We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
An AQM Controller Based on Feed-Forward Neural Networks for Stable Internet.
- Authors
Bisoy, Sukant Kishoro; Pattnaik, Prasant Kumar
- Abstract
As the nature of Internet becomes nonlinear and complex, designing an intelligent controller at the router which can learn the traffic pattern of the network and predict the correct value and stabilize the system is a difficult task. In this, a new active queue management (AQM) controller is proposed based on feed-forward neural network called (FFNN-AQM) to control the network congestion efficiently by stabilizing the queue length. It learns the traffic pattern of the nonlinear and dynamic network and predicts the future value of current queue length. The parameters of neurons adjusted depending on the time-varying environment to stabilize the queue length. The NS2 network simulator is used to analyze the performance of FFNN-AQM along with existing techniques. The simulation experiment results demonstrate that FFNN-AQM is stable and achieve faster convergence with small settling time in varying network conditions. The proposed controller outperforms existing AQM proportional integral (PI), intelligent adaptive PI and neural network PI techniques.
- Subjects
ARTIFICIAL neural networks; INTERNET traffic
- Publication
Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ), 2018, Vol 43, Issue 8, p3993
- ISSN
2193-567X
- Publication type
Article
- DOI
10.1007/s13369-017-2767-9