"Austere Internet"

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

Jim Lord has mentioned he's working with some people on something he calls the "austere internet". They would organize shortwave data links using a new, more powerful data compression scheme. He's also mentioned that he's working with a client which wants to start an Internet bank. And that is all I know about these items. has anyone else heard anything about this?

-- J. Yuma (jy@isarebel.com), March 08, 1999


Hi J.,

Ham operators have wide area bulletin board systems in operation (have been for years). They've also been forwarding messages for decades (long before computers). None of these replace the Internet.

And shortwave radio, even VHF and UHF radio, can't replace the internet, especially the WWW. The amount of data that can be reliably sent over a radiotelephone link is appreciably less than the amount that can be sent over a normal 'phone line.

However, various methods (protocols) are being used by hams for digital communications -- it's just a lot slower than the slowest Internet connection. The latest low speed method is called PSK31, and seems to give the best 'connection' for normal typing speeds (3-4 characters/second). Faster methods (RTTY, Pactor, Amtor) need more equipment than just a computer and a transceiver.

Really reliable systems (Internet-style automatic retransmission of damaged packets) haven't worked out very well.

Dean -- (KB0ZDF, ex-W0ZHG)

-- Dean -- from (almost) Duh Moines (dtmiller@nevia.net), March 09, 1999.

Dean, you wrote: "Faster methods (RTTY, Pactor, Amtor) need more equipment than just a computer and a transceiver."True in some cases, but really all you need is a modem between the tranceiver and radio, whether that be a stand-alone type or the computer's soundcard mimicing one. With most stand-alones, you don't need a computer - a dumb terminal will suffice. PSK31 is a robust protocol, yes, but Pactor/Pactor II have about the same level of efficiency, and under certain band conditions are markedly better than PSK31. Everyone knows that 300-baud AX.25 packet is a waste of time on HF due to hits and selective fading, and indeed those networks don't work very efficiently.

However, there are efficient protocols and systems being used. When you're dealing with Pactor or even Amtor (in linked mode, not FEC), used for example in an HF ApLink node, traffic flows very fast and smoothly, even with relatively poor propagation. The main limitation on digital HF communications rates is the restricted bandwith, which the regs state shall not exceed 3 KHz. That limits the data rate to around 300 baud maximum. In contrast, he military commonly uses much higher data rates (and correspondingly greater bandwidth) in some of its spread-spectrum and multiplexing circuits.

-- sparks (wireless@home.com), March 09, 1999.

Hi Sparks,

I was sorta talking off the top of my head, because I haven't used a digital mode (except PSK31) for a few years.

One thing that's always bothered me is: What is the reason the HF modems don't use multi-phase modulation to increase data content? Just as the landline modems use phase quadrature (and more), why not HF radio links? Is it that there's too much signal distortion that can't be corrected? Or is bandwidth the problem -- it can't be limited to 3kHz and still retain the data?

-- Dean -- from (almost) Duh Moines (dtmiller@nevia.net), March 09, 1999.

Hi Dean. I think if I wasn't using digital HF on an almost daily basis, I'd forget all the little dits and dahs myself :)

About using phasing techniques - a lot has to do with the HF environment. Imagine if the internet had to put up with 'spherics, static, phase distortion, selective fading and variable-degradation circuits... throughput would fizzle. Also, the faster the baud rate, the less noise must exist on the link for optimum data transfer.

The type of keying, whether AFSK, FSK or PSK, has no bearing on the bandwidth of a given transmission - that's determined solely by the baud rate.

Thumbnail sketch, hope it helps.


-- sparks (wireless@home.com), March 09, 1999.

So we bootleg 10 meter AMTOR, via tnc instead of modem, set up a series of NATIONAL nodes, a couple of hardwired auto gateways to twisted pair,and, well, make it work. We get to live within the 300 baud limit, and are forced to actually THINK about what we want to kick out as a packet, rather than, like here, blither er blather, [(dit dit dit dit - dit dit) hi].

I also think that there will be an interesting influx in (on) 10 M, to which the current bootleggers may take exception, particularly due to the "legals'"lack of net op skills. Listened to a 10M boot net lately? It's scary!! I wish my SKYWARN guys had that kind of net discipline. And these guys (and gals) are doing this nationally, thank you to the propagation curves. [Gotta git myself a 10M allmode!!]

Chuck, who learned enough code to get by one evening on his way to Tech, so he could play SKYWARN and NOW finds that he needs to make the time to get back to the code.

de N8NLL (and OCCASIONALLY, the voice of WB8CQR on 6M SKYWARN Backbone in Northern Ohio and NW PA)

-- Chuck, a night driver (reinzoo@en.com), March 09, 1999.

Moderation questions? read the FAQ