Technique for adaptive multiuser equalization in code division multiple access systems

• US 20050207477A1
• Filed: 03/16/2005
• Published: 09/22/2005
• Est. Priority Date: 03/16/2004
• Status: Active Grant

• Current Assignee: Peter Monsen
• Original Assignee: Peter Monsen
• Inventors: Monsen, Peter
• Granted Patent: US 7,463,672 B2
• Time in Patent Office: Days
• US Class Current: 375/147
• CPC Class Codes: H04B 1/7105 Joint detection techniques, e.g. linear detectors
• CPC Class Codes: H04B 1/7107 Subtractive interference cancellation
• CPC Class Codes: H04B 2201/70702 Intercell-related aspects
• CPC Class Codes: H04B 7/08 at the receiving station

Abstract

A radio communication method and system for transmitting multiple-user digital data information in a Code Division Multiple Access (CDMA) format over multipath and mutual-interfering channels to a receiving terminal with diversity antennas and equalization signal processing. At the receiving terminal, diversity antenna signals are grouped into time blocks and replicas are generated of the user-unique CDMA sequence signals for a set of K mutually interfering users. Within each time block the multiuser channel is estimated, equalization parameters are calculated, and decision-feedback equalization is used to produce multiuser estimates associated with a subset λ of the K users. These estimates are deinterleaved and error-correction decoded to recover transmitted digital data information for the subset of λ users. The invention includes embodiments for a CDMA sequence period that is equal to or longer than the transmitted data symbol interval with applications to current CDMA reverse link and forward link standards. Also embodiments for both decision-directed and reference-directed channel estimation are given.

Claims

1. A method for receiving digital data information that is user-unique and is transmitted from a plurality of user transmitting terminals over multipath and multiple-access interference channels to a receiving terminal with one or more diversity antennas containing associated received diversity signals, wherein the user transmitting terminal adds error-correcting coding and interleaving to the digital data information to provide a sequence of digital data symbols at a user symbol rate that is the reciprocal of a user integer multiple of a system period and subsequently multiplies the digital data symbols by a user-unique direct-sequence spread spectrum (DSSS) signal to provide a modulated signal for transmission to the receiving terminal, and wherein the receiving terminal produces within a data frame successive user decisions corresponding to transmitted digital data symbols for a set of λ

detected users, said receiving method comprising the steps at the receiving terminal of;

time-sampling and grouping received diversity signals into a receiver time block that has a time duration that is no more than one-half of the interleaving delay;

producing DSSS replicas of the user-unique DSSS signals for K=μ

+λ

mutually-interfering user signals where K>

1 and μ

is equal to the number of interfering users that are not detected;

estimating a set of channel parameters associated with the receiver time block;

formatting the DSSS replicas to provide a set of sequence signals;

sequence combining the received diversity signals and sequence signals to provide sequence matched signals;

channel combining the sequence matched signals and channel parameters to provide channel matched signals, which depend on a subdecision interval that corresponds to a digital data subsymbol equal in duration to the system period and which further depend on subsymbol delay offsets that are relative to the subdecision interval;

processing the channel parameters and sequence signals to provide forward and backward filter parameters;

creating past interference signals that depend on the subdecision interval;

subtracting the past interference signals from the channel matched signals to provide equalization signals that depend on the subdecision interval and subsymbol delay offsets;

forward combining the equalization signals and the forward filter parameters to provide λ

user subsymbol estimates for successive subdecision intervals within the data frame;

summing the user integer multiple successive subsymbol estimates to provide a user symbol estimate of the transmitted user digital data symbol;

selecting, at the user symbol rate for users in the set of λ

detected users, the user decision as the possibly transmitted digital data symbol that is closest to the user symbol estimate;

partitioning the user decisions into subdecisions associated with subdecision intervals;

symbol deinterleaving and error-correcting decoding user symbol estimates to recover digital data information; and

wherein the creating step further comprises;

multiplying the backward filter parameters and subdecisions from previous subdecision intervals to provide the past interference signals.

2. The method of claim 1 wherein the user transmitting terminals are geographically separate and grouped within cell regions associated with central stations and wherein the receiving terminal is located at a central station and wherein the number of detected users λ

is greater than the number of users in the cell region associated with the receiving terminal central station.

3. The method of claim 2 wherein the number μ

of interfering users not detected is zero and wherein the equalization signals in the subtracting step further include;

K present equalization signals with no symbol delay; and

K future equalization signals with a symbol delay of minus one

4. The method of claim 1 wherein the user transmitting terminals are collocated at a plurality of central stations and wherein the receiving terminal is located within a cell region of an associated central station and wherein the number of mutually-interfering user signals K is greater than the number of collocated user transmitting terminals at the associated central station.

5. The method of claim 4 wherein the number μ

of interfering users not detected is greater than zero and wherein the equalization signals in the subtracting step further include;

μ

past equalization signals with a symbol delay of one symbol;

K present equalization signals with no symbol delay; and

K future equalization signals with a symbol delay of minus one symbol.

