Possibility of decoding opensky with d*star equipment

[kb1ipd]

I am wondering if anyone knows much about the technical specs of Icom's D-star vs. Opensky. Obviously I realize the two are not compatible systems but on the surface they look remarkably similar.

Both use AMBE for the vocoder and both use GFSK for modulation, with what appears to be a similar IP based protocol for the switching.

The biggest difference I see thus far is that Opensky uses TDMA where as I do not believe that D-star uses any kind of multiple-access. I'm not entirely sure, however Opensky requires TDMA or if it's just an optional feature that the system supports.

What I am basically getting at is how difficult/possible it might be to get d*star decoders, such as the one that can be added to the icon pcr-2500 to decode unencrypted Opensky. If they're similar enough and it's just an issue of a simple TDMA system, then perhaps that could be decoded in software and then fed back to the AMBE decoder. Or even possibly a firmware hack..

Please don't yell at me. Yes I did search and such. I know they're different and I'm sure there's no "simple" way to do it like "Oh just bridge that resistor and there ya go"

Just wondering if... in theory... maybe?

[N_Jay]

IN theory it would be easier to just develop a decoder in software in a PC.

You can get the AMBE decoder by buying a developers kit from DVSI.

[Unitrunker]

Originally Posted by kb1ipd

Just wondering if... in theory... maybe?

OpenSky is essentially a private CDPD network. The original CDPD networks were data only. The voice was analog. The OpenSky designers leveraged that platform to create OpenSky. You can purchase the standards for CDPD from TIA / IHS. Look for the TSB87 and TIA732 series of documents. I don't know which ones are the "right" ones. That won't tell you exactly how the OpenSky packets are formatted. My guess is they use a lot of the old AMPS control channel signalling. The slotted voice channels would be entirely new territory for someone to figure out.

[kb1ipd]

Hmmmm... I don't suppose there's any existing trunking or decoding software that would be able to decode the channel signaling or any or the trunking or identification?

My guess is not. I wonder how much the DVSI development products cost. I tried e-mailing them some time ago to ask about their software-based decoders. They weren't really that helpful, IIRC.

[Unitrunker]

Originally Posted by kb1ipd

Hmmmm... I don't suppose there's any existing trunking or decoding software that would be able to decode the channel signaling or any or the trunking or identification?

The major CDPD networks were shut down two years ago - all that AMPS/ CDPD test equipment became obsolete. I haven't looked but ... eBay? Honestly, I don't expect the old gear to be that helpful. If it provided a raw 19200 bps dump of the data - that would be a start.

[kb1ipd]

Originally Posted by Unitrunker

The major CDPD networks were shut down two years ago - all that AMPS/ CDPD test equipment became obsolete. I haven't looked but ... eBay? Honestly, I don't expect the old gear to be that helpful. If it provided a raw 19200 bps dump of the data - that would be a start.

Yes that's true. However I believe that CDPD has found it's way into various other things, because I recal hearing that UPS uses it for their wireless communications with delivery trucks... perhaps that's been retired too though.

I would have to look into it further to see exactly how it is implimented and such, but I doubt it would be *that* hard to decode the control and track. But I'll have to look into it further. I would doubt that the carrier bandwidth would be more than a sound card - or a virtual sound card in an icom pcr or a winradio or something like that would be able to sample just fine. I could be wrong though...

[N_Jay]

Originally Posted by kb1ipd

Yes that's true. However I believe that CDPD has found it's way into various other things, ..

CDPD and CDPD protocols are two different things.

CDPD is how packet data was sent over the AMPS cellular network.

The protocols are fairly good so lots of people have reused parts for other data systems. The best known is probably OpenSky which used a lot (but not all) CDPD protocols.

Originally Posted by kb1ipd

. . . because I recal hearing that UPS uses it for their wireless communications with delivery trucks... perhaps that's been retired too though.
. .

No it is not retired, it IS the initial implementation of OpenSky from before Tyco-M/A-COM purchased them.
It is still in use (and will be for a good long time, I would expect.)

[Wyandotte]

A curious question:
Let's just say there was some sort of a way to decode the audio with some sort of software package on a computer through a reciever. Wouldn't there also be a problem with the frequency not "sitting still" long enough to decode it? It's my understanding that OpenSky is moving around a band of freqs kinda like spread spectrum. Am I wrong?

[N_Jay]

Originally Posted by Wyandotte

A curious question:
Let's just say there was some sort of a way to decode the audio with some sort of software package on a computer through a reciever. Wouldn't there also be a problem with the frequency not "sitting still" long enough to decode it? It's my understanding that OpenSky is moving around a band of freqs kinda like spread spectrum. Am I wrong?

It is not spread spectrum, but it is cellular so you will see conversations handed off to other cells, sectors or slots fairly often.

[kb1ipd]

Originally Posted by N_Jay

It is not spread spectrum, but it is cellular so you will see conversations handed off to other cells, sectors or slots fairly often.

That's what I figured. Presumably then, you could at least get started on decoding the transmisson before you have the whole channel and control thing figured out.

But that brings me back to my initial question: If dstar uses GMSK, but I'm not sure exactly how opensky is modulated. The opensky control data, I *think* is multiplexed with the voice, but I;m not sure of that. I'm also not sure if a D-Star decoder will only work on 1.2 ghz, which is the spessified operating channel or if it might actually work on other frequencies.

