DocumentCode :
2560416
Title :
Simulation-framework for purely digital CNN/MRF-architectures
Author :
Stilkerich, Stephan C.
Author_Institution :
Image & Signal Process. Group, EADS Corporate Res. Center, Munich, Germany
fYear :
2005
fDate :
28-30 May 2005
Firstpage :
94
Lastpage :
97
Abstract :
Any kind of hardware-relevant modeling, simulation and analysis of purely digital and massively parallel architectures, which is based on CNN/MRF processing principles is a time consuming, computing resources intensive, fault-prone and complex task. Until now there is no industrially qualified toolkit available to systematically support these tasks in a single closed environment. In this paper we present a novel simulation-framework for purely digital CNN/MRF processing systems. The unique modeling, hardware-relevant simulation and analysis capabilities unified in this simulation-framework allows it to systematically investigate (1) the massively parallel processing dynamic of digital CNN/MRF devices, (2) the model´s convergence behavior and (3) complete CNN/MRF systems with an application-specific size. The paper is finalized by simulation results demonstrating the ability of the framework to handle CNN/MRF processing systems of realistic size and complexity. This manifests the industrial relevance of the proposed CNN/MRF simulation-framework.
Keywords :
Markov processes; cellular neural nets; neural net architecture; parallel processing; random processes; Markov random field; massively parallel processing; purely digital CNN/MRF-architectures; simulation framework; Analytical models; Application specific integrated circuits; Cellular neural networks; Computational modeling; Environmental management; Infrared sensors; Mathematical model; Radiation hardening; Signal processing; Space technology;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Cellular Neural Networks and Their Applications, 2005 9th International Workshop on
Print_ISBN :
0-7803-9185-3
Type :
conf
DOI :
10.1109/CNNA.2005.1543169
Filename :
1543169
Link To Document :
بازگشت