Implementation And Performance of the ATLAS Second Level Jet Trigger Page: 2 of 7
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to UNT Digital Library by the UNT Libraries Government Documents Department.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
The ATLAS trigger system is organized in three levels. LVL1 (first level) is hardware based,
with coarse granularity and a maximum latency of 2 ps. Using only the information from the
muon system and the calorimeters, it does a fast search for high PT objects, reducing the rate
from 40 MHz to about 75 kHz. The High Level Trigger (HLT) formed by the Second Level Trigger
(LVL2) and the Event Filter (EF) is software based. LVL2 uses the information provided by
LVL1 about the position of the high transverse energy energy deposition (seeds) to start the
search for interesting signatures. It is run before the event is built and therefore it asks for the
event fragments to the Read-Out Buffers (ROB). To save time it only reconstructs a region of
the detector around the seed provided by LVL1, that is known as Region of Interest (Rol). It
uses the full granularity of the subdetectors and has a maximum processing time of the order
of 40 ms in a 1.8 GHz processor, reducing the rate by a factor 100, approximately. The EF
may run in seeded mode, receiving the seeds from LVL2, or in full event access. It runs more
sophisticated algorithms and has access to better calibration constants, improving the selection
done in the previous steps. It has to reduce the rate to about 200 Hz.
The EF uses an offline-like environment and algorithms, while LVL2 has fast, dedicated
algorithms that have to run in a multi-threaded environment. In both levels, a sequence of
algorithms is run in order to verify or reject as fast as possible a given signature [1].
2. The ATLAS Jet Trigger
At the HLT, jets with high PT are selected by running a sequence of algorithms. At LVL2, the
first one is a feature extraction algorithm, called TrigT2CaloJet, that uses a basic cone algorithm
to reconstruct a jet in the Rol. After that, a hypothesis algorithm is run in order to verify if
the PT of the jet is over the threshold. At the EF, offline-like algorithms are used. Typically, a
more sophisticated cone algorithm is run, although there is the possibility to choose also kT or
other offline available algorithms.
TrigT2CaloJet is an iterative algorithm that starts by requesting to the Read-Out Buffers
the event fragments in a region around the seed provided by LVL1. It receives the data in a
bytestream format that has to be unpacked (translated into the C++ objects that the algorithm
uses). This is one of the most time consuming parts of the algorithms. To be able to reduce the
unpacking time, three possible granularities are being considered:
" Cells: the standard one. Each cell in the Rol is unpacked.
" FEBs: the energy sums of all the cells belonging to a Front End Board (FEB) are calculated
at the Read-Out Driver (ROD) and sent directly to LVL2, that unpacks only the information
corresponding to the FEB energy sums. For the electromagnetic calorimeter, each FEB
contains 128 cells, reducing considerably the amount of data transfered and unpacked.
" The possibility of using the information of LVL1 Trigger Towers is also under consideration
but its feasibility is not yet demonstrated.
After unpacking the data, a grid with the list of detector elements in the Rol is built. The
algorithm loops in this grid and calculates the energy weighted 'q, 0 position3 of the detector
elements (cells, FEBs) inside a cone with fixed radius4 around the center of the Rol. The center of
the cone is updated and the calculation is iterated few times in order to improve the measurement
of the jet center position. Reducing the granularity by using FEB based energy sums has the
advantage that the time to loop over the detector elements is shorter, improving even further
the algorithm performance.
3 The ATLAS coordinate system is a right-handed system with the x axis pointing towards the center of the
LHC ring, the z axis following the direction of the beam and the y axis pointing upwards. The azimuthal angle,
0 is measured from the positive x axis and the polar angle 0 is measured from the positive z axis. Frequently, the
pseudorapidity variable, defined as = -log (tan 2),is used.
4 The cone radius is measured as a distance in the (n,q) plane, defined as R = (Aq)2 + (A#)2
Upcoming Pages
Here’s what’s next.
Search Inside
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Conde Muino, Patricia; Aracena, I.; Brelier, B.; Cranmer, K.; Delsart, P. A.; Dufour, M. A. et al. Implementation And Performance of the ATLAS Second Level Jet Trigger, article, November 9, 2011; United States. (https://digital.library.unt.edu/ark:/67531/metadc834122/m1/2/: accessed April 25, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.