Speaker
Description
The CMS trigger system is a two-tiered filter designed to reduce the information coming from the detector to a level that can be stored for later analysis. The first of these tiers is the hardware-based Level 1 Trigger (L1T), which quickly parses through events for interesting physics objects to reduce the 40 MHz event rate in the detector to a rate that can be transmitted to the next tier, the High Level Trigger (HLT). As part of the CMS Phase 2 upgrade, track information will be available at the L1T and with this comes the opportunity to trigger on new physics objects such as long-lived particles (LLPs). The existence of LLPs is predicted by many beyond Standard Model theories and are characterized by their ability to travel noticeable distances away from the collision point in the lab frame before decay. These decay products therefore originate not from the primary vertex of the proton-proton collision but from a “displaced vertex” of the long-lived particle. The presence of displaced vertices can be used to flag events of interest and further serve to reduce the incoming data rate from the detector. Presented here is a feasibility study in calculating displaced vertex information using the Level 1 track trigger information. Combinations of tracks are used to form vertex candidates and machine learning techniques are used to distinguish between fake and genuine displaced vertices.
