Accurate Time and Frequency Transfer During Common-View of a GPS Satellite

Summary First, Clock A and a GPS receiver used are to Even though the GPS is primarily a navigation deduce from a GPS sate1 1 ite s ephemeris, from system, if two clocks at known coordinates A and B clock.As location, and from received GPS time are in common-view of a single GPS satellite, decoded from the same satellite, the time differreceivers at these two clock sites may coincidentally receive transmitted GPS clock times. By ence (Clock A- GPS time). This method is the subtracting the received times of arrivals as measured by clocks A and B at the two sites while simplest and least accurate (estimated to be compensating for the propagation delays, one has better than about 100 ns with respect to GPS an accurate measure of the time difference between time),2 but has global coverage, is in the receiveclock A and clock B. When all of the error contributions are only mode, requires no other data, yields receiver assessed, it appears that 1 ns time stability and 10 ns of time accuracy should be achievable in prices that could be competitive on a mass producmeasuring remote clocks--at distances of the order tion basis, and could service an unlimited audiof a fe; thousand km. The primary error sources are as follows: uncertainties in the satellite ence. Also, GPS time wil be referred to UTC(USN0) ephemeris, differential ionospheric delays, uncer- and wil be known with respect to UTC(BIH), tainties in tropospheric delay estimation, and uncertainties in receiver delays. UTC(NBS), and other major timing centers. We have chosen this common-view approach Second, Clock A and Clock B at different because it provides an opportunity for a high accuracy (10 ns) relatively low cost receiver due to the common-mode error cancellation achievable. locations anywhere on earth can be compared by making successive observations of the same GPS