RFC on HF Protocol

Tim Wescott <tim@seemywebsite.com> writes:

> As it stands the sentence is quite meaningful, if you think about
> it. It means that you have to understand the population you're
> proposing to design for, and do what they want and expect.

I agree with your sentiment, Tim. 

These folks have come seeking guidance, and many here are very
qualified to provide it. This is a great opportunity for the denizens
of comp.dsp to show their good will, and it sounds like it would be
a heckuva lotta fun as well.
-- 
%  Randy Yates                  % "Watching all the days go by...    
%% Fuquay-Varina, NC            %  Who are you and who am I?"
%%% 919-577-9882                % 'Mission (A World Record)', 
%%%% <yates@ieee.org>           % *A New World Record*, ELO
http://home.earthlink.net/~yatescr

Tim Wescott wrote:
> Vladimir Vassilevsky wrote:
> 
>>
>>
>> Jerry Avins wrote:

   ...

>>>> The first question to answer is what is this protocol intended for. 
>>>> I.e. what is the concept of the communication system.
>>>> All other stuff is the minor technical details. Those details depend 
>>>> on the answer to the first question.
>>>
>>> I think a sufficient answer to that question is (with a little added 
>>> imagination) given in the first sentence at the URL provided:
>>>
>>> "The ARRL is seeking comments from amateurs concerning development of 
>>> an open-source (non-proprietary) data communications protocol 
>>> suitable for use by radio amateurs over high-frequency (HF) fading 
>>> paths."
>>
>>
>> This is a meaningless sentence.
>> The 2..3 receivers with diversity antennas is all that you may need.
>>
>> VLV
>>
>>
> If you can't take vague requirements and expand them into something 
> useful without pissing people off, how do you stay in business?
> 
> As it stands the sentence is quite meaningful, if you think about it. It 
> means that you have to understand the population you're proposing to 
> design for, and do what they want and expect.
> 
> This means that you need a box that will connect to a transceiver that's 
> connected to an antenna, and have the system work well.  You can put 
> some requirements on turn-around time, but probably nothing tighter than 
> AMTOR.  You must have a system that will  go into a box that costs less 
> than $500, and preferably less than $200 -- remember that the average 
> radio amateur will spend a man day to save $1, and be proud of it. 
> Something that would just plug into a sound card would be better yet, 
> from this perspective.  The system has to be plug and play for the folks 
> who passed their tests by memorizing all possible answers to the 
> questions, yet should also provide enough flexibility that experts can 
> get significantly better performance.  It also has to comply with all 
> applicable FCC regulations, some of which are quite sensible and some of 
> which are absurd.  If the system were robust to having someone naively 
> (or maliciously) transmit a carrier smack in the middle of the signal 
> that would be good, but not essential.
> 
> There's more, but you should be getting the idea.  I suspect that if you 
> really wanted to make a project out of this you'd need to take that 
> "data sheet" definition from the ARRL and turn it into a number of 
> criteria such as I've outlined, agree on them, the work on turning 
> _those_ into numbers, if they're not obvious as they stand.

Tim,

You fleshed out very well what I had in mind when I wrote "a little 
imagination".

Jerry
-- 
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Tim Wescott wrote:

>>>> The first question to answer is what is this protocol intended for. 
>>>> I.e. what is the concept of the communication system.

The one who can't clearly define what he wants should not be surprised 
if he will get something incomprehensible as the result.

> This means that you need a box that will connect to a transceiver that's 

Is it a conventional SSB transceiver?

> connected to an antenna, and have the system work well.

Work what? Voice, data, video, ethernet, internet, windows, linux, QPSK, 
OFDM, opensource, for free and all at once? This is how their RFP looks 
like. Duplex or simplex? Broadcast or point-to-point?  Can the GPS be 
used as the external timing reference? Is it intended for DX or it 
enough to reach the neighbor's house?


BTW, a long time ago I was thinking about the digital voice radio for 
amateur communication. At the marginal conditions, the 1200 bps vocoder 
should be better than the analog SSB. I would be surprised if it is not 
standardized yet.


   You can put
> some requirements on turn-around time, but probably nothing tighter than 
> AMTOR.  You must have a system that will  go into a box that costs less 
> than $500, and preferably less than $200 -- remember that the average 
> radio amateur will spend a man day to save $1, and be proud of it.

  The inexpensive DSP eval boards is in this price range.

> Something that would just plug into a sound card would be better yet, 
> from this perspective.  The system has to be plug and play for the folks 
> who passed their tests by memorizing all possible answers to the 
> questions, yet should also provide enough flexibility that experts can 
> get significantly better performance.  It also has to comply with all 
> applicable FCC regulations, some of which are quite sensible and some of 
> which are absurd. 

To my knowledge, it is not allowed to transmit any non-voice signals via 
the voice path of a radio unless specifically permitted. They will have 
to push whatever new protocol through the FCC.

