The following is a reason why GPS satellites cannot test the one-way speed of light, as they cannot know their clocks are incorrectly synchronized, and that their absolute positions are shifted.
Consider this (ignore time dilation effects to simplify things):
Two satellites A and B are placed north and south either side of the Earth (point E). Let the satellites be 300,000 km from E and the speed of light be c = 300,000,000 m/s (to simplify results the satellites are over ten times the typical GPS distance). Let the speed of the ether flow relative to the Earth be V = 200 km/s in a direction south to north.
According to SR light takes one second to travel from the Earth to either satellite and it takes two seconds to travel between satellites. To synchronize the clocks, a pulse of light is sent to each satellite at Earth time t=0, and satellite clocks register t=1 when they receive the light.
But with the ether flow light takes longer to reach B than it does to reach A, and consequently clock A gets set ahead of Earth time by 0.000666223 seconds and clock B get set behind Earth time by 0.000667111 seconds. These offsets compensate for lights motion through the ether. Both satellites now agree that light takes the same time to travel between them (2.000000889 seconds) and leads to the conclusion that the speed of light is constant both ways. But we know that light travels at speeds c-V and c+V between the satellites. Similarly the time for light to travel between the Earth and the satellites appears equal, but this is not the case as the clocks are not synchronized according to Einstein's method, and the speed of light is not constant.
At these distances the satellites positions disagree with SR by (0.000000889 * c = 266.7) m. But typical GPS satellite orbits would be over ten times less than this, resulting in a difference of about 20 m in the direction perpendicular to the ecliptic plane. GPS satellites are not checked optically to this accuracy and so their true orbital positions are not known.
It can be concluded that the GPS signals that travel between the Earth and satellites cannot be used to test the speed of light in one direction and so cannot be used as an argument to dismiss the proposed one-way experiment. Also from this example it is clear that if we do assume the speed of light to be constant, then the Earth appears to become a special frame, i.e. (ECI). So GPS may indeed be an accurate system, but clearly it is not fool proof.
wisp
- particles of nothingness
