Anyone know the specs for this protocol?

Status
Not open for further replies.

KE7IZL

Member
Joined
Aug 2, 2011
Messages
146
Location
Seattle, WA
The one called GW Pactor or GW Dataplex. I hear it's from Global Wireless and from help I received in a previous thread I believe it may very well be the digital mode I've been hearing. If someone can give me the specs for the protocol GW Pactor or GW Dataplex, I'd be very happy, cause then I think I might be able to do something to decode the signal I heard.
 

ka3jjz

Wiki Admin Emeritus
Joined
Jul 22, 2002
Messages
25,361
Location
Bowie, Md.
While it's possible, it's also quite possible that this is proprietary information, and as such, GW may not be so willing to share the specifications. I'm not going to hold my breath

73 Mike
 

KE7IZL

Member
Joined
Aug 2, 2011
Messages
146
Location
Seattle, WA
While it's possible, it's also quite possible that this is proprietary information, and as such, GW may not be so willing to share the specifications. I'm not going to hold my breath

73 Mike

How do companies like Hoka get the specs needed for their decoder? And what's to prevent them from redistributing the specs?
 

ka3jjz

Wiki Admin Emeritus
Joined
Jul 22, 2002
Messages
25,361
Location
Bowie, Md.
They *may* pay for the rights, then very likely need to sign a NDA (non disclosure agreement) to protect the source code. And I would bet the cost ain't cheap (ka-ching!)...73 Mike
 

Token

Member
Joined
Jun 18, 2010
Messages
2,379
Location
Mojave Desert, California, USA
GW Pactor (and GW Dataplex and GW OFDM) was developed specifically for Global Wireless. I doubt they did it in-house, but that is possible. Regardless, they own it. As such terms like "proprietary", "intellectual property rights", "licensing", "prior approval of disclosure", and "non disclosure agreement" are probably going to accompany any official transfer of information on the specifics of its structure and operation.

It is possible that reverse engineering it to determine its attributes, particularly if followed by distribution of that information, would also be a violation of ownership.

T!
 

KE7IZL

Member
Joined
Aug 2, 2011
Messages
146
Location
Seattle, WA
GW Pactor (and GW Dataplex and GW OFDM) was developed specifically for Global Wireless. I doubt they did it in-house, but that is possible. Regardless, they own it. As such terms like "proprietary", "intellectual property rights", "licensing", "prior approval of disclosure", and "non disclosure agreement" are probably going to accompany any official transfer of information on the specifics of its structure and operation.

It is possible that reverse engineering it to determine its attributes, particularly if followed by distribution of that information, would also be a violation of ownership.

T!

Just who is the creator of GW modes? Is it actually Global Wireless? Because I notice the only place with similar technology (radio equipment for ships) and with a similar name, is actually Globe Wireless.

A license agreement states that you may not reverse engineer something (like stated in the the EULA that accompanies Microsoft or other "brand name" software). However as long as one does not own any Globe Wireless hardware or software then they aren't bound to any license agreement. Also there is no mention (as far as I know) of reverse engineering being inherently illegal in any "intellectual property" laws. The laws do allow enforcement of a contract (the license agreement) between the owner of the intellectual property (which is the hardware or software system in question) and user of software (or hardware), but those laws do not themselves specifically state that reverse engineering is an illegal activity.
 
Last edited:

Token

Member
Joined
Jun 18, 2010
Messages
2,379
Location
Mojave Desert, California, USA
Just who is the creator of GW modes? Is it actually Global Wireless? Because I notice the only place with similar technology (radio equipment for ships) and with a similar name, is actually Globe Wireless.

I suspect the auto complete on my spell check added the "al", I meant Globe Wireless. The documentation I have on the format says variously "developed by" and "developed for", but it all says "is the property of".

A license agreement states that you may not reverse engineer something (like stated in the the EULA that accompanies Microsoft or other "brand name" software). However as long as one does not own any Globe Wireless hardware or software then they aren't bound to any license agreement. Also there is no mention (as far as I know) of reverse engineering being inherently illegal in any "intellectual property" laws. The laws do allow enforcement of a contract (the license agreement) between the owner of the intellectual property (which is the hardware or software system in question) and user of software (or hardware), but those laws do not themselves specifically state that reverse engineering is an illegal activity.

I did not specifically mention a license agreement in relationship to reverse engineering. But yes, reverse engineering in face of a EULA or EUC is almost universally not allowed in situations as have been described. What I did say when I mentioned reverse engineering was "a violation of ownership", by ownership I assume some kind of software or hardware, or the specifications themselves, have been procured from Globe Wireless. And either one would likely prevent Hoka or anyone else with such access from distributing the information.

You asked how Hoka and companies like that got the specs so that they could build decoders, and what is to prevent them from distributing it. I have never heard confirmed specifics on how they get it, but I would think there was a strong possibility that for certain hard to define formats they get it from the originators of the format, and if that is the case they might very possibly be bound by some sort of agreement.

T!
 

KE7IZL

Member
Joined
Aug 2, 2011
Messages
146
Location
Seattle, WA
So basically these specs are kept by their originator as a carefully guarded "company secret" (like the formula for the soda Cocacola)?
 

OCO

Member
Joined
Jul 17, 2011
Messages
928
Location
Central Michigan
Companies spend money to develop protocols to match their business needs. This becomes intellectual property just like the software that you create. If there is a monetary advantage to sharing this IP, they may listen to proposals - but typically a reason that's something like "so I can decode your business traffic" or "so I can set up a system using your encoding/decoding" usually won't get you into their boardroom ,,,,
 

n5ims

Member
Joined
Jul 25, 2004
Messages
3,993
The specs may also be protected by one or more patents to prevent anyone from legally reverse engineering them.
 

