Intelligent congestion control in ATM networks

In large modern telecommunication networks, the amount of traffic and the number of nodes and links are so large that the traditional network control may not be effective due to high complexity. These networks need adaptive and intelligent systems in order to provide high network reliability, accurate traffic prediction, efficient use of channel bandwidth, and optimized network management in relation to various, dynamically changing environments. Neural networks can contribute to this emerging new telecommunication infrastructure by providing fast, flexible, adaptive, and intelligent control that cannot be performed sufficiently well by digital computers. In this paper, we discuss the neural network approaches for solving various control problems in high-speed communication networks, and present our proposed neural network model for the optimized control of input queues in ATM switches.