6. The method of claim 1 wherein the DSSS signals associated with the user transmitting terminals employ a time period that is longer than the system period and wherein the processing step provides successive sets of forward and backward filter parameters for successive subdecision intervals.

7. The method of claim 6 wherein the processing step further includes the steps of:

calculating a correlation matrix as the scaled sum of an identity matrix and a delta matrix;

initializing the column of the delta matrix associated with the detected user as the first order term of a forward filter vector;

iteratively computing the forward filter vector nth order term as the product of the delta matrix and the (n−

1) th term;

combining a unit vector and n, n>

1, forward filter vector terms in a power series expansion to provide the forward filter vector.

8. The method of claim 1 wherein the user transmitting terminals are collocated at a plurality of central stations and wherein the receiving terminal is located within a cell region of an associated central station and wherein the number of detected users λ

is equal to unity and wherein the μ

interfering users, μ

>

0, that are not detected have transmitted signals from central stations associated with adjacent cell regions.

9. The method of claim 8 wherein the DSSS signals for the user transmitting terminals collocated at a central station are mutually orthogonal.

10. The method of claim 1 wherein the processing step further includes the steps of:

calculating a correlation matrix;

inverting the correlation matrix to provide an inverse matrix;

selecting a column of the inverse matrix as a forward filter vector for one of the λ

detected users.

11. The method of claim 10 wherein the forward combining step further includes the step of:

forming successive dot products of the forward filter vector and equalization signals to provide successive subsymbol estimates for one of the λ

detected users.

12. The method of claim 10 wherein the inverting step further includes the step of:

decomposing the correlation matrix with a Cholesky decomposition.

13. The method of claim 12 wherein the decomposing step further includes the steps of:

generating a level structure;

computing a permutation matrix;

premultiplying and post-multiplying the correlation matrix by the permutation matrix to provide a modified correlation matrix, that has nonzero elements concentrated in a bandwidth centered on the diagonal;

truncating the Cholesky decomposition to the nonzero element bandwidth.

14. The method of claim 1 wherein the received diversity signals in the time-sampling and grouping step are time sampled at an integer multiple of the DSSS symbol rate where the integer is greater than unity.

15. The method of claim 1 wherein the estimating step further includes the steps of:

modulating user decisions from the previous data frame with the associated DSSS replica to provide a set of reference replicas;

correlating received diversity signals from the previous receiver time block and the reference replicas to provide coarse channel parameters;

reference processing the reference replicas to provide a correction matrix;

multiplying the correction matrix and coarse channel parameters to provide the channel parameters for the present receiver time block.

16. The method of claim 1 wherein the user transmitting terminal further includes transmission of reference data that is known at the receiver;

and wherein the estimating step further includes the steps of;

producing a plurality of reference replicas that are each associated with the reference data of a user transmitting terminal;

correlating received diversity signals and the reference replicas to provide coarse channel parameters;

reference processing the reference replicas to provide a correction matrix;

multiplying the correction matrix and coarse channel parameters to provide the channel parameters.

17. The method of claim 1 wherein a plurality of central stations include pilot transmitting terminals for transmission of reference data that is unique to the central station and known at the receiver;

and wherein the estimating step further includes the steps of;

producing a plurality of reference replicas that are each associated with the reference data of a pilot transmitting terminal;

correlating received diversity signals and the reference replicas to provide coarse channel parameters;

reference processing the reference replicas to provide a correction matrix;

multiplying the correction matrix and coarse channel parameters to provide the channel parameters.

18. For use in a multiuser communication system wherein digital data information that is user-unique is transmitted from a plurality of user transmitting terminals, that may be geographically separate or collocated at a central station, over multipath and multiple-access interference channels, and wherein the user transmitting terminal adds error-correcting coding and interleaving to the digital data information to provide a sequence of digital data symbols at a user symbol rate that is the reciprocal of a user integer multiple of a system period and subsequently multiplies the digital data symbols by a user-unique direct-sequence spread spectrum (DSSS) signal to provide a modulated signal for transmission, a receiver apparatus, that is associated with one or more diversity antennas containing associated received diversity signals, for producing within a data frame successive user decisions corresponding to transmitted digital data symbols for a set of λ

detected users, comprising;

converters for time-sampling and grouping received diversity signals into a receiver time block that has a time duration that is no more than one-half of the interleaving delay;

