Accurate frequency distribution through packet switched networks can be understood as an extension of the TDM synchronization network based on a few new building blocks like the Synchronous Ethernet Equipment Clock (EEC) and the Packet-based Equipment Clock (PEC). However, for time and phase distribution applications, where most of the interest resides today, this approach does not work.
ITU-T packet switched network timing and the standards.
Time and phase synchronization require new synchronization architectures. The challenges to deliver the required accuracy level are especially important in these new scenarios. There is also a renewed interest in synchronization testing related both with network commissioning tasks and troubleshooting.
This paper reviews existing testing techniques applicable to frequency distribution and introduces the new techniques for phase and time applications. A minimum description of the technologies that enable accurate phase and time distribution, including the Precision Time Protocol (PTP) is also addressed.
In any synchronization test set, there are at least five critical blocks or functions: the local oscillator, the clock reference, the network emulator, the test engine and the post-processing unit. Each block involves its own challenges for reliable and accurate tests:
- Local oscillator: The local oscillator constitutes an internal frequency source required by any synchronization test equipment. The local oscillator is expected to be accurate within certain limits. It could theoretically be used as an autonomous (internal) reference but most of the time the local oscillator is locked to another clock reference. In this case, the local oscillator inherits some of the properties from the reference. A typical situation is to discipline the internal oscillator with a GNSS source. It is then expected that the local oscillator gets the long term frequency / time accuracy of the GNSS source.
- Clock reference: Sometimes, such as in jitter tests, the clock reference can be recovered from the signal under test by some kind of filtering processes, but more commonly the clock reference constitutes an independent input in the test. Two main alternatives are used in practical scenarios: Primary Reference Clocks (PRCs) / Primary Reference Time Clocks (PRTCs) and Global Navigation Satellite System (GNSS) signals.
- Network emulator: The test set has to be connected to a device or network to measure and, to some extent, it has to be compatible with the system where it is connected. Sometimes it is enough to achieve interface compatibility like with TDM or 1PPS testing, but in other situations both interface and protocol compatibility is required. The most typical example of protocol compatibility is PTP testing that requires the test set to be interoperable not only with the physical interface (usually Ethernet and IP) but also with the PTP protocol itself. Specifically, most often, the test set has to implement some of the functionality of a PTP slave. It has to not only decode timing information from remote PTP entities but also generate different kinds of PTP messages such as signaling messages and delay request messages.
- Test engine: The purpose of any test equipment is to measure and generate a result based on this measurement. In a synchronization tester the measurement results consist of a sequence of numbers computed by comparing a relative or absolute time associated to the device under test and the time from the clock reference. For a typical wander measurement the test unit may generate several tens or hundreds of test results per second. The exception to this rule is jitter testing, as a high frequency phase impairment. Measurement bandwidth for jitter is in the range of kHz or beyond and it requires a different approach.
Differences between TE and TIE. The TE is the difference between two times and the TIE is the difference between two time intervals
- Post processing unit. This building block computes synchronization performance metrics from the raw measurement results. Many impairment sources are either random or difficult to predict (variable waiting time in queues, oscillator noise, variations on GNSS coverage, temperature fluctuations). For this reason, the associated performance metrics are statistical in nature. Some common statistics are general purpose metrics like averages or standard deviations while some others have been defined specifically for synchronization applications such as the Allan Deviation (ADEV) or the Time Deviation (TDEV). Randomness of synchronization test results is a challenge in terms of repeatability. For example, estimations of the standard deviation of some kinds of phase noise does not converge to any specific value, even in very long tests; there is no way to measure (or even to define) an standard deviation for such processes. In other cases, the impairment processes involved in synchronization tests have a very low frequency (hours, days, weeks) or they are not periodic at all. Measurements involving non-stationary processes may be very long and even in this case may not be totally repeatable.