[Namaste-dev] Downlink power management
kb5mu at amsat.org
Thu May 29 10:35:48 PDT 2008
One thing that's unclear to me about the clever variable-link-budget downlink scheme.
How is the decision made about how many repetitions (how much power) to use for any given block of data?
In order to make that decision ideally, you need to know the present capability of the most-disadvantaged listening station who might be interested in the data block.
Let's admit, first of all, that this is, in fully generality, impossible. In an ideal amateur radio environment, it should be possible to listen without transmitting. This is not only a time-honored aspect of amateur radio, but also an important part of what makes an amateur radio mode interesting and successful. To build excitement and form a vibrant user community, it's important that users (especially new users, who are likely to have minimal stations) be able to hear what's going on. Our system has no way at all to know who is trying to listen (only) and so it can't allocate power for their benefit.
Suppose we pay attention only to active listeners -- those who are willing and able to make at least a brief transmission to register their interest in the data stream. With that (unfortunate) restriction, perhaps we can design a protocol that solves the power allocation problem. This is where it gets hazy for me. If we're currently transmitting the data stream at high power (many repetitions), then everybody can hear it and it's easy to see how all the interested stations might register the fact that they see excess link margin, and we could reduce power (repetitions). However, it's not so easy to see how to increase power to accommodate a new more-disadvantaged station. The new station won't be able to copy the data stream in order to find out that he'd be interested in it. We could imagine some sort of metadata channel, transmitted at max power, that tries to describe each available data stream, to clue in these prospective new participants. However, I don't think any auto-generated metadata will be sufficient, and it's even less likely that human-supplied metadata will be kept up-to-date and useful.
Another option would be to allow the uplinking station to set the power level manually. If he only wants a point-to-point link with a well-known other station or group of stations, then no problem. But that's not the usual case in amateur radio. He will have to guess, or rather make a policy decision. He can choose to start at low power, and converse only with other stations of similar or better receive capability. (And everybody else will immediately start complaining about elitism and the hate-and-discontent level will skyrocket.) Or, he can choose to start at high power, in order to converse with anybody and everybody. Which is fine as long as there's power to spare, but which kinda defeats the whole clever variable-power downlink scheme.
Whatever kind of allocation scheme we use, it will make mistakes. The incentive for interested operators will be to equip themselves to be able to receive the lowest-power signal. Put up a big dish. Anybody who is not able to do that (whether because of cost, CC&R problems, portability requirements, or whatever reason) will be effectively a second-class citizen. This is not a happy environment, and not really the one we want to promise the users.
I could go on, but I think you can see my concerns. Let's hear some ideas about how this problem ought to be addressed. I am seriously worried that the whole scheme is really clever but not entirely appropriate for amateur radio use. I'd love to be shown that I'm wrong about that.
kb5mu at amsat.org
More information about the Namaste-dev