DSSS replica generator for producing user-unique DSSS signals for K=μ

+λ

mutually-interfering user signals where K>

1 and μ

is equal to the number of interfering users that are not detected;

channel estimator for estimating channel parameters associated with the receiver time block;

sequence signal generator for formatting the DSSS replicas into a set of sequence signals;

sequence matched filter that combines the received diversity signals and sequence signals to provide a set of sequence matched signals;

channel matched filter that combines the sequence matched signals and channel parameters to provide a set of channel matched signals, which depend on a subdecision interval that corresponds to a digital data subsymbol equal in duration to the system period and which further depend on subsymbol delay offsets that are relative to the subdecision interval;

equalizer calculator that computes a set of forward and backward filter parameters from the channel parameters and sequence signals;

backward filter that generates past interference signals and subtracts the past interference signals from the channel matched signals to provide equalization signals that depend on the subdecision interval and subsymbol delay offsets;

forward filter that combines the equalization signals and the forward filter parameters to provide λ

user subsymbol estimates for successive subdecision intervals within the data frame;

multiuser quantizer that sums the user integer multiple successive subsymbol estimates to provide a symbol estimate of the transmitted user digital data symbol, and further selects, at the user symbol rate for users in the set of λ

detected users, the user decision as the possibly transmitted digital data symbol that is closest to the user symbol estimate, and further partitions the user decisions into subdecisions associated with subdecision intervals;

symbol deinterleaver and error-correcting decoder that processes user symbol estimates to recover digital data information; and

wherein the backward filter further comprises;

means that multiply the backward filter parameters and subdecisions from previous subdecision intervals to provide the past interference signals.

19. A receiver apparatus according to claim 18 wherein the number μ

of interfering users not detected is zero and wherein the backward filter equalization signals further include;

K present equalization signals with no symbol delay; and

K future equalization signals with a symbol delay of minus one

20. A receiver apparatus according to claim 18 wherein the number μ

of interfering users not detected is greater than zero and wherein the backward filter equalization signals further include;

μ

past equalization signals with a symbol delay of one symbol;

K present equalization signals with no symbol delay; and

K future equalization signals with a symbol delay of minus one symbol.

21. A receiver apparatus according to claim 18 wherein the DSSS signals associated with the plurality of user transmitting terminals have time periods that are longer than the system period and the forward and backward filter parameters computed in the equalizer calculator are produced for successive subdecision intervals.

22. A receiver apparatus according to claim 21 wherein the equalizer calculator further includes:

forward filter correlator means for calculating a correlation matrix;

means for calculating a delta matrix from the correlation matrix;

means for iteratively computing the forward filter vector from the delta matrix.

23. A receiver apparatus according to claim 18 wherein the equalizer calculator further includes:

forward filter correlator means for calculating a correlation matrix; and

inverting means for producing the correlation matrix inverse.

24. A receiver apparatus according to claim 23 wherein the inverting means further includes:

means for generating a level structure;

means for computing a permutation matrix and premultiplying and post-multiplying the correlation matrix by the permutation matrix to provide a modified correlation matrix, that has nonzero elements concentrated in a bandwidth centered on the diagonal; and

means for performing a Cholesky decomposition.

25. A receiver apparatus according to claim 23 wherein the equalizer calculator further includes:

forward filter vector generator means for selecting a column of the correlation matrix inverse as a forward filter vector for one of the λ

detected users.

26. A receiver apparatus according to claim 25 wherein the forward filter further includes:

means for forming successive dot products of the forward filter vector and equalization signals to provide the successive subsymbol estimates.

27. A receiver apparatus according to claim 18 wherein the received diversity signals in the converters are time sampled at an integer multiple of the DSSS symbol rate where the integer is greater than unity.

28. A receiver apparatus according to claim 18 wherein the receiver apparatus further comprises:

reference replica generator that modulates user decisions from the previous data frame with the associated DSSS replica to provide a set of reference replicas, and

wherein the channel estimator further includes;

means for correlating received diversity signals from the previous receiver time block and the reference replicas to provide coarse channel parameters;

means for processing the reference replicas to provide a correction matrix;

means for multiplying the correction matrix and coarse channel parameters to provide the channel parameters for the present receiver time block.

29. A receiver apparatus according to claim 18 wherein user transmitting terminals further include transmission of user-unique reference data that are known at the receiver;

and

wherein the receiver apparatus further comprises;

reference replica generator that provides a plurality of reference replicas associated with user transmitting terminals, and

wherein the channel estimator further includes;

means for correlating received diversity signals and the reference replicas to provide coarse channel parameters;

means for reference processing the reference replicas to provide a correction matrix;

means for multiplying the correction matrix and coarse channel parameters to provide the channel parameters.

30. A receiver apparatus according to claim 18 wherein a plurality of central stations include pilot transmitting terminals for transmission of station-unique reference data that are known at the receiver;

and wherein the receiver apparatus further comprises;

reference replica generator that provides a plurality of reference replicas associated with pilot transmitting terminals, and

wherein the channel estimator further includes;

means for correlating received diversity signals and the reference replicas to provide coarse channel parameters;

means for reference processing the reference replicas to provide a correction matrix;

means for multiplying the correction matrix and coarse channel parameters to provide the channel parameters.