I have just posted the following on Chris Anderson's blog about multiple IR blimp beacons at the DIY drones sight I don't suppose anyone will read it but I think you might be interested...............

I was very interested to read G Bulmer's ideas about ultrasound. If you were to move the ultrasound receiver on the blimp to the top of the envelope pointing upwards not only would you be able to use the perfectly flat & unobstructed ceiling as a height datum but you would also be able to bounce signals off it transmitted from an upward-pointing ultrasound beacon on the floor.
Ultrasound propagates a bit like the light from an electric torch- not very focussed and only kind of directional. But you can time it to a whisker. It moves at a SNAIL'S PACE in electronic terms. Using the ceiling as a kind of ionosphere would I think give a beacon quite a useful range regardless of the direction of the blimp, for optimal results you would want the blimp flying as near to the floor as is consistant with not bumping into the furniture.
Ultrasound propagates best through air at 38 kHz, the same frequency as your IR modulation. You would have to modify your beacon to do a 'ping' to an ultrasound transducer. You probably wouldn't need to use a transistor even. Just an ultrasound transducer; they work as transmitters or receivers.
I'm assuming the beacon runs off a microcontroller so that just means writing a bit of code to do a 38 KHz pulse on an unused output pin. You can check if it's working by connecting another transducer to an oscilloscope. You won't need an amplifier or anything. Just the transducer & two bits of wire.
You would need to get your beacon to do some kind of unique IR pulse which means 'sending ping now'. Then you would need to write a ping-detecting & timing routine for your blimp which would give it an exact (triangular) distance from the ultrasound transmitter on the floor.
If you got that to work you could add a second ultrasound transmitter to another pin of your beacon microcontroller; then it would be necessary to devise another IR pulse which would mean 'sending ping on the other ultrasound transmitter now'.
You now have one IR beacon with two ultrasound beacons attached to it by two lengths of wire. Assuming you place the ultrasound beacons a known distanca apart (say about ten feet) this should give your blimp its exact position anywhere it is withinn range of both beacons.
Ultrasound is very easy to work with. Much easier than you would think.

If you have a sillyscope a good way to start is connect an ultrasound transducer to its input then connect another one to something that generates a 38 KHz signal eg a 555 timer or a signal generator if you have one.
Now just try looking at the received signal bounced off walls, coffee cups etc.

I think you may find you get the same ultrasound signal strength without a transistor. The ultrasound transducers have an extremely high impedance. I forget what it is but it's tens of thousands of kilohms.

because my knowledge of electronics is self-taught it has a lot of holes in it & one particularly gaping one is how to increase the strength of an ultrasound signal. Because I never found out how to do this I have always focussed more on how to detect really weak ultrasound pulses
If you create the pulses using software you can do it in a very controlled way, I found for example that I could usefully detect a reflection from one single oscillation of a transducer.

thereby hangs a very strange tail. But I won't trouble you with it just now.
The simplest way to receve it is to connect an RC diode across the transducer (ie in parallel) also ye olde voltmeter. I suppose if you modulated the transmitted signal at an audio frequency you would be able to use a crystal earpiece instead of the voltmeter. I've only just thought of that. Never tried it.
Yes I used an op amp, as far as I can remember I used the RF diode as the feedback component. Where you would usually put a resistor to set the gain.
In fact I used an LM324N which is four op-amps on one (very cheap) component, I cascaded them together: I put the RF diode on the first stage then just connected the outputs to the imputs without any feedback resistors at all on any of the other op-amps. So that was just two components. I think in fact I put a resistor in parallel with the RF diode.
The limiting factor with it was I found that it tended to feed back & switch its final output permanantly on, an unfortunate side-effect of this was that it would overheat & die on the slightest pretext. As they only cost (then- about 15 years ago) about 10 pence this wasn't a problem.
I used to get round this by only switching the power supply on instantaneously just before the amp was due to detect a ping.
Anyway I daresay you know how amplifiers are supposed to work & will adopt some more conventional approach.
Nowadays ultrasound modules tend to be self-contained with a PWM output. I wonder if this is how the one on the blimpduino works. If so it won't be possible to program it to listen out for externally generated pings.