Title :
Modeling of network switch controlled by neural network
Author :
Polivka, Michal ; Skorpil, Vladislav
Author_Institution :
Dept. of Telecommun., Brno Univ. of Technol., Brno, Czech Republic
Abstract :
Conventional switches are usually controlled by a manually defined algorithm implemented in the network processor. The conventional methods of control are fast, reliable and transparent. The article deals with alternative method of switch control - based on a neural network, focused on Quality of Service. The article describes the designed model, simulation and summarizes the results.
Keywords :
backpropagation; feedforward neural nets; neural chips; switching networks; feedforward backpropagation neural network; network processor; network switch controlled modeling; quality-of-service; Arrays; Fabrics; Generators; Mathematical model; Neural networks; Switches; MATLAB; modeling; neural network; switch; switching algorithm;
Conference_Titel :
Telecommunications and Signal Processing (TSP), 2012 35th International Conference on
Conference_Location :
Prague
Print_ISBN :
978-1-4673-1117-5
DOI :
10.1109/TSP.2012.6256274
