【摘要】 提出一种基于ＢＰ神经网络的ＡＴＭ接入允许控制方案．该方案不仅具有自适应实时控制的优点，而且不依赖于呼入的种类个数，同时也无需呼入参数的更详细描述，因而比以往提出的方案具有更大的灵活性和适应面．模拟结果表明该方案有充分大的接入区域和抗干扰能力更多还原

【Abstract】 A call admission control (CAC) scheme for ATM networks based on BP neural network is presented. Being independent on call class number, the scheme has advantage of adaptive control and no extra parameter descriptions of calls. The scheme can be applied easily and widely. The extreme large acceptance region and good robustness are obtained in the simulation.更多还原