Shortwave HF Station: Filling in the gaps

Syed · 2014-11-11T16:45:35.277Z

This might just be one of my favorite threads so far. I’ll try to make this a comprehensive post, but I’m sure I’m forgetting something, so feel free to ping me if I do.

Shortwave is important–really important–and it will definitely be a part of Outernet’s multi-frequency approach. Although the precise frequency list is not yet available, reception of Outernet by a receiver with multiple tuners will likely the following bands:

Ku-band (downcoverted to L-band by an LNB)
L-band
UHF (around 433 MHz)
Shortwave (2 MHz - 30 MHz)
MW/AM (500 kHz - 1700 kHz)
Carrier pigeons

As you may have noticed, Outernet has no allegiance to technology, infrastructure, or modulation scheme. Our goal is to build a universal information service that is accessible to all of humanity. We use whatever hammers and shovels allows us to do this is the fastest and most cost effective way possible.

Surprisingly (at least to me), broadcasting bits by satellites is considerably cheaper than doing the same over a 100 kW shortwave transmitter. The largest private shortwave installation in the country, WRMI in Florida, is very keen to work with Outernet, but we still need to work through the issue of cost, considering that it will take at least four antennas to delivery a global MB. And 100,000 watts of power is a lot of electricity.

If anyone is curious, we would not be using DRM with these transmitters because high-power DRM exciters are actually not available for purchase. Instead, we would essentially be broadcasting audible noise that can be demodulated by even a feature phone. We think it’s possible to manufacturer a $10 audio jack receiver, which would feed the signal to a simple Java app. The benefit of this approach–which is just one of several for shortwave–is to allow feature phones and existing consumer-grade shortwave radios to receive the signal. The downside to this is the bitrate: 60 bps. But over the course of a day, that’s 500 kB, or an entire ebook.

To learn more about this approach, check out the work of Dr Kim Elliott at the VOA. http://voaradiogram.net

To be clear, Outernet is in no way affiliated with VOA. We’re just pointing out some very interesting research that is being done there.

But like I said, this is only one path–and all other paths are definitely on the table.

Syed · 2014-11-11T16:45:53.462Z

cj5 · 2014-11-11T18:19:54.579Z

Indeed, this is my favorite thread as well.

I had no idea you guys were teamed up with VOA. I know that Florida station very well, as I hear it all the time from California. Granted AM transmitters are expensive, and is the main reason why they are not used as much anymore as a broadcast medium, the delivery of data could be cheaper if the data was distributed. Maybe because I am excited to hear that there is an opportunity for shortwave radio to have a place in a modern application, that I am biased towards utilizing it, but I have to agree with @KR0SIV, that SW radio would be a great way to complement the entire Outernet network.

A lot of Hams I communicate with use packet radio, ALE, WSPR, JT65, PSK31, RTTY, and various FSK modes to communicate, and these would be great modes to use as an exchange. In other words, these modes, since they can’t handle lots of data in a short amount of time, would be great for simply supplying a user’s receiver with a status of the network. If propagation does not allow for optimal conditions, supply the receiver with satellite information to link up. If conditions are good enough to go over SW, have the receiver link up to several SW stations. This group is doing a lot of work already in this area http://robust-packet.net/ and the Robust-Packet is a lot faster than HF packet.

What is Outernet’s goal with having a single ground-based transmission point? To me that seems that we’d be limiting ourselves to a single source of transmission, and that’s not what the mission is, right? Ground-based transmissions would not need to derive from a single source, and would not require anywhere near 100kW output. I’ve worked QRP (low power voice and data transmitting) in Ham radio, and you would be surprised what even 2.5 watts could do. If you planted a small unmanned station in each US state running at 100 watts, supplied by a small solar or wind power source, you could easily get data anywhere within North America over HF. Using an AM radio stations infrastructure from a single point with that much power is unreasonable.

Satellite sources are just as important to a well engineered network, but let’s not depend too much on one thing, like you said. Obviously, UHF, L and Ku bands can deliver more bandwidth over a single source, but going with the theme of not having to depend on a single tech infrastructure is the smart way to go, and utilizing as many technologies as possible to perform one task or various ones would definitely be a benefit to Outernet’s vision.

