一种基于排队论的主动队列拥塞控制算法

A new congestion control algorithm for active queue based Queuing theory

  • 摘要: 主动队列(AQM)算法存在诸多的不足,如排队延时大、时延抖动性强、数据进队与出队速率不匹配等,从而导致数据的滞留、丢失和振荡.基于此,在BLUE算法的基础上,引入M/M/m (n)排队系统的思想,提出一种新的算法---PBLUE.该算法根据稳定状态下的平衡方程来保证队列长度的稳定性,增加扩充因子调节路由器的缓存来快速恢复丢失的数据.通过仿真实验,改进的算法降低了丢包率,提高了带宽利用率,并稳定了队列长度.

     

    Abstract: AQM algorithm has many deficiencies, such as a large queuing delay, strong delay jitter, the rate of the data into and out the team not matched, and thus it always leads to data retention, loss and oscillation. Based on this, the M/M/m (n) queuing system thought based on the BLUE algorithm is introduced which is called PBLUE. The algorithm ensures the stability of the queue length according to equilibrium equation under stable state. Meanwhile, it increases the expansion factor and adjusts the router's cache for quickly recovering of the lost data. In the simulation, the improved algorithm reduced the packet loss rate and increased the bandwidth utilization and stabled the queue length.

     

/

返回文章
返回