Estimating data-dependent jitter of a general LTI system from step response

Abstract

We present a method for estimating data dependent jitter (DDJ) introduced by a general LTI system, based on the system's step response. A perturbation technique is used to generalize the analytical expression for DDJ. Different scales of DDJ are defined that characterize the probability distribution of jitter. In particular, we identify a dominant prior bit that signifies the well-known distribution of DDJ, the two impulse functions. We also highlight that system bandwidth is not a complete measure for predicting DDJ. We verify our generalized analytical expression of DDJ experimentally and show that estimation errors are less than 7.5%.