Real-Time Particle Identification in Liquid Xenon

We present a field-programmable gate array (FPGA)-based technique for the on-line identification of highly ionizing particles in a liquid xenon (LXe) detector. The method was developed and successfully exploited to select \alpha particles emitted by 241 Am sources submerged in LXe in an overwhelming, mostly beam-related, \gamma -ray background. After revising the main features of xenon and other liquid noble gases as ultra-violet (UV) scintillating media, we describe the algorithm idea and its firmware implementation. We then present the results in terms of efficiency and background suppression for the real-time \alpha -particle tagging and the limits of the MEG trigger configuration. Finally, we show that in MEG II we are going to overcome the main issues and further improve the performances.