Jitter extraction in a noisy signal by fast Fourier transform and time lag correlation

Abstract

Jitter in an electronic signal is any deviation in, or displacement of, the signal in time. This paper investigates on decomposition of two types of jitter, namely, periodic and random jitter in noisy signals. Generally, an oscilloscope generates an eye diagram by overlaying sweeps of different segments of a long data stream driven by the reference clock signal. We use the fast Fourier transform with time lag correlation of the signal since we do not have a clock reference signal and apply this technique to simulated noisy signals. We separately injected a random jitter (of known amount), periodic jitter (with known frequency and amount), and both together to various modulation frequencies of sinusoidal signals. The approach is validated by several experiments with numerous values in jitter parameters. When we separately inject random jitter (5 ps) and periodic jitter (5 ps at 4.37 MHz) to the signal, we obtained the results (4.52±0.25 ps) and (4.93±0.04 ps at 4.40±0.04 MHz), respectively.