Replies to This Discussion

Permalink Reply by Jake Stew on July 20, 2012 at 5:14pm

I wouldn't invest in old tech when the new generation of radios (like the 3DR) are so much better in every spec and cheaper to boot.

There are 1W 3DR type radios coming out that will perform a lot better than even the Xtend.

At 9.6kbps you're not going to be happy.

• ▶ Reply

Permalink Reply by Justin on July 20, 2012 at 8:46pm

Do you know what the ETA is for the 1W 3DR? What about things like AES128, I can't find anything for the 3DR that say's that or similar is supported?  How does the 3DR handle interference compared to the Xbee?

I'm just hesitant going for a platform I haven't used when I know the Xbee has and will perform as expected as I've used them before.

I assumed that 9.6k might not be enough, and was happy enough to live with the pro 900, but if you know of a good source of info for the 3DR I may investigate (Especially if they have a 1w unit coming out).

• ▶ Reply

Permalink Reply by Jake Stew on July 21, 2012 at 12:00am

http://diydrones.com/forum/topics/rfd900-new-long-range-radio-modem

http://rfdesign.com.au/index.php/rfd900

http://store.rfdesign.com.au/rfd-900-radio-modem/

Check out the last link.  Looks like it's ready to go now.

If you're using this with the APM 3DR is the way to go for sure.  Lots of extra features and special support for them in MP.

The Si100x chips can do up to AES256, they actually have a lot more processing power than the APM itself. However this isn't implemented yet in the firmware we're using.

I wouldn't be too worried about data interception or hostile takeover unless you think the FBI/CIA/military/etc. is trying to go after you.  They'd need the same radio and the all same settings to get your signal, unless they had a lot of fancy gear and knew what they're doing.

IMHO the current firmware hasn't even scratched the surface of what the chip can do.  So you can expect the radio to get better and better as the firmware matures, and they already work great.

You should note that 1W is the max legal ISM output, and since they have the best receive sensitivity of any radio on the market... they have the max range you can get in a legal transmitter.

• ▶ Reply

Permalink Reply by Justin on July 21, 2012 at 12:24am

Alright, I'm sold on at least giving them a shot, if I hate them I can always go back to Xbee (Mainly sold over on the fact I can get a 1W unit with the price in double digits and that they can have the processing power to do encryption)

I'd like AES on them, not because I fear that some XYZ department wants to monitor my use, but just from a interception/takeover/interference point of view, to me it would seem quite critical that encryption is enabled as just hoping that no one else has the same/similar unit does not seem like a safe thing to do for a cautious person like me, especially when they have such a long range which would increase the chance of the above.

I'm guessing by your interest in the radio's that you have one of these?  Would you be able to tell me what else I need to connect this to an APM2 board? (It has a lot more pins than I'm used to and outside of the TX/RX/Power ports I'm not entirely sure what else is required for that to function)

Also would you happen to know when encryption may be on the timeline?

• ▶ Reply

Permalink Reply by Jake Stew on July 21, 2012 at 2:46am

Tridge or Seppo would be the ones to ask about the encryption timeline.

As far as connecting, the 3DR radio just uses the 4 standard pins.  The extra pins on the RFD900 are just pins connected to the processor.  The SiLabs chip has a really nice feature called a digital crossbar.  This means that almost any function can be mapped to almost any pin.  Most processors can only map a function to one, maybe two different pins.  What that means is that you can do almost anything the processor can on any of the pins.  So you can use the A/D converter, GPIO, PWM, interrupts, etc. on those extra pins that the RFD900 provides.  That would all be stuff that would have to be programmed in the radio firmware.

I asked Seppo about  plans to use these as PWM outputs so that you could replace your regular RC RX with the radio.  It sounded like this was one of the ideas behind it.  But there's all kinds of potential uses and I'm sure they'll implement simple commands for at least controlling GPIO on them. (digital on/off)

As for the encryption, I'd like to see that also.  SiLabs has example code for it ready to go, so it wouldn't be difficult to implement.  With these radios though you have a lot of variables that would prevent interception/takeover.  You have your ID number, start/end frequencies, number of frequency hops, air data rate, etc..  If they aren't all the same then the evil radio will be hopping all over the wrong frequencies at the wrong time and will never lock on.

