An equalizer is a common component of many modern acoustic communication systems. In
the past, it has been assumed that using channel estimate
based methods are better than direct adaptation methods for determining
the best equalizer coefficients in the mean squared error sense.
However, these assumptions are based on work using a Rayleigh fading
channel model and never provided any analytical explanation. This talk
will provide analytical results for why one method is better than the
other. Also, it is shown that for the signal-to-noise ratios common in
underwater communication, these two methods are approximately equivalent.