Hi Glen, Yes indeed it can be trained to any channel you like. We've been doing some work for the VHF/UHF land mobile radio (LMR) space where it's trained to pass through analog FM radios (see freedv.org blog posts, search on BBFM).
Greg - I suspect the final form of this technology arc will be something like you suggest, resulting in high quality speech at very low SNRs. Current work on RADE V2 has the ML doing the sync (e.g. equalisation for the HF channel, fine timing, frame sync) rather than classical DSP, so that can now be trained rather than designed for specific channels. - David On Sat, 2025-08-16 at 16:36 +1000, glen english LIST wrote: > Hi David > Have just read all the papers.. as i read it , ( > https://arxiv.org/pdf/2505.06671) > > we could alternatively train the system for say aurora channel > characteristsics on 2m , or rainscatter characteristics on 10 GHz > > and could expand or use different training sets for different > propagation modes on different bands, > ....depending on how general they are. > > IE this is a workiong deomonstration for AWGN and HF style multipath > channels, but not at all limited it > > best regards > -glen > > > On 11/08/2025 16:44, david wrote: > > Hi Richard, > > > > Yes it certainly is planned, but as per the Radio Autoencoder page: > > _______________________________________________ > Freetel-codec2 mailing list > Freetel-codec2@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 _______________________________________________ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freetel-codec2
