Online Transverse Beam Instability Detection in the LHC : High-Throughput Real-Time Parallel Data Analysis

• Main Author: Söderén, Martin
• Format: Others
• Language: English
• Published: Linköpings universitet, Programvara och system 2017
• Subjects: parallel; computing; instability; detection; cern; hilbert; pipeline; dsp; Computer Systems; Datorsystem
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-143440

This thesis presents the ADT transverse instability detection system, the next generation of instability detection in the LHC at CERN, Geneva. The system is presented after a thorough study of underlying causes for instabilities in high energy particle accelerators, current parallel programming paradigms, the available hardware and software at CERN and possible instability detection techniques. The requirements for the system involve handling vast amounts of data which need to be analyzed in real-time and in this data detect rapid amplitude growth while limiting the computational resources required to a minimum. The result of this thesis was a system that could generate a trigger when an instability was detected, which was used to save data from observation instruments around the LHC. A fixed display in the CERN control centre was also created which allows scientists and operators at CERN to monitor the oscillation amplitude of all particle bunches. The conclusion is that the complete system will be a valuable asset at CERN to help further develop the LHC.