Experimental particle physics involves the construction and operation of large detectors. The detector acts as a microscope, probing the interactions of particles from high-energy accelerators at the smallest measurable distances. These detectors must be sensitive to the unique properties of long-lived particles such as the electron, muon, photon, neutrino, proton, neutron, and pion. From measurements of the individual particles or related groups of particles we can deduce the short-lived particles created in the original collision. A typical detector today may have over a million independent measuring devices synchronized to detect particle interactions at rates well in excess of a megahertz. To handle the data and select interactions of scientific interest, a multi-stage set of electronics and software known as the trigger samples a fraction of the data and makes a decision to acquire and pass on the data for further analysis.
My research presently involves work with the D0 detector at the Fermi National Accelerator Laboratory, which is located less than one hour from NIU. I have developed the design of the second state trigger (of three) to identify the tracks from muons emerging from particle interactions. This muon trigger samples about 5 kilobytes of data at a rate of 10 kilohertz and makes a decision in less than 100 microseconds. The key element of my design is a real-time processor card using digital signal processors and field programmable gate arrays to provide up to 8,000 MIPs of processing per card. The processors require specialized algorithms to find tracks from individual elements, and we are working on the simulation and timing of these efficient algorithms.
Experimental particle physics continues to develop more powerful accelerators and detectors to measure their interactions. One such new apparatus is the ATLAS detector at CERN near Geneva, Switzerland. ATLAS will be faced with data rates and trigger sampling an order of magnitude greater than the current generation of machines. High speed buffers and switches make it possible to design a trigger that will permit efficient identification of interesting events that probe interactions at ever smaller distances. Future upgrades to the Fermi accelerator and detectors will also require more sophisticated triggers and algorithms.
Selected Publications