Thanks, for sharing that information.

Syed · 2014-11-11T19:17:20.675Z

cj5:

I had no idea you guys were teamed up with VOA.

To be clear, we have no official working relationship. But they are aware of us and we are aware of them. And I’ve spoken with Dr. Elliott to learn more about what he is doing with existing transmission equipment, cheap shortwave receivers, GFSK, and the open source fldigi

cj5:

What is Outernet’s goal with having a single ground-based transmission point?

The goal is to simply have a global beacon–exactly as you described. It would be ideal if there was a permanent channel that could relay basic system information. If we can do that exact thing at lower cost and better quality of service with a truly global ground network of lower power transmitters, then all the better!

cj5:

I’ve worked QRP (low power voice and data transmitting) in Ham radio, and you would be surprised what even 2.5 watts could do.

How about those qQRP guys? Bouncing signals across the world from and Altoids tin and less than 1W!

Satellite sources are just as important to a well engineered network, but let’s not depend too much on one thing, like you said.

In my dreams, we have autonomous, solar-powered, sea-faring whitespace multicasters. Those broadcast bots are in addition to backpacks on pigeons.

Trevor · 2014-11-11T20:09:47.155Z

Kind of like Firechat for ham radio.

KR0SIV · 2014-11-11T21:45:43.017Z

The nice thing is that a lot of amateur based digital radio modes already have decoders available for both android and iOS. This would allow anyone with a cheap shortwave reciever and a cellphone/tablet to decode our transmissions.

That being said of course a raspi, computer, etc can do this as well.

It would be great to have a SW station for primary SW broadcast, however in addition hams could use the 30meter band (a 10mhz digital only no contest band) to relay the transmissions in parts of the world the major SW station can’t hit.

Hams could also transmit more frequently on a volunteer basis without the cost of the broadcast station.

Trevor · 2014-11-11T22:24:53.821Z

Would it be possible to use this band to extend the range of a local Pillar?

Syed · 2014-11-12T01:37:35.797Z

Can you describe your scenario in a little more detail? I want to make sure I understand it correctly.

One thing to remember, though, is that shortwave just can’t propagate the avalanche of bits that Ku-band can.

KR0SIV · 2014-11-12T09:00:33.649Z

Indeed, we’re talking 300 baud here… more or less depending on the mode and if we use error correction.
So a condensed version would likely be needed for SW/HF

cj5 · 2014-11-12T15:36:19.589Z

@KR0SIV yes! Hams have been busy building their own decoders for many years. =)

I was thinking more along the lines of automating this. If you are aware of the COTHEN system, it has manned stations as well as a majority of unmanned automated relay stations (nodes). I think at most you would need a controller software that would respond to smaller message packets like an ALE call (i.e. similar to WSPR, JT65 modes), and that receiver could ideally push/pull data from the Outernet, and serve info to receivers in the field over a more robust mode (this probably is similar to internet network hubs). The onsite automated stations would house a receiver/transmitter for controls, a receiver/transmitter for data, a tower, an power source, and possibly a file server that mirrors whatever data a primary station sends (this way there is no need to wait for data to be received from the “mothership”, and then be sent to user’s receiver).

The 30 meter band would be a perfect spot to test on.

KR0SIV · 2014-11-12T18:12:47.200Z

I hadn’t heard of the Cothen system, sounds interesting.
It might also be beneficial to have hams setup UHF/VHF nodes in their towns that transmit data pulled from the outernet satellites or the HF modes.

So for example I could have a receive station in my town with a VHF rig about 200’ up on a tower that could then transmit this data over AX25 (much faster mode) where the range could be increased to cover several counties or in the case of using the 6 meter band possible the whole state.

cj5 · 2014-11-12T18:52:54.900Z

Yes. Exactly. These set ups could be leased too (similar to leasing solar panels and wind generators to power agencies). Rent would cover maintenance and such by the amateur radio operator. I would think these stations would be great for being primary nodes to/from satellite based data. I like the AX25 idea.

