The CMS High Level Trigger

CMS Trigger,Data Acquisition Group

DOI: https://doi.org/10.1140/epjc/s2006-02495-8

2006-01-01

Abstract:At the Large Hadron Collider at CERN the proton bunches cross at a rate of 40MHz. At the Compact Muon Solenoid experiment the original collision rate is reduced by a factor of O (1000) using a Level-1 hardware trigger. A subsequent factor of O(1000) data reduction is obtained by a software-implemented High Level Trigger (HLT) selection that is executed on a multi-processor farm. In this review we present in detail prototype CMS HLT physics selection algorithms, expected trigger rates and trigger performance in terms of both physics efficiency and timing.

High Energy Physics - Experiment

What problem does this paper attempt to address?

### Problems the paper attempts to solve

This paper aims to solve the problem of how to efficiently screen out physically significant events from a vast number of collision events in the Large Hadron Collider (LHC). Specifically, the paper discusses the design and performance optimization of the High - Level Trigger (HLT) in the CMS experiment.

#### Background and challenges

1. **High collision rate**: The proton beams in the LHC collide at a rate of approximately 40 MHz per second, which means a large amount of data is generated every second.
2. **Huge amount of data**: The amount of data for each event is approximately 1 MB, so the amount of raw data generated per second is extremely large.
3. **Limited storage and processing capabilities**: Due to limited storage and processing resources, it is necessary to quickly screen out meaningful events to reduce the amount of data to a manageable level.

#### Solutions

To address the above challenges, the paper proposes and describes in detail the multi - level trigger system in the CMS experiment, especially the software - implemented High - Level Trigger (HLT). Specific measures include:

1. **Level - 1 Trigger**:
- Implemented by hardware, reducing the data rate from 40 MHz to approximately several thousand hertz (kHz).
- Using specific low - level analysis algorithms to quickly screen events in custom - made trigger processors.

2. **High - Level Trigger (HLT)**:
- Execute more complex software filtering algorithms on a multi - processor cluster to further reduce the data rate to several hundred hertz (Hz).
- Implement partial event reconstruction, only reconstructing the information of physical objects required for selection to save CPU resources.
- The algorithm design takes into account different types of physical objects such as electrons, photons, muons, jets, etc., and sets thresholds according to the requirements of different physical processes.

3. **Optimization strategies**:
- **Fast confirmation**: For example, use the hit information in the pixel detector to quickly confirm electron candidates.
- **Partial reconstruction**: Only reconstruct the parts of the event related to the objects of interest, rather than the entire event.
- **Dynamic adjustment**: Dynamically adjust the thresholds and allocate bandwidth of the trigger according to different luminosity conditions (low luminosity and high luminosity).

#### Goals

The ultimate goal is to ensure that valuable physical events are screened out as efficiently as possible without affecting the physical efficiency, for subsequent offline analysis. This requires not only efficient algorithm design but also a comprehensive evaluation and optimization of the performance of the trigger system.

Through these measures, the paper shows that the CMS experiment has developed an efficient and flexible trigger structure that can select the physical signals of interest under the harsh conditions of the LHC.