David - right but you can't beat the physics.. 4 levels compared with two.....
The reason why any of these sort of N-FSK FM baseband systems work at all is that there are never called to work down near in the FM demod threshold .... Going to a narrow bandwidth (say div 2) sure wins you 3dB , but the increased complexity of demod (more bits per symbol) will hurt much more than the 3dB you got. Bruce- multipath performance- well depends if you are talking short delay (where the whole channel disappears in the fade) or long delay( the channel itself suffers selective frequency distortion- say >100uS) For long delay multipath- In the FM demod case, you are pretty much screwed because the channel errors that occur at IF turn themselves into horrendous intermod /harmonic distortions at demodulated baseband. So, I think the result is similar for GMSK, audio OFDM etc. For short delay, where the channel disappears, nothing much you can do about that for a stationary device except provide some sort of frequency or space diversity in either the transmitter or receiver (or both) . Dean- have a look at my post a few posts back- That begins with " "Hi David , I think whether one can do better than FM in the real world 100% depends" Yeah I think the audio ofdm has the best chance of getting through the average radio mic/speaker system This is why FreeDV works so well. The simplicity of Free DV is obtained by the nature that an SSB channel is a very linear thing compared with an FM mod-demod chain (which are things of the devil) . Life is not so easy in the FM mod demod chain, so there are some mods that would improve it so it deals with pre-emphasis-de emphasis cleanly, and also takes advantage of the fm demod output high SNR in the low frequency audio band. The clicks at the fm detector threshold will tend to rile it though, some interleaving will be necessary so that a click (which affects all bins simultaneously) doesn't cause a complete loss of frame. 73 Glen VK1XX On 23/12/2014 1:49 PM, David Rowe wrote: > Some tables I looked at (http://www.atlantarf.com/FSK_Modulation.php) > suggest 4FSK bas a bandwidth efficiency of 0.4 bits/s/Hz, BT=0.5 GMSK > about 0.5 (99% power). > > Coherent GMSK is about 2dB more efficient than non-coherently detected > 4FSK (I don't have any figures for coh 4FSK). So not much in it. It > would be nice to know the reasons, and maybe have a go at 4FSK. I > imagine the demod is easier, but we only need to write that once. > > To me they are both pretty close to the best physics allows. > > - David > > ------------------------------------------------------------------------------ > Dive into the World of Parallel Programming! The Go Parallel Website, > sponsored by Intel and developed in partnership with Slashdot Media, is your > hub for all things parallel software development, from weekly thought > leadership blogs to news, videos, case studies, tutorials and more. Take a > look and join the conversation now. http://goparallel.sourceforge.net > _______________________________________________ > Freetel-codec2 mailing list > Freetel-codec2@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 > -- - Glen English RF Communications and Electronics Engineer CORTEX RF & Pacific Media Technologies Pty Ltd ABN 40 075 532 008 PO Box 5231 Lyneham ACT 2602, Australia. au mobile : +61 (0)418 975077 ------------------------------------------------------------------------------ Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net _______________________________________________ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freetel-codec2
