Track and Vertex Reconstruction in CMS for Key Physics Processes

Track and vertex finding in LHC experiments are challenging tasks: combinatorial pattern recognition algorithms have to be made fast enough to allow the use of tracks and vertices at trigger level, in spite of the high charged particle multiplicity expected. In addition, precise estimation of track and vertex parameters is rendered difficult by the large background of soft tracks, noise hits, and non-Gaussian tails of the hit resolutions and of multiple scattering. In this paper, we describe track and vertex finding in the CMS experiment, with an emphasis on their application at High-Level Trigger. We also describe the application of robust fitting techniques to track and vertex reconstruction in CMS, in order to reduce the effect of noise and non-Gaussian tails.