论文部分内容阅读
主动队列管理通过网络中间节点有控制的分组丢弃实现了较低的排队延时和较高的有效吞吐量,是TCP端到端拥塞控制近来研究的一个技术热点.已有的大多数算法在判定分组丢弃时大都沿袭了RED的概率丢弃机制,具有一定计算复杂度的随机数生成过程不利于路由器性能的优化.在本文中,我们首先定义了拥塞指数这一新的测度变量来量化描述网络的拥塞状态.接着,利用模糊逻辑设计了一种新的智能分组丢弃机制,离线的合成推理使得分组丢弃的判定仅需要简单的查表操作和比较运算即可完成,为优化路由器的性能提供了便利.数字仿真的结果表明:智能分组丢弃机制的性能优于经典的RED算法,控制队列的能力强,鲁棒性好,稳定工作域大,能很好地抵抗突发性和非弹性业务的干扰,适合工作在瞬息万变的动态网络环境中.
Active queue management achieves lower queuing delay and higher effective throughput by controlling packet discarding by intermediate nodes in the network, which is a hotspot of recent research on TCP end-to-end congestion control.Most of the existing algorithms are in the process of determining In the dissertation, we firstly define the new measure variable of congestion index to quantify the network performance. In this paper, we first define the probability of discarding mechanism of RED, and the process of generating random numbers with some computational complexity is not conducive to the optimization of router performance.In this paper, Congestion state.Secondly, a new intelligent packet discarding mechanism is designed by using fuzzy logic. The off-line synthesis reasoning makes the decision of packet discarding only need simple look-up table operation and comparison operation, which is convenient for optimizing the performance of the router The results of digital simulation show that the performance of intelligent packet discarding mechanism is superior to that of classic RED algorithm, the ability to control queues is strong, robustness is good, and the stable work area is large, which can well resist the interference of burst and inelastic services , Suitable for working in a rapidly changing dynamic network environment.