On Sat, Apr 17, 2010 at 4:49 PM, hteller <[email protected]> wrote:
> Brian, > > > > But when the path provides less loss, you can take advantage of that > extra S:N to do one of two things: increase speed (less energy needed per > bit) or decrease power. And who cares how fast someone types. Why should the > transmitter be on for the entire time the person is typing? Why not transmit > when some or all of the message is ready? > > This is the principle of the "typeahead" buffer, which almost all PSK31, > and other digital programs, programs already have. Some even have an option > to transmit when the Enter key is pressed. The PROBLEM is that the > frequency is occupied, whether or not anyone is currently sending. That is > the purpose of transmitting the QRL and waiting for a reply. > Media Access Control depends upon control of access, Yeah, that is what its name implies. :-) > but on the ham bands, access is random, and not controllable. Why? You don't even need everyone participating for it to work. And don't stop with thinking that your MAC just controls *when* you transmit but also *where* (frequency) you transmit and maybe even *how* you transmit. Our method of "control" is listening first, or the QRL, and courteous > operating practices. So? That is CSMA. > This is the way that we SHARE frequencies to make most efficient use of > them by a lot of stations which may fade in and out. Statistical heuristic on CSMA. If you can control access then you can share resources, but if access by a > central control is not possible, then we must rely on the human brain to > control access. Why? Do you have any idea just how complex distributed control is in the Internet and the cell phone systems? There is no central control. Yet it works very well with many orders of magnitude more complexity than what we are talking about here. If you think about it, what we do as human media access control computers is actually very algorithmic. I bet you could write down the steps you use to find a frequency and operate. But given the power of the computers we have there is no reason why we can't automate that process. Here is an example of how traditional thinking skews the process. You probably think that if a station tunes to a different frequency and you don't know it, you aren't going to receive their query about whether the frequency is in use. But now think about CW skimmer and the PSK browser. We have the ability to receive and decode an ENTIRE BAND at once. We can see when a frequency is occupied and when it isn't. We can be frequency and time agile. Why should an entire QSO take place on a single frequency? Why not adaptively QSY? Why not beacon, "we are going to be using this frequency for this long." > Heck, when I worked at Western Electric 50 years ago, we were given a > demonstration of multiplexing over the phone lines, and what happened if too > many users tried to get on at the same time - everyone's syllables were > clipped. Of course, this is all covered by queuing theory, as you point out > by the cell phone access control. > Yeah. They have the problem with blocking because they required a 64kbps channel. But look at the Internet. The more people who use it, it just degrades gradually for everyone. Still, the traffic gets through. > > If you can design a mode that does better than what we have today, then do > it! I assure you it will be welcomed and you can show us all how it should > be done properly! > I am considering doing just that. Step one is to investigate the medium and how to build an efficient adaptive link that runs over that medium. It may run as slowly as a few bits-per-second, or as fast as several thousand bps depending on band conditions. It will not require the operator to make that decision. Once I have that I can layer the MAC and the routing. Oh, and it will support kbd-to-kbd. My problem is that, while I am quite good with protocols, I don't have the signal processing and coding background. That was all phy stuff that was done by someone else. But once someone gave me a path to move bits from here to there, then I could make them sit up and beg. Now I want to understand the phy so I can get it right. But right now I am trying to get you guys to stop thinking in the same path. Radios are changing. These changes bring about capabilities we have only just dreamed of. Now they are here. We need to start thinking about taking advantage of them. -- 73 de Brian, WB6RQN/J79BPL
_______________________________________________ fldigi-alpha mailing list [email protected] https://lists.berlios.de/mailman/listinfo/fldigi-alpha