KE7IZL

Member
Joined
Aug 2, 2011
Messages
146
Location
Seattle, WA
Actually it's not business traffic for the company itself as much as it is maritime communication. So there's several different places in the world with becon channels (things that transmit the station name indefinitely) it seems. As the pattern for each one seems to to be the same from one packet to the next (suggesting it is an ID signal using this mode, much like CW ID uses morse code to send a call sign). It seems these are all over the place, and (from what I can tell) COMPLETELY replaced CW ID station markers.

It used to be that WEFAX on HF stations would transmitt some time but when not in use (durring times that no TX of weather maps was scheduled) would have a CW ID. It would sould like 4 packets of a warbling tone (like a carrier FM modulated by a pure sine wave of maybe 30Hz or so and occupied about 200Hz bandwidth. and lasting about 1 second each). After 4 consecutive such tone bursts (which clearly contained no data) you'd hear a 3 letter CW ID. This cycle repeated until it was time to TX some more maps Which means on a given station you'd hear this warble warble warble warbble CW ID warble warble warble warble CW ID, over and over again for a couple hours. This was a couple years ago.

Over time I noticed these 4xWarble+CWID station marker signals slowly stopped being used and instead I've heard the above mentioned GW Dataplex signals. It seems the old ID transmitters got replaced with newer tech. Also I began to hear the GW Dataplex all over Marine bands (not just on the WEFAX stations' frequencies). It seems both ships and land stations of all types (basically all of maritime communication) is being now conducted with GW Dataplex (except for a few maritime band signals using encrypted 75baud 850hz spaced ASCII RTTY transmisions, which are now encrypted).

I've not hear the 4xWarble+CWID station markers now for a couple years!
I guess GW Dataplex has become synonymous with maritime communications recently.
 

OCO

Member
Joined
Jul 17, 2011
Messages
928
Location
Central Michigan
From the DXSOFT Yahoo group:


Re: [dxsoft] mode suggestion

Henrik,

I believe only the Wavecom decoders for government use can decode GW Pactor
itself.

The channel free markers which can be decoded by Hoka and Skysweeper are not
really very interesting - mainly because all they have is a single byte ID
and no-one has properly correlated these to known coast stations.

Here is all I have on it if anyone is interested in taking it further:


Globe Wireless Marker Decoding
==============================

Globe Wireless provides HF marine traffic via coast stations.
They use a proprietary modem which sounds similar to Pactor.

Protocol Description

Unfortunately, the protocol they use on HF is a proprietary one for which
few decoders are available. The signal uses at least two different
modulations: 100 baud, 200Hz shift synchronous FSK and PSK.

Globe Wireless seems to send at least 4 distinct kinds of traffic from their
stations:

*Channel Free Marker
ARQ Data (speculative)
Selcall or FSK data (speculative)
PSK data

*Decoding the Channel Free Marker
There is a repetitive channel free marker broadcast when a channel is idle.
This is in 100bd 200Hz shift FSK.

The total number of bits sent in the channel free marker is 231.
The first 98 bits compose this header:

10100101001000000011100010100011000111110110110000110101001110110100101111110010\
111100101111001011

Following this, an 8 bit quantity repeats 6 times - a form of FEC.

The byte corresponds to the following table:

0x47 HLF, Seoul, South Korea
0x49 KEM, Nikolski, Alaska, USA
0x4e VCS, Halifax, Nova Scotia, Canada / VCT, St. Johns, Newfoundland,
Canada
0x5d KEJ, Honolulu, Hawaii, USA
0x5e CPK, Santa Cruz, Bolivia / VCT, St. Johns, Newfoundland, Canada
0x5f A9M, Hamala, Bahrain
0x63 9HD, Malta
0xc3 XSV, Tianjin, China
0xc9 VIE, Darwin, Australia / ZLA, Awanui, New Zealand
0xcc HEC, Bern, Switzerland / 9MG, Penang, Malaysia / LFI, Rogaland, Norway
0xd2 ZSC, Capetown, South Africa
0xd7 KPH, San Francisco Radio, San Francisco, California, USA / VJS, Perth,
Australia
0xd8 WNU, Slidell Radio, New Orleans, Louisiana, USA
0xdb KHF, Agana, Guam
0xdc KFS, Palo Alto, California, USA
0xdd LSD836, Buenos Aires, Argentina
0xde SAB, Goeteborg, Sweden
0xe3 8PO, Bridgetown, Barbados

There are other stations with unknown bytes.
E.g., 0x3b possibly Pt. Reyes, California, USA.
More work is needed to match all the coast stations.


Credit to Ben Mesander
http://tiki-lounge.com/~ben/radio/gw.html
 

KE7IZL

Member
Joined
Aug 2, 2011
Messages
146
Location
Seattle, WA
From the DXSOFT Yahoo group:

Thanks for the info, that could be VERY helpfull. Just 1 more question. Is the 6 repeat FEC bytes a repeat of the station id byte? Or what byte is it repeating?

Also even if you have 98 bits that is NOT divisible by 8. So I must ask just HOW are the bytes divide up out of this stream of 98 bits? You seem to know enough to know that 98 bits compose the header. But you seem to have gotton something wrong. You claim it's 98 bits out of 231 bits. However I know the 231 bit count is bogus. I myself have used a free program called ABox2 to FSK demodulate it into a bit stream and I counted 152 bits by putting the resulting text file (each line is one bit) into MS Excel and noting how many lines it takes (easily read off of the numbers on the rows of MS Excel).
 
Status
Not open for further replies.
Top