I will discuss methods of comparing and synchronizing clocks and the procedure of characterizing their performance in terms of the two-sample Allan variance. I will describe methods that are used when the device under test and the reference device are in the same facility and when they are at different locations linked by a communications channel. The reference device can be a national time scale, and I will briefly describe how national time scales are defined. In each case I will use the Allan variance as a tool to understand the characteristics and limitations of the various methods. I will also introduce the concepts of traceability and of a cost-benefit analysis into the synchronization procedure.
Handbook of Measurement in Science and Engineering