DocumentCode :
2205531
Title :
Neural networks for multiprocessor real-time scheduling
Author :
Cardeira, Carlos ; Mammeri, Zoubir
Author_Institution :
CRAN, ENSEM, Vandoeuvre les Nancy, France
fYear :
1994
fDate :
15-17 Jun 1994
Firstpage :
59
Lastpage :
64
Abstract :
In recent years, neural networks have become a popular area of research, especially after Hopfield and Tank opened the way for using neural networks for optimization purposes and surprised the scientific community by their paper (Biological Cybernetics, vol. 52, pp. 141-52, 1985) presenting a circuit to give approximate solutions for the classical traveling salesman problem in a few elapsed propagation times of analog amplifiers. In this paper, we analyse Hopfield neural networks from the scheduling viewpoint to see if they can be used to solve real-time scheduling problems. We build a neural network whose topology depends on real-time task constraints, and converges to an approximate solution of the scheduling problem. Finally, we analyse the quality of the result in terms of the convergence rate and the complexity of the algorithm
Keywords :
Hopfield neural nets; computational complexity; convergence; multiprocessing systems; network topology; optimisation; real-time systems; scheduling; Hopfield neural networks; algorithm complexity; analog amplifiers; approximate solution; convergence rate; elapsed propagation times; multiprocessor real-time scheduling; network topology; optimization; real-time task constraints; traveling salesman problem; Algorithm design and analysis; Analog circuits; Cities and towns; Hopfield neural networks; Image converters; Network topology; Neural networks; Shape; Signal processing algorithms; Speech recognition;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Real-Time Systems, 1994. Proceedings., Sixth Euromicro Workshop on
Conference_Location :
Vaesteraas
Print_ISBN :
0-8186-6340-5
Type :
conf
DOI :
10.1109/EMWRTS.1994.336864
Filename :
336864
Link To Document :
بازگشت