Device and method for combining precoding with symbol-rate spectral shaping

• US 5,249,200 A
• Filed: 07/30/1991
• Issued: 09/28/1993
• Est. Priority Date: 07/30/1991
• Status: Expired due to Term

• Current Assignee: CIF Licensing LLC
• Original Assignee: Codex Corporation (Motorola Solutions, Inc.)
• Inventors: Chen, Michael, Eyuboglu, M. Vedat
• Primary Examiner(s): Kuntz, Curtis
• Assistant Examiner(s): Bocure, Tesfaldet
• Application Number:
  US07/737,874
• Time in Patent Office: 791 Days
• Field of Search: 375/18, 375/37, 375/39, 375/57, 375/58, 375/11, 374/43, 374/45, 332/159, 332/107, 329/349, 455/50
• US Class Current: 375/285
• CPC Class Codes: H04L 2025/03808 Transmission of equaliser coefficients
• CPC Class Codes: H04L 25/03343 Arrangements at the transmitter end
• CPC Class Codes: H04L 25/497 by correlative coding, e.g. partial response coding or echo modulation coding transmitters and receivers for partial response systems transversal equalizers at the transmitter end H04L25/03343
• CPC Class Codes: H04L 25/4975 Correlative coding using Tomlinson precoding, Harashima precoding, Trellis precoding or GPRS
• CPC Class Codes: H04L 27/3422 in which the constellation is not the n - fold Cartesian product of a single underlying two-dimensional constellation

Abstract

The device and method provide precoding and symbol-rate transmitter spectral shaping of signals representative of digital information. The invention improves reliability of reception in the presence of non-Gaussian channel noise. The present invention also provides a flexible way of choosing between linear equalization, precoding without spectral shaping, and precoding with spectral shaping.

Claims

1. A device for providing symbol-rate transmit spectral shaping of a digital information sequence in a digital transmission system, comprising at least one of:

modulation means for at least modulating the digital information sequence; and

demodulation means for at least demodulating the modulated digital information sequence to provide an estimate of the digital information sequence;

wherein;

the modulation means comprises at least;

a precoding means operably coupled to receive the digital information sequence, for precoding said sequence utilizing a predetermined monic equivalent channel response h(D) to generate a precoded sequence;

a symbol-rate spectral shaping means with a monic filter response a(D)≠

1, operably coupled to the precoding means, for utilizing a spectral shaping filtering function and a first scaling factor to provide a spectrally shaped sequence of precoded symbols; and

a transmission preparation means operably coupled to the symbol-rate spectral shaping means, for generating a spectrally-shaped transmission sequence for transmission on a selected channel;

and the demodulation means comprises at least;

an equalizing means operably coupled to receive the spectrally-shaped transmission sequence, for providing an equalized spectrally-shaped sequence of precoded symbols; and

a symbol-rate processing means, operably coupled to the equalizing means, comprising;

noise prediction means, operably coupled to receive the equalized spectrally-shaped sequence of precoded symbols, for substantially utilizing a symbol-rate noise prediction filter with response b(D)=h(D)/a(D) and at least a second scaling factor to provide a noise prediction filter output; and

decoding means, operably coupled to the symbol-rate noise prediction means, for providing an estimate of the digital information sequence.

2. The device of claim 1, wherein the predetermined equivalent channel response h(D) is a product of the noise prediction filter with response b(D) and the spectral shaping filter a(D).

3. The device of claim 2, wherein the noise prediction filtering function is a second order filter function of a form b(D)=1+b₁ D+b₂ D² wherein b₁ and b₂ are selected coefficients.

4. The device of claim 1, wherein the noise prediction filter with response b(D) is a first order filter of a form b(D)=1+b₁ z^-1 having a noise prediction coefficient b₁ that is substantially equal to ##EQU8## where g is a noise pole and r denotes a magnitude of g.

5. The device of claim 1, wherein the predetermined equivalent channel response h(D) is substantially one, and the spectral shaping filter with response a(D) is substantially an inverse of the noise prediction filter with response b(D).

6. The device of claim 1 wherein the equalizing means comprises substantially a linear equalizer.

7. The device of claim 1 wherein the first scaling factor is substantially equal to 1/√

A wherein A is the sum of the squares of the coefficients of a(D) and the second scaling factor is substantially equal to √

A.

8. A device in a modem for providing symbol-rate transmit spectral shaping of a digital information sequence in a digital transmission system, comprising at least:

modulation means for at least modulating the digital information sequence; and

demodulation means for at least demodulating the modulated digital information sequence to provide an estimate of the digital information sequence;

wherein;

the modulation means comprises at least;

a precoding means operably coupled to receive the digital information sequence for precoding said sequence utilizing a predetermined monic equivalent channel response h(D) to generate a precoded sequence;

a symbol-rate spectral shaping means with a monic filter response a(D)≠

1, operably coupled to and responsive to the precoding means, for utilizing a spectral shaping filtering function and a first scaling factor to provide a spectrally shaped sequence of precoded symbols; and

a transmission preparation means operably coupled to and responsive to the symbol-rate spectral shaping means, for generating a spectrally-shaped transmission sequence for transmission on a selected channel;

and the demodulation means comprises at least;

an equalizing means operably coupled to receive the spectrally-shaped transmission sequence, being responsive to the transmission preparation means, for providing an equalized spectrally-shaped sequence of precoded symbols; and

a symbol-rate processing means, operably coupled to and responsive to the equalizing means, comprising;

noise prediction means, responsive to the equalizing means, being operably coupled to receive the equalized spectrally-shaped sequence of precoded symbols for substantially utilizing a symbol-rate noise prediction filter with response b(D)=h(D)/a(D) and at least a second scaling factor to provide a noise prediction filter output; and

decoding means, operably coupled to and responsive to the symbol-rate noise prediction means, for providing an estimate of the digital information sequence.

