Method and apparatus for multiuser-interference reduction

• US 5,761,237 A
• Filed: 06/27/1996
• Issued: 06/02/1998
• Est. Priority Date: 02/10/1994
• Status: Expired due to Term

• Current Assignee: GlobalFoundries, Inc.
• Original Assignee: International Business Machines Corporation
• Inventors: Petersen, Brent R., Evangelos, Eleftheriou
• Primary Examiner(s): Chin, Stephen
• Assistant Examiner(s): GLUCK, JEFFREY WILLIAM
• Application Number:
  US08/666,457
• Time in Patent Office: 705 Days
• Field of Search: 375/200, 375/206, 375/207, 375/229, 375/230, 375/232, 375/233, 375/234, 375/348, 375/350, 370/320, 370/335, 370/342, 370/441, 370/479, 364/724.2
• US Class Current: 375/148
• CPC Class Codes: H04B 7/2628 using code-division multiple access [CDMA] or spread spectrum multiple access [SSMA] H04B7/2618 takes precedence
• CPC Class Codes: H04L 25/03044 using fractionally spaced delay lines or combinations of fractionally integrally spaced taps
• CPC Class Codes: Y02D 30/70 in wireless communication networks

Abstract

The present invention concerns an apparatus and method for reducing the multiuser-interference of input signals. The apparatus in accordance with the present invention comprises a multivariate predictor (81) and a decision quantizer (82), said multivariate predictor (81) operating on interference signals η'"'"'(D) provided by means for extracting interference signals (83), said interference signals η'"'"'(D) being obtained from said input signals and output signals b(D) which are available at an output of said decision quantizer (82) and fed back from there to said means for extracting interference signals (83).

Claims

1. An apparatus for reducing the multiuser-interference of input signals x(D) in a CDMA communication system, said apparatus comprising:

a multivariate predictor;

a decision quantizer;

a means for extracting interference signals η

'"'"'(D), said interference signals η

'"'"'(D) obtained from said input signals x(D) and from output signals b(D) from said decision quantizer;

wherein said multivariate predictor operates on said interference signals η

'"'"'(D) provided by the means for extracting interference signals with said multivariate predictor feeding its output signals η

(D) into means for combining with said input signals x(D).

2. The apparatus of claim 1, further comprising

a matrix of forward filters for a first interference reduction by filtering input signals x(D).

3. The apparatus of claim 1, further comprising

a bank of fractional-chip spaced filters for a first interference reduction by filtering input signals x(D).

4. The apparatus of claim 1, wherein said means for generating interference signals subtracts said output signals b(D) from said input signals x(D).

5. The apparatus of claim 1, wherein said multivariate predictor comprises a network of K×

K T-spaced filters.

6. The apparatus of claim 5, wherein said multivariate predictor comprises a means for subtracting the multivariate predictor output signals η

(D), provided at the outputs of said network of K×

K T-spaced filters, from said input signals x(D).

7. The apparatus of claim 2, wherein each forward filter of said matrix is either

a zero-forcing (ZF) linear equalizer, or

a minimum-mean-square error (MMSE) linear equalizer.

8. The apparatus of claim 1, wherein said decision quantizer is a Viterbi decoder for decoding convolutionally encoded data.

9. The apparatus of claim 1,2,3,4,5,6,7 or 8, further comprising a means for determining prediction errors e(D) said prediction errors used for updating the prediction coefficients of said multivariate predictor.

10. The apparatus of claim 1 further comprising means to transmit and receive TDMA-based or FDMA-based traffic.

11. A method for reducing the multiuser-interference of input signals x(D) in a CDMA communication system, comprising the steps:

extracting interference signals η

'"'"'(D) from said input signals x(D) by using output signals b(D) of a decision quantizer,

generating multivariate predictor output signals η

(D) by taking said interference signals η

'"'"'(D) as inputs, and

reducing the noise and residual interference signals of said input signals x(D) by subtracting said multivariate predictor output signals η

(D) from said input signals x(D).

12. The method of claim 11, wherein a first interference reduction step is carried out prior to the step of extracting interference signals η

'"'"'(D).