Feed forward equalization for highly-spectrally-efficient communications
A receiver may be operable to receive a signal. A sequence estimation module of the receiver may generate estimated symbols corresponding to the received signal. The generating of the estimated symbols may use tap information associated with one or both of a pulse shaper via which the signal was transmitted and an input filter of the receiver. The sequence estimation module may generate a reconstructed signal based on the estimated symbols and the tap information. A feed forward equalizer (FFE) of the receiver may adapt a plurality of tap coefficients of the FFE based on the reconstructed signal. The signal may be equalized via the FFE. The adaptation of the tap coefficients of the FFE may be based on a least-mean-square (LMS) process for minimizing a mean square of the error signal. An output signal of the FFE may comprise a power gain compensation.
1. A system comprising:
a receiver, the receiver being operable to:
receive a signal in the receiver;
generate, by a sequence estimation module of the receiver, estimated symbols corresponding to the received signal, wherein the generating of the estimated symbols uses tap information associated with one or both of a pulse shaper via which the received signal was transmitted and an input filter of the receiver;
generate, by the sequence estimation module, a reconstructed signal, wherein:
the reconstructed signal is generated based on the estimated symbols and the tap information;
the reconstructed signal is generated based on convolution of the estimated symbols with the tap information; and
the generation of the reconstructed signal incorporates a non-linear model that approximates nonlinearity generated in a transmitter from which said received signal originated;
adapt a plurality of tap coefficients of a feed forward equalizer (FFE) of the receiver based on the reconstructed signal; and
equalize the received signal via the FFE.
2. The system according to claim 1 , wherein a plurality of taps of the FFE comprise a center tap, one or more early taps and one or more late taps.
3. The system according to claim 1 , wherein the FFE is operable to adapt the plurality of tap coefficients of the FFE based on a least-mean-square (LMS) process for minimizing a mean square error.
4. The system according to claim 1 , wherein an output signal of the FFE comprises a power gain compensation.
5. The system according to claim 1 , wherein the FFE is operable to utilize a determined loop gain to adjust a convergence time of the adaptation of the plurality of tap coefficients of the FFE.
6. The system of claim 1 , wherein each of said generated estimated symbols is generated using information from multiple symbols of the received signal.
7. The system of claim 6 , wherein said sequence estimation module uses a Viterbi algorithm.
8. The system of claim 1 , wherein said one or both of a pulse shaper via which the received signal was transmitted and an input filter of the receiver consists only of the pulse shaper via which the received signal was transmitted.
9. The system of claim 8 , wherein the pulse shaper is a partial response pulse shaper which introduced a substantial, controlled amount of inter-symbol interference in the received signal.
10. The system of claim 1 , wherein said non-linear model is a model of non-linearity experienced by said received signal in a transmitter of said received signal.
11. The system of claim 1 , wherein said non-linear model is a model of non-linearity experienced by said received signal en route to said sequence estimation module.