Combinatorial Kalman filter and high level trigger reconstruction for the Belle II experiment

One of the most important steps in both the online and offline event reconstruction is the search for the charged decay products using the tracking detectors. Tracking is the only reconstruction part able to extract the parameters of all charged final state particles with a high precision. The increased instantaneous luminosity comes with a drastically enlarged beam-induced background, especially in the innermost silicon-based tracking detectors, which imposes major challenges to the tracking. To address these issues, a combinatorial Kalman filter (CKF) was introduced into the Belle II software through the work of this thesis. It enables a wide range of applications during track finding and is used for attachment of hits and tracks in the pixel detector (PXD) and the strip-vertex detector (SVD). The obtained improvements to the vertex resolution are crucial in allowing a larger range of physics channels to be studied and a higher precision in all measurements. The Belle II detector is introduced in Chapter 2 with a focus on the tracking subdetectors and the trigger setup required for this thesis. The foundations behind the development of tracking algorithms and the concept applied at Belle II are described in Chapter 3. In Chapter 4, the detailed processing time measurements for the HLT reconstruction as well as the developed trigger setup are discussed together with the new implementation for the multiprocessing framework. The first-developed precise event time determination required for a successful data acquisition is described in Chapter 5 together with its performance. The combinatorial Kalman filter developed during this work is presented in Chapter 6. Its large beneficial impact on the resolution of the reconstruction and other important figures of merit as well as its performance on the first recorded data is outlined in the chapter. The thesis ends with conclusions on the performed work in Chapter 7.