OMEGA-NETWORK-BASED ATM SWITCH WITH NEURAL NETWORK-CONTROLLED BYPASS QUEUING AND MULTIPLEXING

Multistage interconnection networks (MIN's) have long been studied for use in switching networks, Since they have a unique path between sour ce and destination and the intermediate nodes of the paths are shared, internal blocking can cause very poor throughput, This paper proposes a high throughput ATM switch consisting of an Omega network with a ne w form of input queues called bypass queues, We also improve the switc h throughput by partitioning the input buffers into disjoint buffer se ts and multiplexing several sets of nonblocking cells within a time sl ot, assuming that the routing switch operates only a couple of times f aster than the transmission rate, A neural network model is presented as a controller for cell scheduling and muitiplexing in the switch. Ou r simulation results under uniform traffic show that the proposed appr oach achieves almost 100% of potential switch throughput,