Dynamic control of communication systems based on simple recurrent neural networks

A simple recurrent neural network called diagonal recurrent neural network (DRNN) is used for dynamic control of communication systems, particularly to dynamic congestion control in broadband ATM networks. The architecture of DRNN is a modified model of the fully connected recurrent neural network with one hidden layer, and the hidden layer is comprised of self-recurrent neurons. Two DRNNs´ are utilized in the control system, one as an identifier called diagonal recurrent neuroidentifier (DRNI) and the other as a controller called diagonal recurrent neurocontroller (DRNC). The DRNI is used to behave like the real network and to translate the distal error signal between the required performance bound and the performance observed from the real network. The DRNC is used to control adaptively regulating access of external traffic into the network to guarantee the desired performance given in the form of performance bound. Simple dynamic queuing models for the presentation of real networks are used to test the performance of the suggested control scheme