• efficiency
• simulation
• 2015
• rate
• trigger
• level-0
• kaon
• NA62

This master thesis work focuses on a simulation of the lowest level trigger system (L0) of the NA62 experiment at CERN.

The NA62 experiment is designed to measure the K^+->πνν_bar branching ratio with signal to background ratio of about 10:1. This decay mode is particularly interesting for its sensitivity to new physics signals, and for its potential to increase the accuracy on the knowledge of CKM matrix elements values.
The high rate in the experiment doesn't allow to store all the data, thus a fast and efficient trigger system is required.

The simulation aims to give an offline replication of the real hardware trigger system, in order to have a tool for trigger checking, but also for its optimization.
The real trigger system is based on informations (called primitives) on detected signals coming from different detectors. These informations depend on the detector, thus primitive generation is different for each involved detector. Primitives from different detectors are sent to a central processor which elaborates the trigger decision by comparing them with some predefined sets (called masks): in case of a positive match, a signal is transmitted to all detectors and data are readout. This is only the lowest hardware trigger level, after which the higher trigger levels (L1/L2) implemented in software take the definitive decision on data storage.

The NA62 experiment presents an innovative feature in trigger system structure: the same data used by the hardware trigger system to generate primitives and retrieve the trigger decision, are completely stored on disk. This means that the offline simulation and the hardware trigger system act both on the same informations and potentially a perfect offline reconstruction of the hardware system can be performed. A trigger simulation allows, in this case, a deep investigation on the real system performance.

The strength of this offline trigger simulation is the ability on acting both on acquired data and MonteCarlo samples.
Acting on acquired data, it allows a comparison between the real and the expected performance and so it can be used for debugging, trigger optimization and inefficiency estimations.
On the other hand, running the simulation on MonteCarlo samples is an useful tool for trigger conditions optimization, signal efficiency and background rejection estimation and trigger rate studies.

The first part of this project consists in an approach to the general understanding of the NA62 experiment, studying its tasks and structure. In this part I focused on a study of trigger conditions: an extensive set of primitive conditions for each detector included in the L0 trigger system is formulated, and a set of trigger masks for selection of new physics signals is proposed. Rate estimates for these trigger masks were estimated from existing simulations and rough evaluations.

The core part of the thesis consists in implementation of the simulation and its comparison with the real trigger system during 2015 run.
This part includes also a monitoring and checking of the performance of the real trigger system by means of an the analysis of data collected during this run which highlighted some problems in the system, some of which are still under discussion. The comparison of the simulated trigger with the real one led to the evaluations of simulation limits and real trigger system inefficiencies.

In the last part of the work the simulation is used for trigger conditions optimization for K^+->πνν_bar signal detection. Trigger rate and signal efficiencies are estimated comparing the performances of a trigger mask based on arrangement during 2015 data with one based on ideal trigger conditions which will be implemented in the next future at NA62 experiment.