Digital equalization method and apparatus

• US 5,164,959 A
• Filed: 01/22/1991
• Issued: 11/17/1992
• Est. Priority Date: 01/22/1991
• Status: Expired due to Term

• Current Assignee: Microelectronics Technology Inc (MTI Laboratory Inc)
• Original Assignee: Hughes Aircraft Company (Rtx Corporation)
• Inventors: Cai, Khiem V., Dell-Imagine, Robert A.
• Primary Examiner(s): Tarcza, Thomas H.
• Assistant Examiner(s): CAIN, DAVID C
• Application Number:
  US07/643,969
• Time in Patent Office: 665 Days
• Field of Search: 375/1
• US Class Current: 375/150
• CPC Class Codes: H04B 1/7093 Matched filter type
• CPC Class Codes: H04B 1/715 Interference-related aspects
• CPC Class Codes: H04B 2001/7152 with means for suppressing interference
• CPC Class Codes: H04L 27/2331 wherein the received signal is demodulated using one or more delayed versions of itself

Abstract

A digital equalizer and equalization method for a spread spectrum communication link that has a dispersive channel and other systems that process signals having a sufficient degree of redundancy in time and/or phase. No training signal is required, except for an initialization bit sequence. The receiver incorporating the equalizer uses correlation over a bit time between a received signal, that includes data modulation, and a locally generated spreading signal that does not. The output of the correlator is a set of complex digital samples (I and Q-channel) taken at the spreading sequence bit rate. In the absence of multipath dispersion, there is perfect cross-correlation function, and hence only one large sample per transmitted data symbol. With multipath dispersion, the received energy is spread over two or more samples of the correlation. The equalizer combines the signal energy contained in the multiple output samples, converting the concatenation of the channel and equalizer into an equivalent nondispersive channel. The equalizer uses the output of its demodulator function to convert the data bearing signals into a known signal structure, which is used to determine the (unknown) channel impulse response. The equalizer employs optimum filtering of the individual correlation samples averaged over many bit times, to determine a set of multiplicative weights that time and phase align the samples so that they are coherently added. The equalizer improves the effective signal to noise ratio to a point that is close to the theoretical nondispersive channel without increasing signal power or reducing data rate.

Claims

1. A digital equalizer for use with a spread spectrum communications system, said digital equalizer comprising:

data input means for receiving an input signal comprising a correlated output signal corresponding to a complex I and Q-channel digital signal that has a data phase associated with a non-dispersive channel and a random phase associated with a dispersive channel;

time delay means for delaying the input signal to produce a time delayed input signal;

demodulation means for determining the data phase contained in the input signal

first combining means for combining the time delayed input signal with the data phase of the input signal to produce an estimated matched filter signal;

accumulating means for computing an accumulated estimated matched filter signal;

second combining means for combining the correlated output signal with the accumulated matched filter signal to eliminated the random phase contained in the correlated output signal;

integrating means for integrating the correlated output signal to increase the signal to noise ration thereof and for providing a time equalized signal.

2. The equalizer of claim 1 which further comprises:

conjugation means for providing the complex conjugate of the time delayed input signal.

3. The equalizer of claim 1 which further comprises:

first conjugation means for providing the complex conjugate of the data phase signal; and

second conjugation means for providing the complex conjugate of the accumulated estimated matched filter signal.

4. The equalizer of claim 1 which further comprises:

noise removal means for removing noise from the correlated output signals by setting the values of the accumulated matched filter signals that are less than a predetermined threshold to zero.

5. The equalizer of claim 4 wherein the noise removal means comprises:

means for sorting the accumulated matched filter signals;

means for comparing the amplitudes of the sorted accumulated matched filter signals to the predetermined threshold; and

means for setting the values of all accumulated matched filter signals that are less than or equal to the predetermined threshold to zero.

6. The equalizer of claim 2 which further comprises:

noise removal means for removing noise from the correlated output signals by setting the values of the accumulated matched filter signals that are less than a predetermined threshold to zero.

7. The equalizer of claim 6 wherein the noise removal means comprises:

means for sorting the accumulated matched filter signals;

means for comparing the amplitudes of the sorted accumulated matched filter signals to the predetermined threshold; and

means for setting the values of all accumulated matched filter signals that are less than or equal to the predetermined threshold to zero.

8. The equalizer of claim 3 which further comprises:

noise removal means for removing noise from the correlated output signals by setting the values of the accumulated matched filter signals that are less than a predetermined threshold to zero.

9. The equalizer of claim 8 wherein the noise removal means comprises:

means for sorting the accumulated matched filter signals;

means for comparing the amplitudes of the sorted accumulated matched filter signals to the predetermined threshold; and

means for setting the values of all accumulated matched filter signals that are less than or equal to the predetermined threshold to zero.

10. A method of digitally equalizing a data symbol transmitted using a spread spectrum communications system, said method comprising the steps of:

processing the data symbol to determine its phase;

time-delaying the data symbol;

combining the phase with the time-delayed data symbol to strip the phase from time-delayed symbol;

accumulating the combined signal to produce a matched filter signal;

combining the matched filter signal with the non time-delayed data symbol to produce a combined signal; and

integrating the combined signal to produce an equalized signal.

11. The method of claim 10 wherein the data symbol is transmitted over a dispersive channel using a spread spectrum communications system, and wherein the method further comprises the step of:

correlating a received spread spectrum communications signal with a locally generated reference signal to provide a correlated output signal comprising the data symbol encoded with the data phase and a random phase associated with the dispersive channel.

12. The method of claim 11 which further comprises the steps of:

complex conjugating the time delayed correlated output signal.

13. The method of claim 11 which further comprises the steps of:

complex conjugating the data phase signal; and

complex conjugating the accumulated estimated matched filter signal.

14. The method of claim 11 which further comprises the step of:

removing noise from the accumulated signals by setting the values of all channel impulse response signals that are less than a predetermined threshold to zero.

15. The method of claim 14 wherein the noise removing step comprises the steps of:

sorting the channel impulse response signals;

comparing the amplitudes of the sorted channel impulse response signals to the predetermined threshold; and

setting the values of all channel impulse response signals that are less than or equal to the predetermined threshold to zero.