9. The device of claim 8, wherein the predetermined equivalent channel response h(D) is substantially a product of the noise prediction filter with response b(D) and the spectral shaping filter with response a(D).

10. The device of claim 9, wherein the noise prediction filtering function is a second order filter function, substantially a transfer function of a form b(D)=1+b₁ D+b₂ D² wherein b₁ and b₂ are selected coefficients.

11. The device of claim 8, wherein the noise prediction filter with response b(D) is a first order filter with a response of a form b(D)=1+b₁ D and having a noise prediction coefficient b₁ that is substantially equal to ##EQU9## where g is a noise pole and r denotes a magnitude of g.

12. The device of claim 8, wherein the predetermined equivalent channel response h(D) is substantially one, and the spectral shaping filter with response a(D) is substantially an inverse of the noise prediction filter with response b(D).

13. The device of claim 8 wherein the equalizing means comprises substantially a linear equalizer.

14. The device of claim 8 wherein the first scaling factor is substantially equal to 1/√

A wherein A is the sum of the squares of the coefficients of a(D) and the second scaling factor is substantially equal to √

A.

15. The device of claim 8, wherein the predetermined equivalent channel response h(D) is substantially one, and the spectral shaping filter with response a(D) is substantially an inverse of the noise prediction filter with response b(D).

16. The device of claim 8, wherein at least part of one of:

the precoding means;

the spectral shaping means;

the transmission preparation means;

the equalizing means; and

the symbol-rate processing means,

is implemented utilizing a computer program storage medium having a computer program to be executed by a digital computer stored thereon, the computer program comprising at least one of;

first unit for converting information into a selected number of bits to provide encoded symbols;

second unit for utilizing the encoded symbols to generate precoded, encoded symbols, termed a first symbol;

third unit for substantially spectrally shaping the first symbol to obtain substantially a shaped symbol and for transmitting the shaped symbol, where desired;

fourth unit for receiving information symbols, for at least utilizing an equalizing unit to generate equalized filtered, received information symbols;

fifth unit for utilizing a symbol-rate processing unit to obtain selectably filtered, received noise predicted information symbols;

sixth unit for determining substantially the estimated information data input symbol sequence for the received information symbols; and

a

seventh unit for substantially obtaining estimated information bits of information input.

17. The device of claim 8 wherein at least part of one of:

the precoding means;

the spectral shaping means;

the transmission preparation means;

the equalizing means; and

the symbol-rate processing means, is implemented utilizing at least a first digital signal processor configured to provide at least one of;

first unit for converting information into a selected number of bits to provide encoded symbols;

second unit for utilizing the encoded symbols to generate precoded, encoded symbols, termed a first symbol;

third unit for substantially spectrally shaping the first symbol to obtain substantially a shaped symbol and for transmitting the shaped symbol, where desired;

fourth unit for receiving information symbols, for at least utilizing an equalizing unit to generate equalized filtered, received information symbols;

fifth unit for utilizing a symbol-rate processing unit to obtain selectably filtered, received noise predicted information symbols;

sixth unit for determining substantially the estimated information data input symbol sequence for the received information symbols; and

a

seventh unit substantially obtaining estimated information bits of information input.

18. A method for providing symbol-rate transmit spectral shaping of a digital information sequence in a digital transmission system, comprising at least one of the sets of steps:

(A) precoding the digital information sequence utilizing a predetermined monic equivalent channel response h(D) to generate a precoded sequence;

utilizing a symbol-rate spectral shaping filter with a monic filter response a(D)≠

1, and a first scaling factor to provide a spectrally shaped sequence of precoded symbols;

generating a spectrally-shaped sequence for transmission on a selected channel; and

(B) utilizing an equalization unit to provide an equalized spectrally-shaped sequence of said precoded symbols; and

utilizing a symbol-rate noise prediction filter with response b(D)=h(D)/a(D) and at least a second scaling factor to provide a noise prediction filter output, and providing an estimate of the digital information sequence.

19. The method of claim 18, wherein, in the step of precoding, the predetermined equivalent channel response h(D) is a product of the noise prediction filter with response b(D) and the spectral shaping filter with response a(D).

20. The method of claim 19, wherein the noise prediction filtering function is a second order filter function of a form b(D)=1+b₁ D+b₂ D² wherein b₁ and b₂ are selected coefficients.

21. The method of claim 19, wherein the predetermined equivalent channel response h(D) is substantially one, and the spectral shaping filter with response a(D) is substantially an inverse of the noise prediction filter with response b(D).

22. The method of claim 18, wherein the noise prediction filter with response b(D) is a first order filter of a form b(D)=1+b₁ z^-1 having a noise prediction coefficient b₁ that is substantially equal to ##EQU10## where g is a noise pole and r denotes a magnitude of g.