System for soft symbol decoding with MIMO log-map detection

1. A method for determining soft metrics in a multiple-input multiple-output (MIMO) communication system, wherein a soft metric is determined for each bit, b i , encoded by a multi-dimensional symbol, wherein the soft metrics indicate discrepancies in the MIMO communication system, the method comprising:

transmitting a sample of the multi-dimensional symbol, wherein the multi-dimensional symbol is represent-able as a complex, K-dimensional vector x, K being an integer corresponding to the number of transmit antennas in the MIMO system, wherein each vector component of the vector x represents a signal transmitted with one of the K transmit antennas, the multi-dimensional symbol being transmitted through a communication channel;

receiving from the communication channel, a sample represented by a complex, N-dimensional vector y, N being an integer corresponding to the number of receive antennas in the MIMO system, and wherein vector y is related to vector x according to an equation: y=Hx+n, where H is an N×K channel matrix modeling phase shifts and amplitude changes on paths between the K transmit and N receive antennas, n is a complex N-dimensional vector representing communication channel noise; and

determining a soft metric L(b i ) for each bit b i encoded by x according to the equation: L(b i )=σ −2 ·(min x j |b i =−1 ∥y−Hx j ∥ 2 −min x j |b i =+1 ∥y−Hx j ∥ 2 ), where σ is a standard deviation of the noise and/or interference, and x j represents the j-th possible value for x.

2. The method of claim 1 wherein the soft metric indicates a likelihood that each bit, b i , encoded by a multi-dimensional symbol has a 0 or 1 state or equivalent states thereof.

3. The method of claim 1 wherein the MIMO system has two transmit and two receive antennas forming at least a 2 by 2 matrix channel.

4. The method of claim 1 wherein all possible values of x are given by 2 B(K−1) , such that complexity of computations is reduced, where B is the number of bits transmitted per symbol.

5. In a MIMO communication system, a reduced complexity method for determining a soft metric L(b i ) for each bit b i encoded by a complex vector x, wherein the soft metrics indicate discrepancies in the MIMO communication system, the method comprising:

receiving via a communication channel, a sample represented by a complex vector y, where complex vector y is related to the complex vector x; and

using complex vector y to determine soft metrics L(b i ) for each bit b i encoded by x according to the equation: L(b i )=σ −2 ·(min x j |b i =−1 ∥y−Hx j ∥ 2 −min x j |b i =+1 ∥y−Hx j ∥ 2 ) where H is a channel matrix modeling phase shifts and amplitude changes for the communication channel; and x j represents the j-th possible value of x from no more than 2 B(K−1) possible values for x, where B is the number of bits transmitted per symbol, and where σ is a standard deviation of noise/interference in a received signal.

6. The method of claim 5 wherein the complex vector x includes a first element and a second element, the method further comprising:

determining a first symbol value by making a hard decision for each bit b i of the first element;

using the first symbol value to determine soft metrics L(b i ) for each bit b i encoded by the second element;

determining a second symbol value by making a hard decision for each bit b i of the second element; and

using the second symbol value to determine soft metrics L(b i ) for each bit b i encoded by the first element.

7. In a MIMO communication system, a reduced complexity method for determining a soft metric L(b i ) for each bit b i encoded by a complex vector x, wherein the soft metrics indicate discrepancies in the MIMO communication system, the method comprising:

receiving via a communication channel, a sample represented by a complex vector y;

determining a vector estimate of vector x using the received vector y, a first element and a second element being estimated from the vector estimate;

making a hard decision on values for each bit b i of the first element;

using the values of the first element to determine a soft metric L(b i ) for each bit b i encoded by the second element;

making a hard decision on values for each bit b i of the second element; and

using the values of the second element to determine a soft metric L(b i ) for each bit b i encoded by the first element.

8. The method of claim 7 wherein the complex vector x is K-dimensional, the method further comprising:

prior to the act of receiving, transmitting a sample of a multi-dimensional symbol, wherein the multi-dimensional symbol is represent-able as the complex, K-dimensional vector x, K being an integer corresponding to the number of antennas in the MIMO system, wherein each vector element of the vector x represents a signal transmitted with one of the K transmit antennas.

9. The method of claim 8 wherein the complex vector y is N-dimensional, N being an integer corresponding to the number of receive antennas in the MIMO system, and wherein vector y is related to vector x according to an equation: y=Hx+n, where H is an N×K channel matrix modeling phase shifts and amplitude changes on paths between the K transmit and N receive antennas, n is a complex N-dimensional vector representing communication channel noise.

