参数自适应的随机早期检测算法

Enhanced RED algorithm with adaptive parameters

主动队列管理(AQM)算法是最近网络拥塞控制研究的重点,随机早期检测(RED)算法作为AQM算法的代表受到广泛的关注。分析了随机早期检测算法的原理和局限性,针对RED算法参数配置困难的问题,提出了一种参数自适应的随机早期检测算法。该算法能够根据网络负荷的变化动态的调节RED算法的参数,从而能够在突发业务下对拥塞做出及时有效的反应。通过仿真表明该算法能够有效地降低丢包率,降低数据包排队延时,提高了RED算法的网络适应性。

Active queue management （AQM） algorithm is the stress on network research in recent years. As the representative of AQM algorithm, the random early detection （RED） has got abroad attention. The elements and limitation of RED algorithm are investigated. To overcome the difficulty in right parameters determination, an enhanced RED algorithm with adaptive parameters is proposed, which can adjust the parameters according to the variation of network load. So this algorithm can respond to the congestion more effectively in time. The simulation results show that this enhanced RED algorithm can effectively reduce packet loss and the delay of the queuing packet. The algorithm is improved the adaptability of the RED algorithm.