The Belle II experiment, the upgrade of the Belle experiment, at KEK (High Energy Accelerator Research Organization) in Tsukuba, Japan, will be built to answer fundamental questions that are not covered by the Standard Model of particle physics. For this reason, decays should be observed with high precision.To be able to measure all decay products with a very accurate vertex resolution, it was decided to add a Pixel Detector (PXD) with an inner radius of only 14 mm in short distance around the beam (outer radius 12.5 mm). This increases the vertex resolution and it is possible to improve the reconstruction efficiency and accuracy. Because of the short distance to the interaction point, we expect to have a background induced occupancy of up to 3% on the pixel detector. This generates an expected data rate of about 20 GB/s and exceeds the bandwidth limitations of the data storage. Based on hits in the outer detectors, back projections of particle tracks are performed and Region of Interests (ROI) on the PXD sensors are calculated. Based on those ROIs the data are reduced.In this thesis I present my development of the ROI based data reduction algorithm as well as my feasibility studies about a future 6-layer tracking.Online Data Reduction for Belle IIA first test with the whole DAQ integration and prototype sensors of PXD and SVD had been performed at DESY. For the verification of the ROI selection logic a full recording of in- and output data was included. With this setup I recorded 1.2·10^6 events containing in total 4.8·10^8 hits. The occupancy of originally 0.80% was reduced with my ROI selection logic by a factor of 6.9 to 0.12% by rejecting all hits outside any ROI.In addition I investigated the ROI positioning and got a result of a distance between ROI center and hit of 17.624±0.029 with a main offset direction of pi/2 and 3pi/2. With a more accurate position of the ROIs their size could be reduced which would optimize the reduction factor.Feasibility Studies of 6-Layer Tracking at Belle IIIn this part I investigated the limitations of the currently planned system with ROI selection and cluster rescue. Furthermore I investigated if it would be possible to increase the quality of data selection by adding a 6-layer tracking, a concept which would use all unreduced data from PXD and SVD to perform a tracking. The recent system does not include the PXD data in their tracking. In my studies I focused on two cases which could cause problems at the reconstruction of events: secondary vertex and low momentum charged particles.As an example for secondary vertex events I chose events including a K0S . I investigated that a 6-layer tracking can increase the number of reconstructable pions from K0S about 0.75% compared to the recently system. For transversal momenta of about p_t=200 MeV this increase can reach even 2%.For an example of low momentum particles I chose a decay channel of Zc(3900)- via D*- which decays in D0 and pi-. This pion has a very small momentum and may not reach enough SVD layer for a tracking without PXD. Here my result is, that the pions with such a low transversal momentum will be below the minimum ionizing momentum region. Therefore the cluster rescue covers those pions already. The increase of a 6-layer tracking will be only 0.12%.
展开▼
