Explanation of satellite links
The uplink antenna diameter of the cell phone is imaginary. The number (diameter) is artificial and you need to simply adjust this number to get whatever transmit gain you think is realistic. I have estimated 0 dBi which means omni-directional. This is not unreasonable in that the satellite may not be exactly overhead and the cell phone is unlikely to have a sharp null upwards.
Choose a suitable mobile uplink frequency, say 890 - 915 MHz or 1805 - 1880 MHz (or both, if you fancy a multi band system)
I have guessed the cell phone power output to be 200 mW and carrier bandwidth 170000 Hz. This can easily be adjusted up and down for your refinement. 200 mW may be representative of the conversation part of a call but if you want to receive just the initial acquisitions and ringing then the phone is likely to output 2 watts at 890-915MHz or 1 watt at 1805-1880MHz.
Update Aug 2016: I wrote this page in 2004 and amending it in 2016, and 2019. I have seen publicity about the MUOS system of 4 geostationary satellites providing a cell phone service for military users. The operational satellites are at 177W MUOS 1, 100W MUOS 2, 15.5W MUOS 3 and 75E MUOS 4. MUOS 5 has recently launched and is intended for longitude 72° E. See: https://en.wikipedia.org/wiki/Mobile_User_Objective_System MUOS is stated to work in range 300 MHz to 3 GHz and the picture of the feed system suggests a feed array of 15 tapered broad-band helical UHF antennas.
For the spacecraft uplink antenna I have guessed at 30m diameter and 15 individual spot beams. Think in terms of a deployable mesh dish like INMARSAT 4. You need to run the uplink G/T calculation for this separately: e.g. 30m diameter at 1.85 GHz, 65% efficiency and 300K noise temp gives a G/T of about +28.5 dBK The noise temp is high because it looks down at the warm earth, although it will be cooler if you point the satellite uplink antenna at the sea. There could well be so many ground terminals operating that the uplink is full of noise from all the terminals combined. In which case you need to work out hoe you will detect the wanted uplink site, for example by exact frequency, spread spectrum code match etc..
For a more sensitive system with global coverage consider more ans smaller spot beams. e.g a multi spot beam design and the diameter of each beam or cell. The earth is 17 deg across so you could have 55 beams each approx 2.1 deg dia, each with a gain of 33 dBi (5m dia at 1.8GHz) G/T= 33 - 10log(300) = 8 dB/K
On the downlink to to the teleport hub site I have assumed some suitably unusual frequency (34 GHz) for fun. Change it to anything you want.
For the satellite downlink eirp you need to adjust the satellite eirp per 200000 Hz bandwidth till you get an acceptable downlink C/N. In the default example below a downlink satellite eirp of +16 dBW per 200000Hz is just about enough. If the downlink satellite antenna is 43 dBi (e.g. 0.5m dia at 34 GHz) then you need a spacecraft tx power of -27 dBW per 200000 Hz which should not be too difficult to achieve.
Typically you would want the satellite to receive the full bandwidth 890-915MHz (25 MHz) so the satellite output amplifier needs to be able to support 124 FDMA channels, each of 200kHz. For a factor of 124 add 21 dB to the wanted single carrier power. The result is the composite multi-carrier power. Add a further 3 dB for linearity. Similarly for the wider 75 MHz bandwidth of 1805-1880MHz.
Amend all the white boxes freely as you wish and then click any calculate button to obtain results in the green boxes.
Any problems or comments please e-mail email@example.com