Geiger-mode avalanche photodiode ladar receiver performance characteristics and detection statistics

Abstract

The performance of single and multielement Geiger-mode avalanche photodiode (GM-APD) devices are investigated as a function of the detector's reset or dead time. The theoretical results, developed herein, capture the effects of both quantum fluctuations and speckle noise and are shown to agree with Monte Carlo simulation measurements. First, a theory for the mean response or count rate to an arbitrary input flux is developed. The probability that the GM-APD is armed is shown to be the ratio of this mean response to the input flux. This arm probability, P(A), is then utilized to derive the signal photon detection efficiency (SPDE), which is the fraction of signal photons that are detected. The SPDE is a function of the input flux, the arm probability, and the dead time. When the dead time is zero, GM-APDs behave linearly, P(A) is unity, and the SPDE theory is simplified to the detector's effective quantum efficiency. When the dead time is long compared to the acquisition gate time, the theory converges to previously published "infinite" dead-time theories. The SPDE theory is then applied to develop other key ladar performance metrics, e.g., signal-to-noise ratio and detection statistics. The GM-APD detection statistics are shown to converge to that of a linear photon counting device when the combined signal and noise flux is much less than the reset rate. For higher flux levels, the SPDE degrades, due to a decreased arm probability, and the detection probability degrades relative to that of a linear device.

Universal Performance Prediction for Gated and Free-Running Geiger-Mode Avalanche Photodiodes

“…To verify the accuracy of the UDR method, results predicted by the UDR method for gated GM-APDs were compared with results predicted by the iteration discrete-time (ID) method [17] and the look-back-upon (LTR) method [15]. Results predicted by the UDR method for free-running GM-APDs were compared with results predicted by the UDR method for gated GM-APDs by using a long gate time to simulate a free-running situation.…”

“…Define r i (t) as the incoming distribution of the count rate of primary electrons (PEs) at the detector's photo conversion front end where PEs come from light absorbed by imperfect quantum efficiency and the dark noise inside the GM-APD. Then, the mean number of PEs at the n th bin, K n , is given by [12]:…”

“…For gated GM-APDs, the initial value of P AN (n), 0 ≤ n ≤ N , is zero and the initial value of p 1 (n) is still given by (12). For free-running GM-APDs with a constant rate of incoming events before the estimation time, p 1 considering afterpulsing can be calculated by a quadratic equation deduced as follows:…”

“…However, they cannot model the timecorrelation of the dead time. To model the time-correlation of the dead time, several models [9], [10], [11], [12], [13], [14], [15] have been proposed for gated GM-APDs without the afterpulsing effect. The gate-limit assumption is the basic assumption for these models, where the need to consider dead time only exists within the gate.…”

Photon counting ladar work at FOI, Sweden

“…Linear-mode photon counting detectors, however, offer advantages over GM detection [5,6] for single event properties like dynamic range and pulse to pulse range resolution. Due to the limited reset time for GM APD's (0.1-1 µsec) the detector is unable to capture both the obscurant and the target for a range interval between 15 to 150 meters.…”