Track clustering and vertexing algorithm for L1 trigger

One of the keystones of the canceled BTeV experiment (proposed at Fermilab's Tevatron) was its sophisticated three-level trigger[1]. The trigger was designed to reject 99.9% of light-quark background events and retain a large number of B decays. The BTeV Pixel Detector provided a 3-dimensional, high resolution tracking system to detect B signatures. The Level 1 pixel detector trigger was proposed as a two stage process, a track-segment finder and a vertex finder which analyzed every accelerator crossing. In simulations the track-segment finder stage outputs an average of 200 track-segments per accelerator crossing (2.5MHz). The vertexing stage rinds vertices and associates track-segments with the vertices found. This paper proposes a novel adaptive pattern recognition model to find the number and the estimated location of vertices, and to cluster track-segments around those vertices. The track clustering and vertex finding is done in parallel. The pattern recognition model also generates the estimate of other important parameters such as the covariance matrix of the cluster vertices and the minimum distances from the tracks to the vertices needed to compute detached tracks.