论文部分内容阅读
提出了一种基于Hopfield神经网络 (HNN)信元调度的多重输入队列ATM交换结构 (ASF) ,消除了队头 (HOL)阻塞造成的性能恶化 .计算机仿真结果显示 ,与单先入先出 (FIFO)队列和开窗输入缓冲ASF相比 ,该方案大大提高了吞吐率并减少了信元时延 .
A multi-input queue ATM switching fabric (ASF) based on Hopfield neural network (HNN) cell scheduling is proposed to eliminate the performance degradation caused by HOL congestion. Computer simulation results show that the performance of single input-first- ) Compared to the open-window input buffer ASF, the proposed scheme greatly improves the throughput and reduces the cell delay.