video/data link for UAV

[jmdejoanelli]

Hey guys,

I am building a long range data link, and I want to know if using two separate links of two different but near frequencies will interfere with each other.
One would be an Applied Wireless 3Mbps 922MHz FSK data single direction transmitter and the other would be a Digi International 115Kbps XTEND 900MHz FSK transceiver (with frequency hopping).

If this setup will not work, can you suggest a single RF transceiver module(or solution) with the data rate of the Applied Wireless module and the range of the XTEND module.

This is my thinking:
One link should be longer range, be more robust, but data rate is not that important (115Kbps would be plenty).
The second one should be high speed around 2-3Mbps, suitable for video, but range is not as (range would be a plus).
The video link doesn't need to be robust like the low speed link since it only streams video and the low speed link will still maintain a connection since it is the primary control link (for a surveying UAV platform).

I need at least 20Km range, which is my big issue. I'm happy to use directional antennas on the ground, but everything on the UAV platform must be small and lightweight.

This is for a university research project I'm working on.

Any help would be appreciated.

Cheers,
Josef

[sonicj]

sorry to to be the bearer of bad news, but 3mbps from a aerial platform at 20km with consumer equipment ain't gonna happen.     fortunately, i doubt you will really need that kind of bandwidth, but then again, im not sure exactly what you're trying to do....

anywho, first off, what country are you located? do you or your school hold any licenses to transmit at higher powers? what size ship will this equipment be riding on?

all long range setups im aware of use different bands for uplink & downlink.  the lower frequency stuff is better suited for long range due to propagation. personally, i would do something like 35Mhz for control and 900Mhz for video/telemetry. the distances you are shooting for will require power levels far exceeding unlicensed power limits. you will need a amateur license or academic license (if available) to legally transmit at these power levels.

you may want to have a look at some of the long range equipment by Thomas Scherrer TSLRS gear link
i believe he is also heavily involved/partners with Danish Aviation Systems
-sj

[jmdejoanelli]

Awe... That is bad news.

I wanted to transmit 480p(at most) video from the UAV on a one way link. Then have a 2 way, low data rate control and telemetry link.
What kind of data rate would I be looking at do you think for this resolution (at say 20 or so fps)?

I'm in Australia, hence the love of the 900MHz band, and I think 1 Watt is the limit for me because I don't have any radio licenses. Also, anything over 1 Watt is likely to be too heavy at the aircraft end anyway.

I wonder if I should maybe modify my specs a little. Could I transmit video at 1 or 2 frames per second plus control and telemetry over one 115kbps link do you think? The XTEND modules I've been looking at do 115kbps at 24km or 64km with high gain antennas (manufacturer quoted).

Also, it may be relevent to note that when I say "control data" I mean way point data which would be sent to the craft every so often, since the craft has a (albeit limited) autopilot system.

The craft is an RC Airplane.

Thanks for your reply, I really hope I can get this figured out, because I've been stressing over it.

Cheers,
Josef

[jmdejoanelli]

By the way,

Assume I have a tracking antenna system that can aim a 15-22dBi (depending of freq.) directional antenna to within 1 degree of the aircraft.

[Psi]

Even at 480p with compression you need around 1mbit. Maybe 500kbit if you don't mind a low quality image or 15fps.

There is gear available to stream 720P from an RC plane. And 720P requires around 3mbit.
But i very much doubt it's able to do 20km.

To do 20km @ 720p you'd need to get a HAM license to transmit at the required power legally

[jmdejoanelli]

If I need around 1Mbps for 480p is that for 30fps?

If that's the case, the data rate required for 480p at 2fps would be ~1000/30 * 2 = ~66.7Kbps I figure.

I'm starting to think now that 1 or 2fps will probably be plenty fast enough for my application.

[jmdejoanelli]

Also, now that I think about it (this is all very much thought in process), I may not need 20Km range at all, infact, I don't think my UAV platform has the fuel, batteries to go that far...

Maybe around 3-6Km is a more reasonable request.

lol my original specs are dwindling now aren't they...

[sonicj]

im not familiar with digital or progressive scan ATV so i can't be of much help calculating your transmitter or bandwidth requirements... 

are you using the downlinked video as your surveying imagery?  or just as visual confirmation of waypoints, etc?  if you are planning on using the video as your "collected product", why not send up a high res point & shoot with intervalometer? or if it needs to be video, a go-pro is capable of shooting decent HD video while simultaneously outputting analog sd video. you could feed the analog video signal to the transmitter for real-time monitoring and collect your HD roll at the end of the flight. telemetry can go on the audio channel or overlay it on top of the video before the transmitter.

fuel is definitely a concern. weight is the other biggie. 3km sounds like a very reasonable goal...
-sj

[RCMR]

If you only want 3-6Kms then just use one of the "storebought" FPV video solutions.

