Low complexity multi-channel modulation method and apparatus

• US 7,876,838 B2
• Filed: 06/17/2004
• Issued: 01/25/2011
• Est. Priority Date: 06/19/2003
• Status: Expired due to Fees

• Current Assignee: The University of Queensland, University of Sydney
• Original Assignee: The University of Queensland, University of Sydney
• Inventors: Collings, Iain Bruce, Clarkson, Ian Vaughan Lyndon
• Primary Examiner(s): Pathak; Sudhanshu C
• Application Number:
  US10/869,066
• Publication Number:
  US 20050047323A1
• Time in Patent Office: 2,413 Days
• Field of Search: 375/265, 375/260, 358/426.12, 708/404
• US Class Current: 375/260
• CPC Class Codes: H04L 27/2627 Modulators
• CPC Class Codes: H04L 27/2649 Demodulators

Abstract

A digital communications apparatus for processing a data vector includes a transmitter for communicating with a receiver across a channel having channel characteristics. A transmit vector is computed from the data vector and from values defining a basis for a transmit lattice. The transmit lattice defines a set of transmit symbols of length N and is related to an orthogonal receive lattice by the channel characteristics. The computation is performed according to an algorithm of order N complexity.

Claims

1. A digital communications apparatus for processing a data vector, said apparatus including a transmitter for communicating with a receiver across a channel having channel characteristics, said apparatus including:

at least one computational processor;

at least one memory device storing instructions, which, when executed by the at least one computational processor cause the apparatus to;

compute a transmit vector from the data vector and from values defining a basis for a transmit lattice, said transmit lattice defining a set of transmit symbols of length N and related to an orthogonal receive lattice by the channel characteristics, and

perform the computation according to an algorithm of no more than order N complexity.

2. An apparatus according to claim 1, wherein the instructions further cause the apparatus to determine the channel characteristics.

3. An apparatus according to claim 2, wherein the instructions further cause the apparatus to calculate the values defining the basis for the transmit lattice.

4. An apparatus according to claim 1, wherein the instructions further cause the apparatus to calculate the values defining the basis for the transmit lattice in order that the transmit symbols avoid clipping by the transmitter.

5. An apparatus according to claim 1, wherein the instructions further cause the apparatus to determine the basis for the transmit lattice from a decomposition of a channel matrix of the channel characteristics.

6. An apparatus according to claim 1, wherein the instructions further cause the apparatus to compute the transmit lattice points on the basis of the channel characteristics and a receive lattice having points spaced to correspond to a desired bit-error-rate (BER).

7. An apparatus according to claim 1 further comprising a converter configured to convert binary code data into the data vector.

8. An apparatus according to claim 7 wherein the converter comprises a Gray-code converter.

9. An apparatus according to claim 1 including a receiver to process a receive data vector.

10. An apparatus according to claim 9 wherein the receiver is configured to compute a filtered vector from the receive data vector and values defining a basis for the orthogonal receive lattice.

11. An apparatus according to claim 10, wherein the receiver is configured to compute the filtered vector according to an algorithm of order N complexity.

12. An apparatus according to claim 11, wherein the receiver is further configured to process the filtered vector to determine a maximum-likelihood estimate of the data vector.

13. An apparatus according to claim 12 further including a converter configured to convert elements of the maximum-likelihood estimate of the data vector to a binary code format.

14. An apparatus according to claim 13 wherein the converter comprises a binary-to-Gray-code converter.

15. A method for operating a processor of a digital communications apparatus to transmit digital data across a channel having channel characteristics, the method including the steps of:

calculating, by the processor, a basis for a transmit lattice, the transmit lattice defining a set of transmit symbols of length N and related to an orthogonal receive lattice by the channel characteristics,

wherein the calculation is performed by the processor according to an algorithm of no more than order N complexity.

16. A method according to claim 15, including calculating the receive lattice as a function of a desired bit-error-rate (BER).

17. A method according to claim 16, including calculating the basis for the transmit lattice in order that clipping is avoided in a transmitter operated by the digital communications apparatus.

