Multiplex communications – Communication over free space – Combining or distributing information via code word channels...
Reexamination Certificate
1999-05-19
2002-09-03
Hunter, Daniel (Department: 2684)
Multiplex communications
Communication over free space
Combining or distributing information via code word channels...
C370S479000, C370S335000, C370S320000
Reexamination Certificate
active
06445692
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to CDMA receivers and to blind adaptive equalizers for a multiuser CDMA communication system.
2. Related Art
Recently Code Division Multiple Access (CDMA) telecommunication systems have gained ground and popularity against rival TDMA/FDMA solutions, despite the increased complexity of the receiver. In the CDMA framework, different users employ distinct spreading codes but transmit at the same time and frequency. CDMA systems break up digital communications into bits at the sending end, and spray the bits across a slice of the spectrum. At the receiving end, the bits are reassembled based on signatures, or wave forms, associated with the bits of the corresponding users. A problem with CDMA systems arises because of interference from other users. This is especially prevalent where there are weak signal users and/or there are reflections from buildings, mountains or other structures that distort signals (multi-path distortions). Various interference suppression schemes have been used in the past to separate out Multi-User Interference (MUI) to obtain a clean signal at the receiving end. In one such scheme, signals are filtered out, one by one, from the strongest to the weakest.
Significant effort has focused recently on designing multiuser receivers in order to suppress MUI and deliver the promised capacity gains of CDMA technology. While optimal solutions have been derived, their exponentially increasing complexity renders them inappropriate for systems with a large number of users. For this reason, recent efforts have concentrated on suboptimal linear solutions and have investigated ways to derive the receiver parameters from the data.
Multiuser receiver design can be divided into two categories based on whether training sequences are used or not. In the first category, non-blind MMSE receivers can be adaptively implemented if the desired signal is known at the receiver, while in the second category, a blind approach has to be employed. Blind subspace methods show good performance, but the required SVD of large matrices leads to a heavy computational load. On the contrary, blind adaptive multiuser detectors significantly reduce the computational burden. A scheme was presented by Honig, et al., “Blind Adaptive Multiuser Detection” IEEE Transactions on Information Theory, Vol. 41, No. 4, pp. 944-960, July 1995, for the case when multipath interference is absent and showed performance equivalent to that of the MMSE receiver. In Wang, et al., “Blind Multiuser Detection: A Subspace Approach” IEEE, Tans. On the Information Theory, Vol. 44, No.2, pp. 677-690, March 1998, an adaptive implementation based on subspace tracking method was shown to improve performance at the expense of more computational complexity. The method of Honig, et al. was later extended by adding more constraints by Schodorf, et al., “A constrained Optimization Approach to Multiuser Detection” IEEE Transactions on Signal Processing, Vol. 45, No. 1, pp. 258-262, January 1997. Unfortunately, in the presence of multipath and multichip interference, minimum variance solutions are known to be sensitive to signature mismatch. A solution for that case was attempted by Tsatsanis, “Inverse Filtering Criteria for CDMA Systems, IEEE Trans. On Signal Processing, Vol. 45, No. 1, pp. 102-112, January, 1997, and later by Shen, et al., Blind Adaptive Multiuser CDMA Detection Based on a Linear Projection Constraint” Proc. SPAWC'97, pp. 261-264, La Villette, Paris, France, Apr. 16-18, 1997, by forcing the receiver response to delayed copies of the signal of interest to zero. With these additional constraints, minimum variance techniques are applicable, but have inferior performance since they treat part of the useful signal as interference. Recently however, constrained optimization solutions were developed which combine all multipath components of the signal of interest and jointly minimize MUI while maximizing the signal component at the receiver output, Tsatsanis, et al., “Performance Analysis of Minimum Variance CDMA Receivers,” Proc. 13st Intl. Conf. On Digital Signal Processing (DSP97), Vol. 1, pp. 379-382, Santorini-Hellas, Greece, Jul. 2-4, 1997. A related development from a different viewpoint was also reported in Liu, et al., “Low Complexity Receivers for CDMA Communications over Frequency-selective Channels,” Proc. SPAWC'97, pp.265-268, La Villette, Paris, France, Apr. 16-18, 1997. These blind methods exhibit superior performance which is close to that of the trained MMSE receiver Tsatsanis, et al. Unfortunately however, adaptive implementations of the solution of Tsatsanis, et al. are not straightforward, as they correspond to the optimization of a non-linear cost function.
Other efforts in this area include:
Kawabe, et al., U.S. Pat. No. 5,394,434 multiple-access (CDMA) demodulator which stores received chip data in a separate memory for each transmitting station. Each memory area holds data for a plurality of symbols. When the receiving station finishes receiving a signal from one transmitting station the corresponding area is designated for processing. All data stored in the designated area is correlated with the spreading code of the transmission station to generate correlated values for a plurality of symbols. Estimated symbols are derived from the correlated value and the temporarily stored. The difference between each estimated symbol value and the previous estimate is multiplied by the spreading code of the designated station to generate remaining interference value. The remaining interference values are subtracted from all chip data stored in the memory except for data stored in a designated area thereby updating the memory contents. A demodulated value is output for the oldest symbols stored in a designated memory area.
Divsalar, et al., U.S. Pat. No. 5,644,592, discloses a method of decoding a spread spectrum composite signal to provide an improved method of interference cancellation for CDMA communications. The system comprises a parallel interference cancellation (IC) system that reduces the degrading effect of multiuser interference. The patent discloses a method of decoding a spread spectrum composite signal comprising plural user signals spread with plural respective codes, wherein each user signal is despread, filtered to produce a signal value, analyzed to produce a tentative decision value, respread, summed with other re-spread signals to produce combined interference signals, combining a portion of each signal value with one of the combined interference signals and the composite signal to produce an estimate of a respective user signal to preserve information contained in the user signal. Each signal is processed for each user by a matched filter which would be dominated by the corresponding user signal and will produce a maximum signal to noise ratio (SNR). Thereafter, a re-scaling amplifier amplifies the signal with an estimate of the transmitted power. The method attempts to cancel the multiuser at each stage of the IC device.
Hoff, et al., U.S. Pat. No. 3,619,586 discloses a second-recuversive digital filter optimized by the equation
f
(
n
)=
CX
(
n
)+2(%+&bgr;
2
)
f
(
n−
1)
−f
(
n−
2)
+&bgr;
1
[
%+&bgr;
2
)
f
(
n−
1)
−f
(
n−
2)]
where the sum of %+&bgr;
2
may be provided as a single coefficient K. The filter can be used as a band-pass filter or as a oscillator.
Tanaka, et al., U.S. Pat. No. 5,602,765 discloses an adaptive transfer function estimating method for estimating with a projection algorithm function of an unknown system and its output in an acoustic cancelier, active noise control, or the like. The estimated transfer function correcting vector calculation calculates an estimated transfer function correcting vector on the basis of the error signal and the input signal to the unknown system.
Weng, et al., U.S. Pat. No. 5,652,903 discloses a co-processor for use with a digital signal processor used on an inte
Harry Andrew T
Hunter Daniel
The Trustees of the Stevens Institute of Technology
Wolff & Samson
LandOfFree
Blind adaptive algorithms for optimal minimum variance CDMA... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Blind adaptive algorithms for optimal minimum variance CDMA..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Blind adaptive algorithms for optimal minimum variance CDMA... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2870528