10. The method of claim 9 wherein the act of determining an estimate of vector x further comprises:

determining a scalar number z according to the equation:

z

=

h

1

*

⁢

y

h

1

*

⁢

h

1

=

h

1

*

⁢

y



h

1



2

where h 1 * is the complex conjugate transpose of the first column of channel matrix H.

11. The method of claim 10 wherein the estimate of the first element of the transmitted vector x is formed according to the equation:

x

^

1

=

z

-

h

1

*

⁢

h

2

⁢

x

2

-

…

-

h

1

*

⁢

h

K

⁢

x

K

h

1

*

⁢

h

1

,

wherein

x

^

1

=

z

-

h

1

*

⁢

h

2

⁢

x

2

-

…

-

h

1

*

⁢

h

K

⁢

x

K

h

1

*

⁢

h

1

is rounded to a constellation point nearest to transmitted vector x to get {circumflex over (x)} 1 .

12. A system for determining soft metrics in a multiple-input multiple-output (MIMO) communication system, wherein a soft metric is determined for each bit, b i , encoded by a multi-dimensional symbol, wherein the soft metrics indicate discrepancies in the MIMO communication system, the system comprising:

a transmitter having one or more software instructions for transmitting a sample of the multi-dimensional symbol, wherein the multi-dimensional symbol is represent-able as a complex, K-dimensional vector x, K being an integer corresponding to the number of transmit antennas in the MIMO system, wherein each vector component of the vector x represents a signal transmitted with one of the K transmit antennas, the multi-dimensional symbol being transmitted through a communication channel;

a receiver having one or more software instructions for receiving from the communication channel, a sample represented by a complex, N-dimensional vector y, N being an integer corresponding to the number of receive antennas in the MIMO system, and wherein vector y is related to vector x according to an equation: y=Hx+n, where H is an N×K channel matrix modeling phase shifts and amplitude changes on paths between the K transmit and N receive antennas, n is a complex N-dimensional vector representing communication channel noise; and

one or more software instructions for determining a soft metric L(b i ) for each bit b i encoded by x according to the equation: L(b i )=σ −2 ·(min x j |b i =−1 ∥y−Hx j ∥ 2 −min x j |b i =+1 ∥y−Hx j ∥ 2 ) where x j represents the j-th possible value for x and where σ is a standard deviation of noise/interference in a received signal.

13. In a MIMO communication system, a reduced complexity method for determining a soft metric L(b i ) for each bit b i encoded by a complex vector x, wherein the soft metrics indicate discrepancies in the MIMO communication system, the method comprising:

receiving via a communication channel, a sample represented by a complex vector y;

determining a vector estimate of vector x using the received vector y, a first element and a second element being estimated from the vector estimate;

making a hard decision on values for each bit b i of the first element, as well as a conventional soft decision on

x

^

1

=

z

-

h

1

*

⁢

h

2

⁢

x

2

-

…

-

h

1

*

⁢

h

K

⁢

x

K

h

1

*

⁢

h

1

to get a soft metric for the first element where h 1 * is the complex conjugate transpose of a selected column of channel matrix H; and

using the values of the first element to determine a soft metric L(b i ) for each bit b i encoded by the second element.

14. In a MIMO communication system, a method for determining a soft metrics L(b i ) for each bit b i encoded by a complex vector x, wherein the soft metrics indicate discrepancies in the MIMO communication system, the method comprising:

receiving via a communication channel, a sample represented by a complex vector y, where complex vector y is related to the complex vector x; and

using complex vector y to determine soft metric L(b i ) for each bit b i encoded by x according to the equation: L(b i )=(min x j |b i =−1 ∥(y−Hx j )R −1 (y−Hx j )*∥ 2 −min x j |b i =+1 ∥(y−Hx j )R −1 (y−Hx j )*∥ 2 ) where H is a channel matrix modeling phase shifts and amplitude changes for the communication channel; and x j represents no more than 2 B(K−1) possible values for x, where B is the number of bits transmitted per symbol, and R −1 is the inverse of a spatial noise/interference auto-covariance matrix.

15. A method for determining soft metrics for each bit encoded by a multi-dimensional symbol in a multiple-input multiple-output (MIMO) communication system having K transmit antennas and N receive antennas with K being an integer corresponding to the number of transmit antennas and K being greater than one and N being an integer corresponding to the number of receive antennas, wherein the soft metrics indicate discrepancies in the MIMO communication system, the method comprising:

transmitting a signal corresponding to the multi-dimensional symbol through a communication channel, the multi-dimensional symbol being a symbol representable as a complex, K-dimensional vector x wherein each vector component of the vector x represents a signal transmitted with one of the K transmit antennas and represents one or more bits, b i ;

receiving, from the communication channel, samples represented by a complex, N-dimensional vector, y, wherein the vector y is related to the vector x according to an equation y=Hx+n, where H is an N×K channel matrix that models phase shifts and amplitude changes on paths between the K transmit antennas and the N receive antennas and where n is a complex N-dimensional vector representing channel noise over the communication channel; and

determining a soft metric L(b i ) for each bit b i encoded by the vector x from measurements of the vector y and the matrix H.

