A good first read is An introduction to RF telemetry systems, by Gale Allen (pdf link, 11 pages).
To summarize, with the same type of antenna, the higher the frequency the lower the range. We should prefer a 500MHz band to a 2.4GHz for example. However directive antennas with higher gain are more practical in higher frequencies because the wavelength is shorter and antennas are sized to the wavelength. They may also be cheaper. With a high gain antenna, a higher frequency link can reach the same range as a lower frequency link with a unity gain antenna.
There are some license-free radio frequency bands available, the ISM bands (Industrial Scientific and Medical) :
- In the EU: 433MHz – up to 10mW power, 868MHz: multiple channels with power output up to 500mW. 2.4GHz with outputs up to 10mW.
- In the US: 433MHz up to 1mW output, 315MHz up to 10mW output, 915MHz up to 500mW (with restrictions on protocol – spread spectrum).
The ~900MHz band seems more promising since it allows for relatively high power without requiring a license, and the bandwidth will be more than enough if no video relay is considered on the telemetry link. Here is an example of RF module of 500mW on 869MHz ($100), here a telemetry module of 1W on 902-928 MHz ($90).
Weather balloons are launched very often (more than 850 twice a day around the world) and the most used product is the Radiosonde Vaisala RS92 and variants. Their RF output is 200mW for the 1680MHz version, and at least 40mW for the 403MHz version. Amateur radio operators have reported catching signals from those radiosondes several hundreds of kilometres away, so we definitely don't need more than the allowed 500mW ISM RF power. That may however require a high quality reception station with high gain antennas and low-noise amplifiers.
A RC kit is announced for 172km range with a 500mW module on 869MHz band. Antennas used for this range are a omnidirectional antenna (+5dBi) for the reception and a patch antenna (+9dBi) for the transmission.
Amateur radio satellites can be easily received from the ground, although their transmit power can be quite low. They use 145 MHz and 435 MHz bands in various uplink/downlink configurations. For example, the Saudi-OSCAR 50 satellite uses a 250 mW UHF transmitter with a 1/4 wave antenna on the 435MHz band, and it can be received, with quite some noise, with a low cost radio and a 2.15dBi gain 1/2 wave antenna.
Amateur satellites have to declare their orbit and frequencies to the International Telecommunication Union (ITU). This can be done for free now.
List of emission modules available on the 869MHz ISM band, 500mW power
- TIMWO HP868, also known as ARF7581AA, 1-page PDF link
- ARF35 (IP65 package), price 650 EUR.
- ARF27, price 67 EUR TX, 30 EUR RX.
- ARF 29, price 115 EUR.
- Yishi YS-C30L
- Mellow ml808
- HOPE RF RFM12BP SPI-controlled 500mW module
- XBee-PRO 868HP, announced for 80km with RF line of sight, but limited to 315mW
- Radiometrix TX3H transmitter, requires coding circuitry, 450mW
Reception equipment for the 869MHz band
Three kinds of choices are offered to us for reception:
- the reception module matching the emission module, some of them are indeed developed and sold together; the advantage is that we know what is the sensitivity of the receptor and we know that it will operate without issue on the same band,
- an amateur radio equipment,
- a software defined radio equipment (SDR), like the populars FunCube Dongle Pro+, bladeRF and the hackRF. SDR allows a large range of frequencies to be received and kind of encoding to be decoded. All the work and control is done by a computer, contrary to amateur radio equipment that does it in hardware. SDR interfaces are generally USB dongles on which an antenna is plugged.
In any case, a high gain directive antenna operating in the 869MHz band will be required to pickup the signal that far away, or even send some data upstream. It will need to be directed towards, which can prove difficult when objects are behind clouds or in a not well known orbit.
We may need a low noise amplifier too, depending on the chosen reception equipment.
A general principle about antennas is that the narrower the beam, the higher the gain. Indeed, it would take 100 times more power to cover the whole sphere around an ideally isotropic antenna than with an antenna covering a 1/100th of this sphere, with the same perceived power at the same distance. Wikipedia's article on High-gain antennas is quite instructive. Here is a quote: high-gain antennas must be physically large, since according to the diffraction limit, the narrower the beam desired, the larger the antenna must be (measured in wavelengths).
This wikibooks article has a large list of antenna types and their main characteristics.
This link is a good introduction too, with most properties of antennas explained.
Flying object tracking
It may not be easy to track a flying object with a directional antenna, even inside the atmosphere. If it passes behind clouds for example, you lose the ability to track it visually and it may be complicated to find it again later. In the case of a high altitude balloon with clear sky, that can be done easily if winds don't push it hundreds of miles away. Otherwise, it may get behind mountains and the line of sight can be lost if it's not high enough in altitude.
There are two main solutions to this problem. The first is to let the aircraft provide its position through the telemetry link, which is then used to refine the pointing of the tracking antenna. The issue with this solution is that bad weather may make the radio link or GPS lock unstable, and still result in failure of the tracking. It also requires the aircraft to know its position quite well, but IMU coupled with GPS should be reliable enough. For rockets however, that may be more complicated to have an accurate location information with amateur sensors.
The second solution is to have several ground stations to triangulate the position and speed of the emitter, using received signal strength indication (RSSI) and Doppler shift. This technique is often used for tracking indoors. Using the RSSI as control loop input with only one station can be done, but bad weather affects it too, and since there are 4 possible actions (2 on each axis), it would be a guess-and-try type of tracking, with a number of missed information.
If the tracking is lost at some point, a wider beam antenna may be used, like a patch antenna (really lost in that case) to try to get a position information. Since these antennas have a lower gain than highly directional antennas, they may not be able to catch the data correctly, but they can still provide a cone of plausible localization.
An introduction to RF telemetry systems, by Gale Allen (pdf link, 11 pages).
A more complete reading is the Telemetry Systems Radio Frequency Handbook, US military document, 2008 (pdf link, 133 pages).