Systematic afterpulsing-estimation algorithms for gated avalanche photodiodes

Abstract

We present a method designed to efficiently extract optical signals from InGaAs avalanche photodiodes (APDs) operated in gated mode. In particular, our method permits an estimation of the fraction of counts that actually results from the signal being measured, as opposed to being produced by noise mechanisms, specifically by afterpulsing. Our method in principle allows the use of InGaAs APDs at high detection efficiencies, with the full operation bandwidth, either with or without resorting to the application of a dead-time. As we show below, our method can be used in configurations where afterpulsing exceeds the genuine signal by orders of magnitude, even near saturation. The algorithms that we have developed are suitable to be used either in real-time processing of raw detection probabilities or in post-processing applications, after a calibration step has been performed. The algorithms that we propose here can complement technologies designed for the reduction of afterpulsing.

Robust diagnostics of dark counts for quantum networks

“…Light sources, as well as dark noise, corresponding to Poisson processes, correspond to the probability density for the time intervals between samples, which is approximately described by a decaying exponent [8,9]:…”

Robust diagnostics of dark counts for quantum networks

“…Light sources, as well as dark noise, corresponding to Poisson processes, correspond to the probability density for the time intervals between samples, which is approximately described by a decaying exponent [12,13]:…”

Sequences of the ranged amplitudes as a universal method for fast noninvasive characterization of SPAD dark counts

“…Electronic mail: s.a.moiseev@kazanqc.org. method 16 , temporal distribution (background decay) 17,18 and many other techniques for various experimental installations [19][20][21][22] (see also review 1 ). The time interval analysis is widely used.…”

“…where N is the total number of points in the sample (in our case N = 10 3 ). By taking into account that many light sources as well as the dark noise counts corresponding to the Poisson processes 28 , the probability density for time intervals being between the counts will be approximately described by the damped exponent (see also theoretical review 21,29 ):…”