Joint adaptive optimization of soft decision device and feedback equalizer
The present invention uses a novel adaptive soft decision device in order to jointly optimize decision device and DFE operation. The soft decision device receives the input and output samples of the slicer and generates a feedback sample by non-linearly combining them with respect to a single decision reference parameter. Moreover, the soft decision device provides novel error terms used to adapt equalizer coefficients in order to jointly optimize decision device and equalizer coefficients.
1. A method for operating forward and feedback filters in a communications receiver having a decision feedback equalizer, said communications receiver having a decision feedback equalizer, said communications receiver responsive to a received signal to form slicer input samples and hard decision samples corresponding to said slicer output samples, said method comprising:
non-linearly combining said slicer input samples and said hard decision samples to form a soft decision samples in a soft decision device; and
the said soft decision device being adaptively adjusted to control contributions of said slicer input samples and said hard decision samples to form said soft decision samples; and
said soft decision device generating error signals to adjust said forward and feedback filters; and
operating said forward and feedback filters by coupling a composite decision samples to said feedback filter and by adapting said forward and feedback filters with said error signals.
2. The method of claim 1 , wherein said decision feedback equalizer operates in passband.
3. The method of claim 1 , wherein said feedback filter operates in baseband.
4. A receiver having a decision feedback equalizer, said receiver responsive to a received signal to form slicer input samples and hard decision samples corresponding to said slicer output samples, said receiver comprising:
a forward filter;
a feedback filter;
a slicer that is capable of generating the slicer output samples; and
a soft decision device that is capable of non-linearly combining said slicer input samples and said hard decision samples to form soft decision samples, said soft decision device further being adaptively adjusted to control contributions of said slicer input samples and said hard decision samples to form said soft decision samples, further wherein said soft decision device is capable of generating error signals to adjust said forward and feedback filters;
wherein said forward and feedback filters are capable of coupling a composite decision samples to said feedback filter and by adapting said forward and feedback filters with said error signals.
5. The receiver of claim 4, wherein said decision feedback equalizer operates in passband.
6. The receiver of claim 4, wherein said feedback filter operates in baseband.
7. The receiver of claim 5, wherein said forward filter and said feedback filter operate in the passband.
8. The receiver of claim 6, wherein said forward filter operates in passband.
9. The receiver of claim 4 wherein said forward filter has a plurality of coefficients that are updated using a stochastic gradient descent rule.