A Correlation Based Pulse Detection Technique for Liquid Scintillation Detectors

A new correlation-based pulse-detection technique is presented that significantly improves the probability of detection for low signal-to-noise ratio (SNR) neutron and gamma-ray pulses generated by organic liquid scintillation detectors. Current methods utilize amplitude thresholding to detect the arrival of a pulse as well as to start acquisition of data; however, this approach faces two major limitations. First, to capture low-energy pulses the threshold amplitude should be set close to the amplitude of the noise from the electronics, and false detection is more likely to occur. Second, a large amount of measured data needs to be stored on a computer for post-processing to determine pulse characteristics²hence, the number of events that can be recorded is also limited by the hardware. In this work, we propose performing a normalized cross correlation of incoming scintillator data to a predefined pulse template, and using the correlation result to trigger the detection of a pulse. The predefined pulse template is an average of high energy pulses measured from a particular liquid scintillation detector which then can be used as a template pulse to detect low energy pulses. This technique is more robust to noise, allowing detection of lower SNR pulses. Furthermore, the cross correlation can be computed on a field programmable gate array (FPGA) using correlator hardware accelerators, so that pulses may be detected in real-time as opposed to after post-processing of acquired data. This hardware is similar to what is used for code division multiple access (CDMA) wireless communication. The correlator implemented on an FPGA also allows recognition of the pulse peaks and time-of- arrival in real-time. Therefore, we are only required to store the peak amplitude and time-of- arrival for a pulse as opposed to complete pulse waveforms²resulting in substantial reduction in required storage space. This technique is well suited for nonproliferation and homeland security applications where it is necessary to detect a large number of low SNR pulses.