Following up on my last posts, I wanted to point to Tim O’Shea’s recent work on GNU Radio channel models.
He looked into the flat fader block and came up with patches to fix problems regarding the autocorrelation, which I experienced earlier. With these changes, I’m very postivie that we have a validated and trustworthy fading implementation for GNU Radio.
In addition to that, he came up with a new tapped delay line channel model that introduces some randomness in the delay profile (usually only the amplitudes and phases of the individual paths change over time). So checkout his great post and watch the video.