Neural Network-Based Flow Control for ABR Services in ATM Networks

The aim of Neural Network Flow control for ABR (Available Bit Rate) services in Asynchronous Transfer Mode (ATM) Networks is to develop an efficient rate control mechanism to address the real-time transport (RTT) variation of RM-Cell caused by the dynamic drift of the bottleneck in a multicast tree. By optimizing rate control, we seek to mitigate the impact of RTT variations on network performance. Additionally, our focus includes enhancing buffer occupancy to improve throughput. This involves implementing a rate control system based on neural networks (NN), leveraging their adaptive capabilities to adjust transmission rates in response to network conditions dynamically. Through this project, we aim to achieve more stable and reliable data transmission within multicast tree networks, ultimately enhancing overall network efficiency and performance. The potential impact of this research is significant, inspiring further exploration and development in the field.