High performance equalizer having reduced complexity
View Patent ↗An apparatus and method for implementing an equalizer which combines the benefits of a decision feedback equalizer (DFE) with a maximum-a-posterori (MAP) equalizer (or a maximum likelihood sequence estimator, MLSE) to provide an equalization device with significantly lower complexity than a full-state MAP device, but which still provides improved performance over a conventional DFE. The equalizer architecture includes two DFE-like structures, followed by a MAP equalizer. The first DFE forms tentative symbol decisions. The second DFE is used thereafter to truncate the channel response to a desired memory of L 1 symbols, which is less than the total delay spread of L symbols of the channel. The MAP equalizer operates over a channel with memory of L 1 symbols (where L 1 <=L), and therefore the overall complexity of the equalizer is significantly reduced.
1. An equalizer comprising:
a decision feedback equalizer configured to receive an input signal, the decision feedback equalizer comprising a decision device and a first feedback filter coupled to the decision device in feedback fashion;
a second feedback filter, separate from the first feedback filter, configured to receive a decision from the decision device and to use the decision to produce a cancellation signal; and
a non-linear equalizer configured to receive the cancellation signal and to produce an output signal based on the cancellation signal.
2. The equalizer of claim 1 , wherein the second feedback filter is further configured to reduce the energy of post-cursor signal elements in the input signal that exist beyond a predetermined channel memory.
3. The equalizer of claim 1 , wherein the non-linear equalizer comprises:
a maximum likelihood sequence estimation (MLSE) equalizer.
4. The equalizer of claim 1 , wherein the non-linear equalizer comprises:
a maximum a posteriori (MAP) equalizer.
5. The equalizer of claim 1 , wherein the second feedback filter is further configured to cancel post-cursor signal elements in the input signal that exist beyond a predetermined channel memory.
6. The equalizer of claim 1 further comprising:
a feedforward filter configured to receive an incoming signal and to increase a signal-to-noise ratio of the incoming signal to provide the input signal.
7. A method of processing a communication signal, comprising:
processing, by a decision feedback equalizer, an input signal to produce a decision signal, the decision feedback equalizer including:
a decision device; and
a first feedback filter coupled to the decision device in feedback fashion;
processing, by a second feedback filter, the decision signal to produce a cancellation signal, wherein the second feedback filter is separate from the first feedback filter; and
generating, by a nonlinear equalizer, an output signal based on the cancellation signal.
8. The method of claim 7 , wherein the processing by the second feedback filter further comprises:
reducing the energy of post-cursor signal elements in the input signal that exist beyond a predetermined channel memory.
9. The method of claim 7 , wherein the generating comprises:
generating the output signal using a maximum likelihood sequence estimation (MLSE) non-linear equalizer.
10. The method of claim 7 , wherein the generating comprises:
generating the output signal using a maximum a posteriori (MAP) equalizer.
11. The method of claim 7 , wherein the processing by the second feedback filler further comprises:
cancelling post-cursor signal elements in the input signal that exist beyond a predetermined channel memory.
12. The method of claim 7 , further comprising:
processing an incoming signal using a feedforward filter to increase a signal-to-noise ratio of the incoming signal to provide the input signal.
13. An equalizer, comprising:
a feed forward filter configured to filter a signal according to a plurality of filter coefficients to provide an input signal, the plurality of filter coefficients being based on a channel estimate of the signal;
a first decision feedback equalizer configured to provide a decision signal,
wherein the decision signal is based on a difference between the input signal and a filtered signal, and
wherein the filtered signal is provided by filtering the decision signal according to the plurality of filter coefficients;
a second decision feedback equalizer configured to provide a cancellation signal by filtering the decision signal according to a subset of the plurality of filter coefficients; and
a non-linear equalizer configured to provide an output signal based on a difference between the input signal and the cancellation signal.
14. The equalizer of claim 13 , further comprising:
a feedback filter configured to filter the output signal to provide a second cancellation signal; and
a second non-linear equalizer configured to produce a second output signal based on the second cancellation signal.
15. The equalizer of claim 13 , wherein the non-linear equalizer comprises:
a maximum likelihood sequence estimation (MLSE) non-linear equalizer; or a maximum a posteriori (MAP) equalizer.
16. The equalizer of claim 1 , further comprising:
a third feedback filter configured to filter the output signal to provide a second cancellation signal; and
a second non-linear equalizer configured to produce a second output signal based on the second cancellation signal.
17. The equalizer of claim 1 , wherein the decision device is further configured to provide the decision as part of a decision signal,
wherein the first feedback filter is further configured to filter the decision signal according to a plurality of filter coefficients; and
wherein the second feedback filter is farther configured to produce the cancellation signal by filtering the decision signal according to a subset of the plurality of filter coefficients.
18. The equalizer of claim 1 , wherein the non-linear equalizer is further configured to process a filtered signal according to a state space, the filtered signal being representative of a difference between the cancellation signal and the input signal to provide the output signal, and
wherein the state space is less than a full state space corresponding to M L , M being representative of a size of a symbol alphabet used for modulation of the input signal, and L being representative of a channel memory corresponding to the input signal.
19. The method according to claim 7 , further comprising:
processing, by a third feedback filter, the output signal to produce a second cancellation signal; and
generating, by a second non-linear equalizer, a second output signal based on the second cancellation signal.
20. The method according to claim 7 , wherein the processing by the decision feedback equalizer comprises:
filtering the decision signal according to a plurality of filter coefficients to provide a filtered signal; and
subtracting the filtered signal from the input signal to provide an input to the decision device, and
wherein the processing by the second feedback filter comprises:
filtering the decision signal according to a subset of the plurality of filter coefficients to provide the cancellation signal.