When analyzing a modulated signal, the error vector magnitude may be higher for the following cases:
The measurement implements an adaptive feed-forward equalizer, which implies that the equalizer filter taps adapt their coefficients to compensate for the action of the channel filter or any other filter. This adaptation ensures that the convolution of the channel filter and the equalizer coefficient filter yields a delta function, thereby removing any inter symbol interference in the equalized output complex waveform. The adaptive feed-forward equalizer uses a feed-forward adaptive least-mean-squared (LMS) algorithm to adjust the equalizer taps.
ASK, PSK, and QAM modulation types support equalization.
At the start of the equalization process, you must set the equalizer mode to TRAIN for iterations specified by training count. The equalizer tap spacing is specified by the number of taps per symbol. The incremental amount by which the equalizer taps adapt is controlled by specifying the convergence factor. A smaller value of the convergence factor requires a larger number of iterations to adapt the equalizer, and it provides a better estimate of the channel. The training starts with an impulse as the initial coefficients.
After training, the equalizer coefficients are applied before the demodulation process.
To continue training of the equalizer in further iterations, specify the previous equalizer coefficients as the initial coefficients, with equalizer mode set to TRAIN.
Refer to the Custom Filters topic for more information about how to specify the coefficients and the coefficient spacing.
After verifying that the channel has been accurately represented by the equalizer, the equalizer mode can be set to HOLD to stop adapting and use the coefficients prior to the demodulation process.