Any suggestions on appropriate/inappropriate ways of writing log(mag(fft(i,q))) to SigMF, rather than (i,q)?
We want to sweep 0.5 - 6 GHz
repeatedly 24x7
with an Ettus X310, concurrently using one antenna/UBX160 each inside and outside a hangar, to capture local spectra and see how well the existing structure shields in<->out.
We don't want to capture samples to disk then post process, but rather convert from samples to spectra in real time, do in software the equivalent of a SpecAn max-hold function for a while, then write those traces to disk. This is so we don't have to buy truckloads of disk drives to capture all the intermittent emissions for a week. ;-)
We can do this in what seem to us the obvious ways, but would rather not reinvent wheels, especially if there are any conventions on how this should be done in the spirit of SigMF.
Thanks!