  If the system were robust to having someone naively
> (or maliciously) transmit a carrier smack in the middle of the signal 
> that would be good, but not essential.
> 
> There's more, but you should be getting the idea.  I suspect that if you 
> really wanted to make a project out of this you'd need to take that 
> "data sheet" definition from the ARRL and turn it into a number of 
> criteria such as I've outlined, agree on them, the work on turning 
> _those_ into numbers, if they're not obvious as they stand.


Here is the second question: how much time and money do they have?


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

"Vladimir Vassilevsky" <antispam_bogus@hotmail.com> wrote in message 
news:Nj9Eh.468$tv6.147@newssvr19.news.prodigy.net...
>
>
> Tim Wescott wrote:
>
>>>>> The first question to answer is what is this protocol intended for. 
>>>>> I.e. what is the concept of the communication system.
>
> The one who can't clearly define what he wants should not be surprised if 
> he will get something incomprehensible as the result.
>
>> This means that you need a box that will connect to a transceiver that's
>
> Is it a conventional SSB transceiver?
>
>> connected to an antenna, and have the system work well.
>
> Work what? Voice, data, video, ethernet, internet, windows, linux, QPSK, 
> OFDM, opensource, for free and all at once? This is how their RFP looks 
> like. Duplex or simplex? Broadcast or point-to-point?  Can the GPS be used 
> as the external timing reference? Is it intended for DX or it enough to 
> reach the neighbor's house?
>
>
> BTW, a long time ago I was thinking about the digital voice radio for 
> amateur communication. At the marginal conditions, the 1200 bps vocoder 
> should be better than the analog SSB. I would be surprised if it is not 
> standardized yet.
>
>
>   You can put
>> some requirements on turn-around time, but probably nothing tighter than 
>> AMTOR.  You must have a system that will  go into a box that costs less 
>> than $500, and preferably less than $200 -- remember that the average 
>> radio amateur will spend a man day to save $1, and be proud of it.
>
>  The inexpensive DSP eval boards is in this price range.
>
>> Something that would just plug into a sound card would be better yet, 
>> from this perspective.  The system has to be plug and play for the folks 
>> who passed their tests by memorizing all possible answers to the 
>> questions, yet should also provide enough flexibility that experts can 
>> get significantly better performance.  It also has to comply with all 
>> applicable FCC regulations, some of which are quite sensible and some of 
>> which are absurd.
>
> To my knowledge, it is not allowed to transmit any non-voice signals via 
> the voice path of a radio unless specifically permitted. They will have to 
> push whatever new protocol through the FCC.

I'm not sure what you mean by this.  Most packet radio boxes connect to the 
radio's mic and headphone.

>
>  If the system were robust to having someone naively
>> (or maliciously) transmit a carrier smack in the middle of the signal 
>> that would be good, but not essential.
>>
>> There's more, but you should be getting the idea.  I suspect that if you 
>> really wanted to make a project out of this you'd need to take that "data 
>> sheet" definition from the ARRL and turn it into a number of criteria 
>> such as I've outlined, agree on them, the work on turning _those_ into 
>> numbers, if they're not obvious as they stand.
>
>
> Here is the second question: how much time and money do they have?
>
>
> Vladimir Vassilevsky
>
> DSP and Mixed Signal Design Consultant
>
> http://www.abvolt.com 

Philip Martel wrote:

>>To my knowledge, it is not allowed to transmit any non-voice signals via 
>>the voice path of a radio unless specifically permitted.  They will have to
>>push whatever new protocol through the FCC.
>  
> I'm not sure what you mean by this.  Most packet radio boxes connect to the 
> radio's mic and headphone.

It does not imply one can connect anything to the mic input.
Those are the FCC allowed boxes which operate in the permitted packet 
mode. And the operation is allowed only in the dedicated parts of ham 
radio bands, and many other restrictions of FCC apply.


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

Let me address some of the issues with some broad generalizaations.

1. Hams tend to have one of two types of tranceivers. The first is 
commercially built 100 watt multi-mode (AM, FM, WFB, SSB, CW). More recent 
versions of these will have line input/output for connecting to soundcards. 
These run into the $100s $1000s to  The second is > 5W( called QRP rigs) CW, 
although a few are SSB, that are home built or from kits. These are >$50. If 
a digital mode could be made to work with transcievers similar to the QRP 
rigs it would be interesting for emergency communications (EMCOMM) since 
they are much more portable and power hungry. Here you get into something 
like the VSAT where a rig of the first type with a good fixed base antenna 
mioght communicate with QRP rigs and temporary antennas.

As an aside, hams have provided emergency communications for nearly 100 
years when needed. Often this was for government agencies but sometimes 
non-government organizations (NGOs). A lot of the push for EMCOMM is for 
digital communications to provide hard copy. It may be that NGOs and 
individuals will be supported more in the future than government agencies. A 
lot depends, in the US, on how well the communication infrastructure 
upgrades work out.

2. The main focus on this RFC IMO is for computer to computer digital 
communications. There are a number of viable chat modes on HF, some even 
adapted to slow digital communications, e.g. PSKMail based on PSK31 chat 
mode. For this DSP group the focus IMO would be on reliably getting a bit 
stream from one computer to another at the highest throughput possible. 
Start from the 3k Hz in the RFC but also determine how much better something 
having a larger bandwidth would work.