18. A method for operating a processor of a digital communications apparatus for transmitting digital data across a channel characterised by a channel matrix H, the method including the steps of:

a) instructing the processor to divide the digital data into N subsequences each of length B_i for i=1, . . . , N;

b) instructing the processor to form a data vector q from the N subsequences;

c) instructing the processor to determine a basis U<sup>−

1</sup>A for a transmit lattice where U<sup>−

1</sup> is an upper triangular matrix determined from a decomposition of the channel matrix and A=diag(2<sup>−

B1</sup>, 2<sup>−

B2</sup>, . . . , 2<sup>−

BN</sup>); and

d) instructing the processor to iteratively calculate a vector x of transmit signals according to the equation

19. A method according to claim 18 including instructing the processor to calculate x_i in each iteration of step d) so that x falls within a pre-determined transmit space comprising a region of N-dimensional space.

20. A method according to claim 19 including selecting the pre-determined region of transmit space to avoid clipping.

21. A method according to claim 18 including steps for instructing the processor to process a receive vector y, said steps including:

e) instructing the processor to calculate a matrix R from the channel matrix H where H=QR and Q is an orthonormal matrix; and

f) instructing the processor to iteratively calculate a filtered receive vector s according to the equation

$s_i=\frac{1}{r_{\mathrm{ii}}}<br><br>{\mathrm{\mu <br><br>}}_i-\underset{j=1}{\overset{i-1}{\sum <br><br>}}<br><br><br><br>r_{\mathrm{ji}}<br><br>s_j$
where μ

=H^Ty.

22. A method according to claim 21 further including instructing the processor to calculate a maximum likelihood estimate {circumflex over (q)} of q by calculating the positive remainder of the division of {circumflex over (p)}_i by 2^Bi where {circumflex over (p)}_i=round(2^Bis_i/d_i) and “

round”

is a function which yields a nearest integer to its argument.

23. A method according to claim 22 including converting the q to estimates of the digital data of step a) by binary-to-Gray-code conversion.

24. A communications apparatus including a software or firmware product in the form of a memory device containing instructions readable by a processor of the communications apparatus to transmit digital data across a channel having channel characteristics the instructions including:

instructions executable by the processor to calculate a basis for a transmit lattice, the transmit lattice defining a set of transmit symbols of length N and related to an orthogonal receive lattice by the channel characteristics,

wherein the calculation is performed by the processor according to an algorithm of no more than order N complexity.

25. A communications apparatus according to claim 24 wherein the instructions include, instructions to calculate the receive lattice as a function of a desired bit-error-rate (BER).

26. A communications apparatus according to claim 24 wherein the instructions include, instructions to calculate the basis of the transmit lattice in order that clipping is avoided in a transmitter operated by the communications apparatus.

27. A communications apparatus according to claim 24 wherein the instructions to calculate the basis for the transmit lattice as a function of the channel characteristics and the receive lattice having predetermined dimensions include:

a) instructions for the processor to divide the digital data into N subsequences each of length B_i for i=1, . . . , N;

b) instructions for the processor to form a data vector q from the N subsequences;

c) instructions for the processor to determine a basis U<sup>−

1</sup>A for a transmit lattice where U<sup>−

1</sup> is an upper triangular matrix determined from a decomposition of the channel matrix and A=diag(2<sup>−

B1</sup>, 2<sup>−

B2</sup>, . . . , 2<sup>−

BN</sup>); and

d) instructions for the processor to iteratively calculate a vector x of transmit signals according to the equation

28. A communications apparatus according to claim 27 wherein said instructions include, instructions for the processor to calculate x_i in each iteration of step d) so that x falls within a pre-determined transmit space comprising a region of N-dimensional space.

29. A communications apparatus according to claim 28 wherein said instructions include, instructions for the processor to select the pre-determined region of transmit space to avoid clipping.

30. A communications apparatus according to claim 27 wherein said instructions include:

e) instructions for the processor to calculate a matrix R from the channel matrix H where H=QR and Q is an orthonormal matrix; and

f) instructions for the processor to iteratively calculate a filtered receive vector s according to the equation

31. A communications apparatus according to claim 30 further including instructions for the processor to calculate a maximum likelihood estimate {circumflex over (q)} of q by calculating the positive remainder of the division of {circumflex over (p)}_i by 2^Bi where {circumflex over (p)}_i=round(2^Bis_i/d_i) and “

round”

is a function which yields a nearest integer to its argument.

32. A communications apparatus according to claim 31 including instructions for the processor to convert the q to estimates of the digital data of step a) by binary-to-Gray-code conversion.