But in any case, I'm assuming they're not similar enough that the D-star dectoder would pickup opensky voice without modification.

So what is the difference?

[N_Jay]

Originally Posted by kb1ipd

That's what I figured. Presumably then, you could at least get started on decoding the transmisson before you have the whole channel and control thing figured out.

But that brings me back to my initial question: If dstar uses GMSK, but I'm not sure exactly how opensky is modulated. The opensky control data, I *think* is multiplexed with the voice, but I;m not sure of that. I'm also not sure if a D-Star decoder will only work on 1.2 ghz, which is the spessified operating channel or if it might actually work on other frequencies.

But in any case, I'm assuming they're not similar enough that the D-star dectoder would pickup opensky voice without modification.

So what is the difference?

There are more differences than similarities.

Yes you could use the pieces of a d-star radio as your stating point but you would still need a sold enough understanding that you could probably start with raw parts and get to the end just as quickly.

Think of it in layers.
the vocoder is just one layer in the network. There are many other layers you have to figure out to decode the transmission.

I am not sure if OpenSky uses any base level of encryption in the protocol. Since it is based on CDPD, and CDPD was natively encrypted, you might never get enough information to decode the signal even if you had working hardware for each module/layer.

[Ray_Air]

You guys are just wasting your time. Open Sky uses end-to-end 192-bit AES encryption. Yes, every transmission as it is formatted into blocks is encrypted to a very high level. If your city, county or state goes Open Sky then you are SOL. Even if you bought an M/A-Com radio that could do Open Sky you would not be able to listen in. As if the full time system encryption wasn't enough, every radio must first be authenticated by the system to be allowed into the system. This goes for the receive portion to, not just TX.

I live in a county that is trying to go to Open Sky countywide. Thankfully ,for the last 8 years they have had nothing but problems and less than a handful of agencies have made the switch. Legitimate authorized radios for the system are all password protected and I am sure they can be turned off real easy if they are lost or stolen.

[N_Jay]

Originally Posted by Ray_Air

You guys are just wasting your time. Open Sky uses end-to-end 192-bit AES encryption. Yes, every transmission as it is formatted into blocks is encrypted to a very high level.

Really Ray?
The references I have found show AES to be optional.
OpenSky may have an encryption layer running all the time from its CDPD roots, but I doubt it is AES.
Do you have any documentation or are you just speaking from your position as a Know-It-All?

Originally Posted by Ray_Air

If your city, county or state goes Open Sky then you are SOL. Even if you bought an M/A-Com radio that could do Open Sky you would not be able to listen in. As if the full time system encryption wasn't enough, every radio must first be authenticated by the system to be allowed into the system. This goes for the receive portion to, not just TX.

Yep, to get an OpenSKy radio to receive it must register.
That would not prevent a person with sufficient equipment and knowledge (I would almost guarantee no one here) from being able to receive and decode the signal off the air provided neither a base encryption nor the OPTIONAL AES is used.
Of course like every cellular architecture systems all you would receive is parts of transmissions unless you are lucky enough to have a unit on the talkgroup you want always registered on the site you are on.

Originally Posted by Ray_Air

I live in a county that is trying to go to Open Sky countywide. Thankfully ,for the last 8 years they have had nothing but problems and less than a handful of agencies have made the switch. Legitimate authorized radios for the system are all password protected and I am sure they can be turned off real easy if they are lost or stolen.

Yep.

[KA1RBI]

If possible, can someone post a sample capture of a confirmed open-sky signal?

Ideally, should be captured at a sample rate of 48,000 (at least) using the standard discriminator tap - using a setup that has been verified to work with other high-speed (4800 or 9600) data...

Thx

Max

[N_Jay]

Originally Posted by KA1RBI

If possible, can someone post a sample capture of a confirmed open-sky signal?

Ideally, should be captured at a sample rate of 48,000 (at least) using the standard discriminator tap - using a setup that has been verified to work with other high-speed (4800 or 9600) data...

Thx

Max

And why would you think that is Good Enough to represent 19,200 GMSK?

[KA1RBI]

Originally Posted by N_Jay

And why would you think that is Good Enough to represent 19,200 GMSK?

19,200 baud or 19,200 bps ?

It may indeed not be Good Enough.

I could ask too for a 250000 sps capture from something like a USRP... ;-)

Max

[N_Jay]

Originally Posted by KA1RBI

19,200 baud or 19,200 bps ?

It may indeed not be Good Enough.

I could ask too for a 250000 sps capture from something like a USRP... ;-)

Max

I believe it is 19200 BPS GMSK.

I believe they also have a version that runs 9600 GMSK (for 900 MHz)

Why not just get a GMSK receiver and find a signal yourself.

You will get much better base-line data.

[Squad10]

Originally Posted by N_Jay

I believe it is 19200 BPS GMSK.

I believe they also have a version that runs 9600 GMSK (for 900 MHz)

Why not just get a GMSK receiver and find a signal yourself.

You will get much better base-line data.

If it's 9.6 or 19.2 kbs GMSK, and you want to tink, consider the CalAmp/Dataradio T-96SR. I've used its modem section to transmit and receive GPS RTK data over 800 Mhz conventional frequencies. It uses the discontinued Motorola Data Set Interface chip for sync/async conversion that works really well.

If you get into P25 data decoding, the Creative Labs Sound Blaster X-Fi card will work.