Your insight never ceases to amaze me Maurice. A quick tutorial on orbital dynamics should help explain. Without getting into the math, satellites are launched into planes. Think of wrapping a string around a balloon. If you space each string some specific distance from the other, pretty soon you have several strings around the balloon. Now each separate string is a plane. Several satellites are launched into each plane. The earth rotates under the satellites. The important point is that it takes a lot of energy to move from one plane to another and typically satellites don't carry the fuel that allow them to change their orbital plane. Moving in plane takes much less fuel. Most of the fuel on board LEO satellites is used in changing the orbital altitude.
The reason it does not take much fuel to maintain orbit is because these satellites are 3-axis stabilized. Torques in the orbit slot are normally controlled by gyros or more generally (because of lifetime & reliability) momentum wheels. Reaction wheels respond to disturbances on the satellite. For example, when a pointing error creates a signal which speeds up the wheel. The torque of the wheel corrects the vehicle. Secular disturbances, such as drag, cause the wheel to drift toward saturation. This is when an external torque, usually a thruster is applied. This is typically done automatically on the satellite. As a result, when in stable orbit, you don't use much fuel.
What this means is that you have a satellite failure in plane x and your spare satellite is in plane y, then you have a problem. You will not have enough fuel to fill the hole with your spare because of the plane change. However, if your spare is in the same plane, then you are OK.
A few comments about colliding satellites and orbital debris. If two satellites collide, think of it as an explosion where lots of shrapnel is generated. Now, all your other satellites in that plane have to fly through that debris. The probability that you could wipe out an entire plane of satellites is pretty high. That is the risk you take, if you don't control the satellites. Also, to simply the algorithms (and therefore software), you want handoff to be deterministic. The system needs to know which satellites are available and when. Probably the most important reason that you want to know where the satellite is located is so that you can provide better service. I am sure that there are billing areas and in order to ensure that the customer is billed properly for any particular area, you need to know his location. He either has to calculate it in the handset or the satellite has to figure out where he is.
I think that pricing minutes according to load is a good idea, but if you look at the demographics of earth, you find most people are clustered in areas and not uniformly distributed. This will contribute to the battery loading issue I discussed and there is no easy answer.
I've been a satellite engineer for over 20 years and whenever somebody says "trust me" I begin to worry. I am sure that the engineers at Loral and Motorola are learning a lot, because frankly, no one has ever flown systems with exactly these type characteristics and these type projected loads. I expect both are good and both will learn from their mistakes. From what I know about Iridium, technology is not the problem, marketing is.
There is an old saying, but very true. "You have to know what business you are in." I am not sure that Iridium has figured that out yet. At least, from what I have read, their marketing seems to be disjointed and inconsistent. It is still too early to tell on G*, because they have not gone in to service.
CSM
PS. The amount of time a satellite can stay in orbit without any adjustments is dependent on its altitude and is a non-linear function (close to logrithmic). For example, a satellite at 400 Km at solar max has a lifetime of a couple of months, at 600 Km about 2 years, at 800Km about 60 years and 1000 Km over 600 years. |
