Upgrade of the CMS Barrel Muon Track Finder for HL-LHC featuring a Kalman Filter algorithm and an ATCA Host Processor with Ultrascale+ FPGAs

Abstract

The Barrel Muon Track finder of the CMS experiment at the Large Hadron Collider uses custom processors to identify muons and measure their momenta in the central region of the CMS detector. An upgrade of the L1 tracking algorithm is presented, featuring a Kalman Filter in FPGAs, implemented using High Level Synthesis tools. The matrix operations are mapped to the DSP cores reducing resource utilization to a level that allows the new algorithm to fit in the same FPGA as the legacy one, thus enabling studies during nominal CMS data taking. The algorithm performance has been verified in CMS collisions during 2018 operations. The algorithm is also proposed for standalone muon tracking at the High Luminosity LHC. The algorithm development is complemented by ATCA processor R&D featuring a large ZYNQ Ultrascale+ SoC with high speed optical links.