16. The method of claim 15 , wherein the soft metric is a maximum a posteriori probability metric.

17. The method of claim 15 , wherein the soft metric is a metric derived from an output of a log MAP decoder.

18. The method of claim 15 , wherein determining the soft metric comprises determining soft metrics that maximize receive diversity order.

19. The method of claim 15 , wherein determining the soft metric comprises computing a likelihood metric by calculating a difference of minimum distances between the received symbol represented by the vector y and symbols from the set of all possible symbols for each combination of Hx j where x j represents the j-th possible value for x.

20. The method of claim 15 , wherein the soft metric L(b i ) is a scaled likelihood metric.

21. The method of claim 20 , wherein the scaled likelihood metric for L(b i ) is computed according to an equation L(b i )=σ −2 ·(min x j |b i =−1 ∥y−Hx j ∥ 2 −min x j |b i =+1 ∥y−Hx j ∥ 2 ) where σ is a standard deviation of noise/interference in a received signal and x j represents the j-th possible value for x.

22. The method of claim 15 , wherein transmitting comprises one or more of bit interleaving and symbol interleaving and receiving comprises one or more of de-interleaving hard metrics and de-interleaving soft metrics.

23. The method of claim 15 , further comprising encoding the signal, prior to transmitting, using forward error correction (FEC) and decoding the signal, after receiving, using a Viterbi decoding process.

24. The method of claim 15 , further comprising interference processing that includes statistics of interference included in log MAP metrics.

25. A method for determining soft metrics for each bit encoded by a multi-dimensional orthogonal frequency division multiplexing (OFDM) symbol in a multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) communication system having K transmit antennas and N receive antennas with K being an integer corresponding to the number of transmit antennas and K being greater than one and with N being an integer corresponding to the number of receive antennas, wherein the soft metrics indicate discrepancies in the MIMO-OFDM communication system, the method comprising:

transmitting a signal corresponding to the multi-dimensional OFDM symbol through a communication channel, a multi-dimensional symbol being a symbol representable as a complex, K-dimensional vector x wherein each vector component of the vector x represents a signal transmitted with one of the K transmit antennas and represents one or more bits, b i , for a single tone of an OFDM waveform;

receiving, from the communication channel, samples represented by a complex, N-dimensional vector, y, wherein the vector y is related to the vector x according to an equation y=Hx+n where H is an N×K channel matrix that models frequency domain phase shifts and amplitude changes on paths between the K transmit antennas and the N receive antennas and where n is a complex N-dimensional vector representing channel noise over the communication channel; and

determining a soft metric L(b i ) for each bit b i encoded by the vector x from measurements of the vector y and the matrix H.

26. The method of claim 25 , wherein the soft metric is a maximum a posteriori probability metric.

27. The method of claim 25 , wherein the soft metric is a metric derived from an output of a log MAP decoder.

28. The method of claim 25 , wherein determining the soft metric comprises determining soft metrics that maximize receive diversity order.

29. The method of claim 28 , wherein determining the soft metric comprises computing a likelihood metric by calculating a difference of minimum distances between the received symbol represented by the vector y and symbols form the set of all possible symbols for each combination of Hx j where x j represents the j-th possible value for x.

30. The method of claim 25 , wherein the soft metric L(b i ) is a scaled likelihood metric.

31. The method of claim 30 , wherein the scaled likelihood metric L(b i ) is computed according to an equation L(b i )=σ −2 ·(min x j |b i =−1 ∥y−Hx j ∥ 2 −min x j |b i =+1 ∥y−Hx j ∥ 2 ) where σ is a standard deviation of noise/interference in a received signal and x j represents the j-th possible value for x.

32. The method of claim 25 , wherein the orthogonal frequency division multiplexing is OFDM modulation according to IEEE 802.11 standards.

33. The method of claim 25 , wherein transmitting comprises one or more of bit interleaving and symbol interleaving and receiving comprises one or more of de-interleaving hard metrics and de-interleaving soft metrics.

34. The method of claim 25 , further comprising encoding the signal, prior to transmitting, using forward error correction (FEC) and decoding the signal, after receiving, using a Viterbi decoding process.

35. The method of claim 25 , further comprising interference processing that includes statistics of interference included in log MAP metrics.