Apparatus and method for a DSSS/CDMA receiver

Details Apparatus and method for a DSSS/CDMA receiver Apparatus and method for a DSSS/CDMA receiver

1999-02-01

2002-04-23

Pham, Chi (Department: 2731)

Pulse or digital communications

Spread spectrum

Direct sequence

C375S142000, C375S143000, C375S144000, C375S148000, C375S150000, C375S152000, C370S335000, C370S342000, C370S320000, C370S350000, C370S441000, C370S479000, C708S300000, C708S320000, C708S322000, C708S323000

Reexamination Certificate

active

06377611

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of Invention
The present invention relates to receivers for DSSS/CDMA communication systems. More particularly, the present invention relates to receivers including an MMSE linear detector implemented using Griffiths' algorithm.
2. Description of the Related Art
Code-division multiple access (CDMA) is one of several methods of multiplexing wireless users. In direct sequence spread spectrum CDMA (DSSS/CDMA) communications systems, users are multiplexed by distinct spreading codes rather than by orthogonal frequency bands, as in frequency-division multiple access (FDMA), or by orthogonal time slots, as in time-division multiple access (TDMA). That is, all CDMA users transmit simultaneously over the same frequency spectrum. In a single-user detector, the conventional receiver operates as a matched correlator to detect the signal of interest by correlating the entire received signal with the particular user's code waveform. Since the received signal is composed of the sum of all users' signals plus additive white Gaussian noise (AWGN), the signals from other users will deteriorate the performance of the matched correlator. This fact results from the random time offsets between users, which make it impossible to design the code waveforms to be completely orthogonal. Multiple access interference (MAI) is thus a factor which limits the capacity and performance of the DSSS/CDMA systems.
The performance of the matched correlator is further degraded by imperfect power control. A nearby interfering user of large power will have a significant impact on the reception of a highly attenuated signal of interest. The so-called near-far problem can be mitigated by an optimal multiuser detector. The detector yields the most likely transmitted sequence of all users to maximize the probability that the estimated sequence was transmitted given the received signal extending over the whole message. This probability is referred to as the joint a posteriori probability, and the receiver is known as a maximum-likelihood sequence (MLS) detector. A problem of the MLS approach is that the implementation complexity is very high. Another disadvantage is that it requires knowledge of the received amplitudes and phases of all users. These values, however, are not known in advance and must be estimated. Therefore, suboptimal but near-far resistant receivers have been developed.
One suboptimal but near-far resistant receiver is a minimum mean-squared error,(MMSE) detector. The MMSE linear detector utilizes the cyclostationarity of the highly structured MAI to mitigate the near-far problem and has a moderate complexity in comparison with the matched correlator. For a particular user, the MMSE detector adopts a finite-impulse response (FIR) filter to estimate the transmitted signal. To mitigate the interference, the FIR filter is designed such that the mean-squared error between the transmitted signal and the estimate is minimized.
A DSSS/CDMA communication system including an MMSE linear detector is disclosed in “MMSE Interference Suppression for Direct-Sequence Spread-Spectrum CDMA” by U. Madhow et al., IEEE Trans. Commun., vol. 42, pp. 3178-3188, December 1994, which is incorporated herein by reference. In a system such as disclosed therein, the received signal vector r(j) of one symbol interval can be expressed by
r
⁡
(
j
)
=
∑
l
=
1
L
⁢
b
l
⁡
[
j
]
⁢
 
