FPGA Implementation of an Artificial Neural Network for Subatomic Physics Experiment Particles Recognition

Abstract

CMOS Pixel Sensors have been used in subatomic physics experiments for charged particles detection. In the International Linear Collider (ILC) vertex detector, the occupancy will be mainly driven by impacts coming from the beam background. This will have a huge impact to the data flow of the system. We propose a design of CMOS pixel sensors with on-chip Artificial Neural Network (ANN) to tag and remove these hits. It is based on different features of hits clusters. In this paper, we will describe the structure of an ANN implemented in an FPGA device. We will show and analyze the distribution of incident angles reconstructed by the ANN.