Fast triggering in high-energy physics experiments using hardware neural networks

@article{Denby2003FastTI,
  title={Fast triggering in high-energy physics experiments using hardware neural networks},
  author={Bruce Denby and Patrick Garda and B. Granado and Ch. Kiesling and Jean-Christophe Pr{\'e}votet and Andreas Wassatsch},
  journal={IEEE transactions on neural networks},
  year={2003},
  volume={14 5},
  pages={
          1010-27
        }
}
High-energy physics experiments require high-speed triggering systems capable of performing complex pattern recognition at rates of Megahertz to Gigahertz. Neural networks implemented in hardware have been the solution of choice for certain experiments. The neural triggering problem is presented here via a detailed look at the H1 level 2 trigger at the HERA accelerator, Hamburg, Germany, followed by a section on the importance of hardware preprocessing for such systems, and finally some new… 
View on PubMed
neurones.espci.fr
13 Citations
Background Citations
3
Methods Citations
1

13 Citations

Citation Type

Citation Type

Stability study of the neural network at particle physics detectors
TLDR
A gas-filled detector was used to study the trigger at different noise levels to find the most stable neural network that tolerates the random hits.
Opportunities for statistical signal processing in high energy physics
  • B. Denby
  • Physics
    IEEE/SP 13th Workshop on Statistical Signal Processing, 2005
  • 2005
TLDR
It will be argued that the falling cost of computing hardware favors the growth of statistical signal processing methods in high energy physics, and the dominant roles of model dependence and systematic errors in final physics analyses render statistical signalprocessing techniques largely inapplicable at this level.
Neural network hardware architecture for pattern recognition in the HESS2 project
TLDR
The problem of implementation of neural network in the context of the level 2 trigger of HESS2 project is considered and a hardware architecture which takes advantage of high parallelism, pipelining and the intrinsic nature of FPGAs is proposed.
A hardware implementation of artificial neural networks using field programmable gate arrays
Exploring liquid computing in a hardware adaptation: construction and operation of a neural network experiment
TLDR
This thesis describes the construction of an experiment dedicated to exploring silicon adaptations of artificial neural network paradigms for their general applicability, power efficiency, and fault-tolerance, and first time adaptation to a CMOS neural network affirms this claim.
Artificial neural networks-based track fitting of cosmic muons through stacked resistive plate chambers
The India-based Neutrino Observatory (INO) collaboration, as part of its detector R\&D program, has developed prototype stacks of resistive plate chambers (RPCs) to study their performance. These…
Stepwise evolutionary training strategies for hardware neural networks
TLDR
A stepwise training approach has been devised that allows for the use of simple evolutionary algorithms to efficiently optimize the synaptic weights of a fast mixed-signal neural network chip and provides efficient means for the training of large networks on the used mixed-Signal chip for demanding real-world classification tasks.
A novel calorimeter trigger concept: The jet trigger of the H1 experiment at HERA
A Survey of Neuromorphic Computing and Neural Networks in Hardware
TLDR
An exhaustive review of the research conducted in neuromorphic computing since the inception of the term is provided to motivate further work by illuminating gaps in the field where new research is needed.
Prototype neural network hardware with analog electronic circuit
TLDR
A model for hardware of neural network with four synaptic input links and corresponding weights constructed with operational amplifier can generate sigmoid functions of varied steepness and outputs for random variations of inputs, weights and threshold.
...
1
2
...

References

SHOWING 1-10 OF 48 REFERENCES
Hardware preprocessing for the H1-Level 2 neural network trigger upgrade
TLDR
An improved "intelligent" preprocessing has been devised to provide increased selectivity at the higher luminosity planned for the HERA upgrade, and a new preprocessing board is currently being designed at the Max Planck Institute for Physics.
Performance of the CDF neural network electron isolation trigger at Fermilab
Realization of a second level neural network trigger for the H1 experiment at HERA
An analog neural network processor with programmable topology
TLDR
The architecture, implementation, and applications of a special-purpose neural network processor are described and the practicality of the chip is demonstrated with an implementation of a neural network for optical character recognition.
Neural Network Adaptations to Hardware Implementations
TLDR
In this section an overview is given of the various issues that are encountered when mapping an ideal neural network model onto a compact and reliable neural network hardware implementation, like quantization, handling nonuniformities and nonideal responses, and restraining computational complexity.
Spert-II: A Vector Microprocessor System
TLDR
A prototype full custom vector microprocessor, TO, is packaged as the Spert-II (Synthetic Perceptron Testbed II) workstation accelerator system, to accelerate multiparameter neural network training for speech recognition research.
The H1 detector at HERA
SYNAPSE - A Neurocomputer that Synthesizes Neural Algorithms on a Parallel Systolic Engine
  • U. Ramacher
  • Computer Science
    J. Parallel Distributed Comput.
  • 1992
DOLFIN-digit online for integration neural networks
  • A. Wassatsch, M. Haase, D. Timmermann
  • Computer Science, Mathematics
    2000 IEEE International Symposium on Circuits and Systems. Emerging Technologies for the 21st Century. Proceedings (IEEE Cat No.00CH36353)
  • 2000
TLDR
The goal of this paper is to develop a strategy for the implementation of different network models with significant advantages over all other digital implementations by using digit online arithmetic in the field of neural network computation.
Design of high speed MOS multiplier and divider using redundant binary representation
TLDR
This work improved the algorithm and the method of implementation, and designed an advanced multiplier and divider for MOS LSI based on a new algorithm that has several excellent features such as high speed addition operations.
...
1
2
3
4
5
...