Implementation of neural network-controlled Banyan network with input-output-queueing

The Banyan network is a kind of multistage interconnection networks (MINs), and the most popular to be used as a basic building block in ATM switch fabric designs. The neural networks scheduling of MINs has been proposed in some papers, but its hardware implementation is complicated. In this paper we propose a high throughput ATM switch which consists of neural network controlled Banyan network with input-output-queueing, and give a method of hardware implementation. Its neural network controller is very simple.