Particle tracking at high luminosities using a novel reconstruction approach

Tracking charged particles with high precision is of vital importance for collider experiment like those operating at the Large Hadron Collider (LHC), CERN. The tracking detector in the CMS experiment is composed of multi-layer silicon based tracker with 3-dimensional position sensitivity. High precision position data from tracker operated in high magnetic field, is used to reconstruct the trajectories of charged particles and obtain their kinematic parameters ($p_t$,$\eta_0$, $\phi_0$) with high accuracy. In this paper, for Phase2 CMS tracker design, we present a novel track reconstruction algorithm for high luminosity (HL) era of the LHC. The algorithm identifies hits associated with each track and utilizes them to accurately determine the kinematic parameters of each track using machine learning (ML) architecture. The proposed algorithm has been applied on a large sample of hard interactions simulated at high luminosity (HL) era of the LHC using Pythia8 and Geant4 framework for equivalent geometry of the outer tracker of the CMS experiment. Performance of the proposed algorithm has been studied using the key indicators such as reconstruction efficiency, fake rate and resolution. Comparison with traditional methods demonstrates robust performance with excellent efficiency and resolution with minimal fake rate.