Potentially someone could rewrite the firmware to scan through all the settings until it locks on.  But at a second or two at a try I think this sort of brute force method would not work.  OTOH if they had a fancy spectrum analyzer with a computer interface and wrote software that would figure out the settings based on the observed radio transmissions it might be possible.  In any case it would take a very dedicated person with good programming skills and a lot of expensive equipment to somehow target you for some reason and be in your flying area at the right time.

It would probably be far easier for them to jam your radio and spoof the GPS signal to sucker your plane down where they want it.

Then again, if you leave all the default settings how they are then anyone with a 3DR or similar radio could lock on and mess with you.

This sort of radio sniffing has been discussed a little before...

http://diydrones.com/forum/topics/sniffing-mavlink-packets

Maybe the odds of getting sniffed/hacked are very remote, but someone somewhere has at least been thinking about it.  I would think you're far more likely to get struck by lightning, but it's at least theoretically possible.  I just figure that if someone actually targeted me like this I'd have bigger problems than them getting my drone, and if they just wanted the drone there's easier/cheaper/quicker ways.

• ▶ Reply

Permalink Reply by Justin on July 21, 2012 at 8:04pm

Ok, so I'm going to order 2 of those modems, 1 have one final question:

The unit uses 2 antennae and uses diversity switching to determine which polarity antenna is the best to transmit from, where do you obtain a reverse polarity antenna as the only differences I've seen have been the connectors?

• ▶ Reply

Permalink Reply by Jake Stew on July 21, 2012 at 10:45pm

I don't have the RFD900, just a pair of regular single antenna 3DR radios.

Usually antenna diversity uses identical antennas though.  You might want to check, but I'm pretty sure you'll be using identical antennas.

• ▶ Reply

Permalink Reply by Seppo on July 22, 2012 at 8:51am

Hi Justin, Jake!,

The RFD900's have 2x RP-SMA antennas on each modem.  You can install any kind of antenna on each of the ports, even leave one port open and use only one antenna if you like.

There are a couple of options for antennas available on the store and a bundle is available which includes 4 antennas (2 x 1/2 wave dipoles, and 2 x 1/4 wave monopoles).

If you would like to use antennas with SMA connectors - there are adapters around which convert RP-SMA, to SMA.  I think the diydrones store has those as well.

If theres anything I can help with, let me know !

Cheers,

Seppo

• ▶ Reply

Permalink Reply by Justin on July 22, 2012 at 3:23pm

Hi Seppo,

I noticed the bundle on your store yesterday so that looks like a good option for me.

Do you know when you'll have encryption available for these radios?

• ▶ Reply

Permalink Reply by Seppo on July 22, 2012 at 7:57pm

There are no plans for encryption on the radio itself at the moment, there are early thoughts to base an updated modem on an Si1034 platform which has double the flash, ram, and encryption built in.

That said, there is no restriction encrypting the data prior to it arriving at the RFD900. We are looking at a few add-on boards to the modems ( one called RCLink ), which will have an STM32 encoding RC PWM channels.  This would have the enough spare processing capacity to encrypt data if required, then send encrypted data to the modems.

• ▶ Reply

Permalink Reply by Justin on July 22, 2012 at 9:24pm

Oh....Jake made it sound like encryption could happen fairly easily/quickly on the current modems.  I'm guessing you have no ETA on these new modems?

Perhaps I'll purchase some Xbee Pro 900's for the interim as I use them on other projects, so once a new radio is out I'll just repurpose the Xbee's and purchase the new radio.

Permalink Reply by Jake Stew on July 22, 2012 at 10:43pm

What I meant was that SiLabs has an encryption library ready to go.  It would not be difficult to implement it if the memory is available.

I think the current firmware uses most of the memory available, so it would require some choices on features to cut out or finding some code optimizations.

It takes 43 bytes of ram and 1056 bytes of code space for the encryption module, and 48/2100 bytes for the decryption module.  You'd have to check with one of the dev team members to see if this amount could be found.

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page