3. Digital modes are allowed in all ham frequency bands as long as the 
protocol is publically documented. The documentation does not even need to 
be sufficient to implement the mode, i.e. Pactor 3. Basically as long as the 
FCC could possibly obtain what is necessary to monitor the transmission it 
would be legal. (But IANAL...)

4. Cost point is somewhere around $100-200 for a stand-alone modem. Using a 
PC sound card is acceptable and somewhat desirable for EMCOMM portable 
operation using a laptop. If a more expensive USB external sound card  <$100 
provides better operation that would okay.

5. Going beyond the transfer of the bit stream, if interested, opens up a 
lot of questions. Point- to-point using the VSAT hub model is feasible, 
especially if the hub can drop data onto the Internet. Point-to-multipoint 
would be nice to allow a mesh type network. One question needing a good 
analysis and explanation is the trades offs between automatic repeat request 
(ARQ) operation and the use of FEC in combination or separately. Included in 
that is fall-back modes to provide to reliable but lower throughput as HF 
conditions deteriorate.

It is realistic to assume that the ARRL headquarters hams are not conversant 
with the techniques, limitations, capabilities, or realistic possibilites of 
HF digital communications. Thus the RFC is vague because they do not know 
the appropriate questions to ask. There are hams that could formulate those 
questions but they are not always heard by the ARRL.  It is a major step 
forward that the ARRL posted the RFC.

-- 

Rud K5RUD
ARES AEC Montgomery County, TX
http://over50.k5rud.us/wiki/



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Rud Merriam wrote:

> Let me address some of the issues with some broad generalizaations.

[...]


So, it is desired to have a wireless network with the functionality 
somewhat similar to 802.11, however it should use the ham transceivers.

If the communication has to go through the voice path, then not much can 
be gained by the advanced processing. A common QAM should probably be 
used, with refinements for the half duplex mode of operation. In the 
ideal conditions, the throughput could probably reach 20kbps, however 
the realistic number would be around 4.8kbps or below. With all 
associated overhead, the channel traffic would be about half less then 
the raw data rate.


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

Why is the voice path a limiting factor? If it were not how much improvement 
is possible?

There is a great deal of interest in SDR, including down in the QRP 
transceiver kit range, so it is feasible to do direct modulation. Even 
getting into the IF of some commercial rigs is possible since hams can 
legally modify gear.

-- 

Rud K5RUD
ARES AEC Montgomery County, TX
http://over50.k5rud.us/wiki/


"Vladimir Vassilevsky" <antispam_bogus@hotmail.com> wrote in message 
news:G9tEh.1076$BE2.741@newssvr27.news.prodigy.net...
>
>
> Rud Merriam wrote:
>
>> Let me address some of the issues with some broad generalizaations.
>
> [...]
>
>
> So, it is desired to have a wireless network with the functionality 
> somewhat similar to 802.11, however it should use the ham transceivers.
>
> If the communication has to go through the voice path, then not much can 
> be gained by the advanced processing. A common QAM should probably be 
> used, with refinements for the half duplex mode of operation. In the ideal 
> conditions, the throughput could probably reach 20kbps, however the 
> realistic number would be around 4.8kbps or below. With all associated 
> overhead, the channel traffic would be about half less then the raw data 
> rate.
>
>
> Vladimir Vassilevsky
>
> DSP and Mixed Signal Design Consultant
>
> http://www.abvolt.com
> 



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Rud Merriam wrote:

   ...

> 5. Going beyond the transfer of the bit stream, if interested, opens up a 
> lot of questions. Point- to-point using the VSAT hub model is feasible, 
> especially if the hub can drop data onto the Internet. Point-to-multipoint 
> would be nice to allow a mesh type network. One question needing a good 
> analysis and explanation is the trades offs between automatic repeat request 
> (ARQ) operation and the use of FEC in combination or separately. Included in 
> that is fall-back modes to provide to reliable but lower throughput as HF 
> conditions deteriorate.

I can't imagine how one might use an ARQ protocol in a 
point-to-multipoint mesh. Has tha been thought through?

   ...

Jerry
-- 
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Rud Merriam wrote:

> Why is the voice path a limiting factor?

It is impossible to gain anything after the FM or AM detector. As for 
the SSB, the bandwidth is too narrow to be useful, and the PAPR is very 
limited too.

> If it were not how much improvement 
> is possible?

It depends. Generally, at the order of several dB.

> There is a great deal of interest in SDR, including down in the QRP 
> transceiver kit range, so it is feasible to do direct modulation. Even 
> getting into the IF of some commercial rigs is possible since hams can 
> legally modify gear.

To get a considerable improvement, the radio has to use the spread 
spectrum modulation. For DX short wave communication, the bandwidth of 
several dozen kHz would be enough. For urban communication, somewhat 50 
MHz is good.

Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com