I'm about to document my FPV backpack project so that anyone can build one.

With 200mW on 5.8GHz and two circularly-polarized cloverleaf antennas you can get nearly 2Kms range.  Use a patch or helical antenna on the receiver and 6Kms is quite an easy target.

Don't be tempted to use 900MHz for the video link -- the commonly available 900MHz video transmitters are very "dirty" and spray noise all over the bands used by GPS so will throw your onboard guidance out the window.

For the data-link, a couple of XBee modules (on 2.4GHz or 900MHz) will get you a couple of Kms with stock (1/4-wave) antennas.  Add some gain to the groundstation and you can easily quadruple that.

[sonicj]

the 900Mhz stuff can be nasty... a notch filter at the GPS frequency, low swr antenna, careful component placement & attention to filtering and shielding can make it play nice with GPS & other sensitive components.  i haven't used one with GPS personally, but i did have to jump through some hoops to get it working right on my plane.
-sj

[jmdejoanelli]

I like the Go Pro idea. HD video recording with SD monitoring.

If I was to use one of those 1W Lawmate, 2.4 GHz video transmitters with a 18-22dBi dish antenna on the ground. Would that give me at least 6km of range?

Also, if I used a pair of 900 MHz XBee Pro modules, with high gain antennas, the range is quoted at around 10km. Are these 900 Mhz modules going to cause issues with the GPS? I'm just confused if you guys are telling me that *analogue* 900 MHz is nasty for GPS or if *digital* 900 MHz is nasty for GPS.

Thanks again for all your input and advice.

Thanks,
Josef

[sonicj]

this paper sums it up nicely. not sure how the xbee performs.

i would do diversity receivers at your ground station. the receivers and antennas are cheap. the diversity switchers are expensive last time i checked. i wrote some arduino code that compares rssi from the receivers and selects the better of the two via a analog switch. gotta hack open the receivers to get to the rssi pins, but its easy.

again, 2.4GHz  doesn't have the propagation that 900MHz offers. plus, lots of stuff going on at 2.4GHz as well. so... range will depend heavily on your environment.  lots of trees and houses with microwaves & routers is bad news for 2.4GHz. if you are out in the middle of a farm and its line of sight the whole flight then 2.4 might be feasible.

do you have a airframe picked out? or are you gonna scratch build?
-sj

[jmdejoanelli]

Another slightly unrelated question:

How bad is it to have a small DC motor near an antenna? In particular, a 900 MHz Yagi and a small 12V DC motor for controlling the direction of the antenna.

Can't do much about changing the placement of the motor, so what kind of measures could I take to reduce noise? (If necessary)

Thanks,
Josef

[sub]

Are you sure about 900MHz in Australia?  I thought in UHF we had only 433Mhz and 2.4GHz, with 900MHz being used for mobile phones.

Edit: Apparently I was misinformed, there is apparently a sliver of ISM there; disregard my last.

[RCMR]

Another slightly unrelated question:

How bad is it to have a small DC motor near an antenna? In particular, a 900 MHz Yagi and a small 12V DC motor for controlling the direction of the antenna.

I assume you'll be using more than just a "small 12V DC motor" to control the direction of the antenna.

For proper tracking purposes you'll need a servo system that uses either steppers or some kind of servo with a feedback loop to allow for precise positioning.

None of these should affect the operation of a Yagi antenna but for long-range video you'd be better advised to go with a circularly polarized setup using a helical antenna on the base station.  This will significantly reduce the losses due to cross-polarization when the craft is banked or turning.

[jmdejoanelli]

I assume you'll be using more than just a "small 12V DC motor" to control the direction of the antenna.

You would be correct.

I am using 1:100 gear motors with 6400 (gear reduction) count encoders, and two PID controllers I've implemented in a ATmega328.
The motors have a rated torque of around 12kg/cm so they should be strong enough.

Cheers,
Josef

[nimisha]

Hi,

 I am doing a uav project. I would like to have a pair of data /video transceivers. my requirements are 200 kbps or more data rate , 2k distance, < 100grams weight , 4 to 6 GHZ freuency range. Can you suggest one for me ?

Thanks,
Nimisha.

[helius]

Thread necromancy detected 

But one thing that needs pointed out about the older posts, is that you cannot scale video bandwidth (of mpeg2 or mpeg4) down to 2fps in a linear way. The bandwidth at 30 or 25 fps assumes great similarity between successive frames so that interframe compression can be used. At 2 fps there is probably no similarity, and only intraframe compression may be possible.