This is machine translation

Translated by Microsoft
Mouseover text to see original. Click the button below to return to the English verison of the page.

Note: This page has been translated by MathWorks. Please click here
To view all translated materals including this page, select Japan from the country navigator on the bottom of this page.


Equalize signal using equalizer object


y = equalize(eqobj,x)
y = equalize(eqobj,x,trainsig)
[y,yd] = equalize(...)
[y,yd,e] = equalize(...)


y = equalize(eqobj,x) processes the baseband signal vector x with equalizer object eqobj and returns the equalized signal vector y. At the end of the process, eqobj contains updated state information such as equalizer weight values and input buffer values. To construct eqobj, use the lineareq or dfe function, as described in Adaptive Algorithms. The equalize function assumes that the signal x is sampled at nsamp samples per symbol, where nsamp is the value of the nSampPerSym property of eqobj. For adaptive algorithms other than CMA, the equalizer adapts in decision-directed mode using a detector specified by the SigConst property of eqobj. The delay of the equalizer is (eqobj.RefTap-1)/eqobj.nSampPerSym, as described in Delays from Equalization.

Note that (eqobj.RefTap-1) must be an integer multiple of nSampPerSym. For a fractionally-spaced equalizer, the taps are spaced at fractions of a symbol period. The reference tap pertains to training symbols, and thus, must coincide with a whole number of symbols (i.e., an integer number of samples per symbol). eqobj.RefTap=1 corresponds to the first symbol, eqobj.RefTap=nSampPerSym+1 to the second, and so on. Therefore (eqobj.RefTap-1) must be an integer multiple of nSampPerSym.

If eqobj.ResetBeforeFiltering is 0, equalize uses the existing state information in eqobj when starting the equalization operation. As a result, equalize(eqobj,[x1 x2]) is equivalent to [equalize(eqobj,x1) equalize(eqobj,x2)]. To reset eqobj manually, apply the reset function to eqobj.

If eqobj.ResetBeforeFiltering is 1, equalize resets eqobj before starting the equalization operation, overwriting any previous state information in eqobj.

y = equalize(eqobj,x,trainsig) initially uses a training sequence to adapt the equalizer. After processing the training sequence, the equalizer adapts in decision-directed mode. The vector length of trainsig must be less than or equal to length(x)-(eqobj.RefTap-1)/eqobj.nSampPerSym.

[y,yd] = equalize(...) returns the vector yd of detected data symbols.

[y,yd,e] = equalize(...) returns the result of the error calculation described in Error Calculation. For adaptive algorithms other than CMA, e is the vector of errors between y and the reference signal, where the reference signal consists of the training sequence or detected symbols.


For examples that use this function, see Equalize Using a Training Sequence in MATLAB, Example: Equalizing Multiple Times, Varying the Mode, and Example: Adaptive Equalization Within a Loop.

See Also

| | | | | | |

Introduced before R2006a

Was this topic helpful?