That waterfall doesn't look like any sensible digital modulation to me, or at least if I were doing HFT, the spectrogram would look totally different.

I think the source is rather an unintentional radiator in this case.