⁢
A
l
⁢
c
l
+
x
⁡
(
j
)
(
1
)
where b
l
[j]&egr;[−1, +1] is the transmitted symbol, A
l
is the received amplitude, c
l
&egr;R
N×1
is the signal srector of the lth user, L is the number of users and x(j) is the vector of AWGN samples. Taking the first user to be the desired transmission, the MMSE linear detector demodulates the symbol of interest as {circumflex over (b)}
l
[j]=sgn(w
T
r(j)) where the FIR filter tap-weight vector w minimizes the mean-squared error (MSE) between the desired symbol and the test statistic. The symbol “T” indicates the matrix transpose operation. The optimum tap-weight vector w
n
can be expressed as
w
o
=R
−1
p
  (2)
where R=E{r(j)r
T
(j)} and p=E{b
l
[j]r(j)} are the correlation matrix of the received signal vector and the cross-correlation vector between the desired signal and the received signal vector, respectively.
The tap-weights of the MMSE linear detector can be obtained by solving the Wiener-Hopf equation. However, the computation load is high since the calculation involves matrix inversion and multiplication. The MMSE linear detector can also be implemented by adaptive algorithms rather than from a direct solution involving matrix computation. The least mean square (LMS) algorithm and the recursive least squares algorithm have been used to train the MMSE detector on the assumption that a pilot signal is available. For practical applications, when ttie interference is time-varying or unknown, the MMSE linear detector can be implemented adaptively by the blind Griffiths' algorithm. The Griffiths' algorithm uses the desired signal vector instead of a training sequence of symbols for initial adaptation. When the desired signal vector is perfectly known and the step-size is small enough, the Griffiths' algorithm will drive the linear detector to converge to the Wiener-Hopf solution. “Blind Adaptation Algorithms for Direct-Sequence Spread-Spectrum CDMA Single-User Detection” by N. Zecevic et al., Proceedings of IEEE Vehicular Technology Conference, pp. 2133-2137, 1991. discloses an adaptive receiver for a DSSS/CDMA system using the Griffiths' algorithm as the adaptation algorithm, and is incorporated herein by reference.
The Griffiths' algorithm, like the LMS algorithm, is an approximate implementation of the method of steepest descent. Generally, the method of steepest descent solves the Wiener-Hopf equation by updating the filter tap-weights recursively. Since the method of steepest descent still requires the correlation matrix, the computation load is not easy to handle. Using the instantaneous estimate instead of the correlation matrix can reduce the computation load. The resultant algorithm is the Griffiths' algorithm.
More particularly, if the correlation matrix R and the cross-correlation vector p are known, the method of steepest descent has an updating equation for a tap-weight vector w expressed as
w
(
j+
1)=
w
(
j
)+&mgr;[
p−Rw
(
j
)]  (3)
where &mgr; is a small fixed step-size. The Griffiths' algorithm simplifies the updating equation (3) by replacing the correlation matrix with an instantaneous estimate r(j)r
T
(j), but still must have the knowledge of the cross-correlation vector p in advance. For the DSSS/CDMA communication system, the cross-correlation vector p becomes A
l
c
l
, where c
1
is the spreading sequence of interest if the timing synchronization between the receiver and the symbol clock of interest is perfect. The Griffiths' algorithm can thus be expressed as
w
(
j+
1)=
w
(
j
)+&mgr;[
c
l
−z
(
j
)
r
(
j
)]  (4)
where &mgr; is a small step-size and z(j)=w
T
(j)r(j) is the output of the detector.
A disadvantage of the Griffiths' algorithm is that, like the LMS algorithm, its convergence properties depend on the eigenvalue spread of the correlation matrix associated with the received signal vectors. As the interference level or the number of users increases, the problem of eigenvalue spread will be severe such that the convergence speed of the Griffiths' algorithm with be very slow.
Several variable step-size (VSS) LMS algorithms have been proposed to accelerate convergence speed. In general, these methods employ a procedure that uses a larger step-size to obtain a faster convergence speed when the error signal is large, and adopt a smaller step-size to obtain a smaller mean-square error as the error signal decreases. For example, “A Variable Step Size LMS Algor

Affiliated with

Also associated with

Rating

Apparatus and method for a DSSS/CDMA receiver does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Apparatus and method for a DSSS/CDMA receiver, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for a DSSS/CDMA receiver will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2888866