On 8/28/21 11:04 PM, Claude Frantz via wsjt-devel wrote: > Hi all, > > These references are interesting, but we have to remember that we are > speaking about HF communication on short-wave bands, in half duplex > mode. None of the actors (the stations on the band) has a full > knowledge of the whole situation of the communication channel in use. > Each station has only knowledge about the stations it can see on its > own receiver during the receiving slots. This is not a global view. > Further, the situation is continuously changing because of the > changing propagation conditions. In this context, with this limited > information, it is very difficult to find the good choice in a global > sense. That's exactly the point I made in my old MACA paper. I argued you can do better by inferring what other stations are hearing than by simply noting when your own receiver is active. If you hear a station turn it over to someone else and then the channel goes silent (to you), you can reasonably infer that you probably should not transmit now. > > PSKreporter is the tool which has a better global overview, but its > use is not mandatory, not always possible and PSKreporter displays a > view of the past because of the technology used. In the context of > continuously changing propagation, this information is of limited > usefulness.
It also means increased dependency on the Internet. I like the Internet as much as anyone, but we should be careful about depending on it too much. We're already dependent on either GPS or the Internet for time. > > When my partner station cannot copy me well, it makes sense to try to > use another TX frequency which seems to be free according to my own > RX. Perhaps, this can help, but it's not sure. If cannot see what the > RX of my partner displays. I cannot verify if my decision was good. If > the situation persists, I can just try another TX frequency again. > Yes, and my idea is to try all those frequencies in a pseudo random sequence that he can duplicate (because he knows my callsign and the time) so I don't have to tell him. And even if one transmission collides, hopefully the next one won't. I haven't yet gotten into methods to keep channels stable under heavy overload, e..g, exponential backoff, but they can come later. Phil _______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel
