Time jitter and phase noise — Now and in the future?

Phase Noise is important for the oscillators in all RF communication systems and total time jitter and stability are important for clocks and clocked digital systems. Time jitter and phase noise are intimately interrelated. The spectra are identical if there are no AM components. It is fundamentally easier to obtain total jitter from the phase noise spectrum rather than the reverse. Asymmetric spectra imply both AM and PM components present simultaneously. The time jitter can then be different on the up and down transitions of the signal waveform. Stability and `wander' measurements take a long time. Suitable algorithms operating on the measurement records can minimize this time. Information theoretic methods promise further improvements in the future. Divider noise can always be removed by `re-clocking' of the divider output. In the future this kind of noise should no longer be an issue. Leeson's phase noise oscillator model can easily be extended to include, Q-multipliers, delay line multimode oscillators, and injection locked oscillators. Phase jumps in the clocks of an ensemble can be removed by three-cornered hat measurements. The physics of phase jumps remains an unanswered question for the future. Software radio techniques (SDR) are offering significant improvements in spectrum and jitter analysis. In the future phase noise could well be seen to apply in several areas of physics such as quanta and sub-atomic particles and gravity waves. Phase Noise will remain a major determinant of electronic system performance for the foreseeable future.