, arrangement and method |
7006579, | Sep 04 2000 | UNWIRED BROADBAND, INC | ISI-robust slot formats for non-orthogonal-based space-time block codes |
7139324, | Jun 02 2000 | Intellectual Ventures I LLC | Closed loop feedback system for improved down link performance |
7200182, | Oct 07 1998 | Intel Corporation | Space time block coded transmit antenna diversity for WCDMA |
7272192, | Apr 14 2000 | BROAD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIVERSITY | Time-reversal block transmit diversity system for channels with intersymbol interference and method |
7366266, | Oct 07 1998 | Intel Corporation | Space time block coded transmit antenna diversity for WCDMA |
7613259, | Oct 07 1998 | Intel Corporation | Mobile receiver phase correction circuit |
7701916, | Feb 25 1999 | Texas Instruments Incorporated | Space time transmit diversity for TDD with cyclic prefix midamble |
20020067309, | |||
20040013211, | |||
20050157683, | |||
20050157684, | |||
20050185734, | |||
RE39165, | Jun 13 1997 | PCTEL, Inc | Method and apparatus for training linear equalizers in a PCM modem |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Mar 18 2013 | AT&T Intellectual Property II, L.P. | (assignment on the face of the patent) | / |
Date | Maintenance Fee Events |
Feb 19 2014 | ASPN: Payor Number Assigned. |
Nov 20 2017 | M1552: Payment of Maintenance Fee, 8th Year, Large Entity. |
Feb 14 2022 | REM: Maintenance Fee Reminder Mailed. |
Aug 01 2022 | EXP: Patent Expired for Failure to Pay Maintenance Fees. |
Date | Maintenance Schedule |
Apr 08 2017 | 4 years fee payment window open |
Oct 08 2017 | 6 months grace period start (w surcharge) |
Apr 08 2018 | patent expiry (for year 4) |
Apr 08 2020 | 2 years to revive unintentionally abandoned end. (for year 4) |
Apr 08 2021 | 8 years fee payment window open |
Oct 08 2021 | 6 months grace period start (w surcharge) |
Apr 08 2022 | patent expiry (for year 8) |
Apr 08 2024 | 2 years to revive unintentionally abandoned end. (for year 8) |
Apr 08 2025 | 12 years fee payment window open |
Oct 08 2025 | 6 months grace period start (w surcharge) |
Apr 08 2026 | patent expiry (for year 12) |
Apr 08 2028 | 2 years to revive unintentionally abandoned end. (for year 12) |