24 GHz Amateur Radio Band
for the first server sky experiments
The first server sky experiments may use the 1.2 centimeter ham band, 24.00 to 25.05 GHz, transmitting to a global network of amateur radio operators for the ground stations.
Assume a test array of 50 12x12 centimeter 2 watt 5 gram thinsats (100 watts total power), deployed from a 2U cubesat. Effective transmit power, perhaps 2 watts, effective transmit antenna area, 12 x 600 centimeters, making a steerable asymmetric beam 0.1 by 0.002 radians, EIRP 125 kW. From GEO (40,000 km path length), with a 2 meter tracking receive dish, and 6 db attenuation loss, that would deliver 5 picowatts to the receiver. Assuming kT = 0.05 eV = 8e-21 J at 600 K effective noise temperature, and 5 kT per bit, that is 4e-20 W/bit. That works out to 5e-12/4e-20 or 125 Mbps. Assume a 100 Mbps full speed downlink, 50% duty cycle (nighttime, antenna seeking); plenty of bandwidth to return a lot of scientific information from the test array.
This band is rarely used by radio amateurs - the horizontal path loss is way too high. Using the band for experimental science, and later for amateur radio experiments, might be allowed by FCC/ITU rules.
Aligning the array antenna
Assume we need to "seek" alignment for the array frequently.
We can presume we know exactly where the array is from computed ephemerides and GPS timing; ditto for the ground station. In addition, presume that the ground station can be (over time) precisely calibrated for pointing direction, presumably against the 1.1 to 1.6 GHz satellites. However, the thinsat orientations and phase alignment will need to be learned and calibrated.
- The pattern at earth distance from GEO is 4000 km by 80 km. Before alignment, the test array computes its angular direction from the intended ground station. Then it rasters out a pattern approximately around the ground station with short packets, sending the array's direction and numbers for the array launch direction. The ground station will return the numbers it hears. Over time, we can learn to predict thinsat orientation, then learn to adjust it with the electrochromic thrusters and emitter timing.
The above is all mad handwaving, of course. It is mostly to allow crude estimates for bandwidth and alignment time.