基于AOS的Hurst-优先级自适应RED与动态调度算法

Hurst and priority adaptive RED combined with dynamic scheduling algorithm based on AOS

为有效传输空间数据系统中多类型、大容量、突发性的数据,基于高级在轨系统(AOS)虚拟信道复用技术,提出了一种自相似参数Hurst-优先级自适应的随机早检测(RED)与动态调度算法——HPRED-DS。该算法在队列管理中设计了基于Hurst参数与优先级的2级丢弃分组策略,在虚拟信道调度中设计了VIP/同步/异步混合的动态调度模式,并将队列管理与虚拟信道调度有效结合起来。实验结果表明该算法在保持高处理效率和吞吐率的基础上,稳定了队列长度,降低了排队延时和延时抖动,并能满足AOS中不同业务的传输要求。

In order to efficiently transmit different types of high-capacity and burst data, based on virtual channel multi- plexing technology on advanced orbiting systems （AOS）, a Hurst and priority adaptive random early detection （RED） combined with dynamic scheduling algorithm named HPRED-DS was proposed. In queue management state, the algo- rithm designed an adaptive two-level packet dropping strategy based on Hurst parameter and priority. In virtual channel scheduling state, a VIP, synchronous and asynchronous mixed virtual channels scheduling mode was designed. Mean- while, the algorithm efficiently combined queue management and virtual channel scheduling. Experimental results show that on the basis of maintaining high treatment efficiency and throughput, the algorithm stabilizes queue length, reduces delay and delay jitter, can meet various services＇ transmitting requirement.