I know of the COTHEN system because of my experience with using Ham ALE (http://hflink.com/). If you have the proper antenna configuration, a quick tuning apparatus, and a radio, establishing links can be done. This allows for optimal conditions depending on which band propagation is favoring. Using the data modes like packet is great because the system will tell you if conditions are getting worse, data will stop, the software will QSY the radio to a better band, and data will pick up where it left off. I assume COTHEN works in the same fashion.

Syed · 2014-11-16T02:59:17.014Z

What is the total cost of a cheap, long-range and autonomous node?

KR0SIV · 2014-11-17T06:29:27.632Z

Are we talking HF shortwave or long wave UHF/VHF stuff?

For a local town node I’d say under $300 if you have somewhere high up to mount it.

DarkStar · 2014-11-18T00:27:52.345Z

Another U.S. amateur checking in here (KU7PDX)!

I love the idea of supplementing Outernet by utilizing HF (aka shortwave) frequencies. Yep, data is fairly slow with typical technologies (packet radio, PSK31, RTTY, etc), however there is always room for experimentation!

One item that us amateurs should keep in mind though is that the FCC requires that our transmissions be directed toward other amateurs (the “no broadcasting” rule). There wouldn’t be anything wrong though with something like a licensed amateur receiving the satellite signal, then sending a digital “QST” of sorts to other amateurs on HF that would receive the data and then make it available on a wireless access point.

I’ll definitely keep following this thread, thanks!

Syed · 2014-11-21T05:15:50.623Z

Sure, as long as you’re making a good faith effort to speak to other licensed amateurs, I don’t see any issues. And if a few Lanterns happen to be listening in at the same time, well, there’s not much that can be done about that.

cj5 · 2014-11-23T15:55:23.849Z

Does this include power and leasing for the space? I’m sure there are a lot of repeater owners who could provide cost for renting a tower and shelter for equipment.

What is everyone’s thoughts on solar or wind power?

cj5 · 2014-11-23T15:59:57.472Z

Welcome @DarkStar!

Thanks for your input. I think Outernet is looking for suggestions on strategy. The Ham community could offer up a lot of advice on what modes and technologies are available for Outernet to use in supplementing satellite data. I wouldn’t worry so much about FCC regulations on broadcasting as I’m sure Outernet will worry about the licensing needed to broadcast data over HF. I certainly don’t think this community would expect Hams to break the law.

On that note, what steps has the Outernet community taken to ensure that there is band allocated for this network?

Syed · 2014-12-22T23:33:17.343Z

Correct. Neither Outernet nor any volunteers will be breaking any American laws in the countries where we operate.

We will begin HF transmissions in January. The modulation schedule will be MFSK, though we’ll be experimenting with others. Basically, this will be a similar demonstration of the work of Dr. Kim Elliot of Voice of America’s Radiogram service.

We should cover most of African and all of the US. The bitrate will be 60 bps to start out, but we’ll be experimenting with different waveforms to increase the speed.

panamman · 2015-04-03T17:58:20.077Z

Great Discussion Syed,

But it does point out one inevitable fact of efficiently distributing data in a heterogeneous low bit rate environment. Band width is band width. Success lies in how efficiently your users can receive the data they need and heterogeneously redistribute it via another mode! Geo / Meo / Leo Satellite,Meshed WiFi, Cellular, Internet, HF, SD card or smoke signals;)

Ultimately all the data we depend on is meshed via heterogeneous connections and storage at every level of it’s distribution and use. It is thus very important to build efficient meshed delivery and storage mechanism upon which an unthinkable number distribution channels can be efficiently exploited.

Efficiently meshed storage can thus overcome many of the problems your posters have focused on such as the timeliness of large software distributions by incorporating automatic meshed data de-duplication which not only reduces storage requirements but also automatically limits meshed transmissions to only the changed parts of the new distributions to be transmitted.

Yes you “eventually” need to transmit the “whole” distribution for “new” users but that could take place by pre-loading all your library receivers with the big stuff users will need to first get up and running and then subsequently stay up to date.

But that’s the point, once you have real meshed and de-duplicated storage at the core of your service you also have the flexibility to address the very wide range of challenges faced by an even wider range of potential users in uniquely challenged environments that you can not yet predict.

Luckily for you most of these meshed de-duplicated storage capabilities are already available via opensource projects.