Model Predictive Neural Control of TCP Flow in AQM Network
Kourosh J. Rahnamai, Kevin Gorman, Andrew Gray · 2006
Many research papers have been published on RED (random early detection) and variants of RED. Recently many articles have been presented on modeling a transmission control protocol (TCP) flow in an active queue management (AQM) of a bottlenecked network link (Dong Lin and R. Morris, 1997), (V. Misra, et al., 2000), (C. Hollot, et al., 2001). Classical control theories have also been applied to achieve or improve stability of the network flow (C. Hollot, et al., 2001). In this paper we present a neural network (NN) model predictive control (MPC) of TCP flows. We show the robust adaptive behavior of the MPC optimal controller under modeling errors and system dynamic changes. We also show the superior transient and steady state behavior as well as general stability of MPC as compared to the classical PI controller