The Algorithm Steering and Trigger Decision mechanism of the ATLAS High Level Trigger

Abstract

Given the extremely high output rate foreseen at LHC and the general-purpose nature of ATLAS experiment, an efficient and flexible way to select events in the High Level Trigger is needed. An extremely flexible solution is proposed that allows for early rejection of unwanted events and an easily configurable way to choose algorithms and to specify the criteria for trigger decisions. It is implemented in the standard ATLAS object-oriented software framework, Athena. The early rejection is achieved by breaking the decision process down into sequential steps. The configuration of each step defines sequences of algorithms which should be used to process the data, and 'trigger menus' that define which physics signatures must be satisfied to continue on to the next step, and ultimately to accept the event. A navigation system has been built on top of the standard Athena transient store (StoreGate) to link the event data together in a tree-like structure. This is fundamental to the seeding mechanism, by which data from one step is presented to the next. The design makes it straightforward to utilize existing off-line reconstruction data classes and algorithms when they are suitable