Why was the chipping rate for the GPS L1 C/A signal chosen to be 1.023 MHz (1023 chips per symbol, one symbol per ms)? It seems to me like you'd want to get one chip as close as possible to an integer number of oscillations of the atomic frequency standard. With a cesium frequency standard, this rate is 9192631770 Hz and a chipping rate of 1.023 MHz gives you 8985.95 oscillations/chip. This isn't far from an integer, but they could have done better with several other choices. For example, a 941 kHz rate (941 symbols per 1 ms chip) gives 9769.0029 oscillations/chip. Does anyone know why they picked 1023?