基于优先级的卫星终端双队列缓存管理算法

Satellite Terminal Dual Queue Buffer Management Algorithm Based on Priority

卫星终端传输的业务种类多、业务量大,容易引起由于业务突发造成的拥塞现象,导致较大的丢包率和较低吞吐量。针对上述问题,提出业务的优先级,在单队列缓存RED算法的基础上进行改进,提出一种采用优先级的双队列缓存管理算法-DBQM-P算法。上述算法对不同业务进行优先级分配,判断双队列缓存业务量与门限值的关系,对不同优先级队列执行不同的丢包概率。仿真结果表明,DBQM-P算法兼顾了不同优先级业务的传输要求,与单队列缓存RED算法对比,有效地降低了业务的丢包率,提高了系统的吞吐量,减小了卫星终端网络突发业务带来的拥塞。

Various business types and large business volume of satellite easily cause congestion, which may result in traffic burst, a large packet loss rate and lower throughput. To solve this problem, this paper proposes a dual queue buffer management algorithm based on the priority - DBQM - P algorithm, which is on the basis of a single queue buffer improved RED algorithm. The algorithm allots the priorities for different businesses, and determines the volume of business with dual queue buffer threshold relations, performs different packet drop probability for different priority queues. Simulation results show that the DBQM - P algorithm takes into account the transmission require- ments of different business priorities. Compared with a single queue buffer RED algorithm, the DBQM - P algorithm effectively reduces the packet loss rate, improves the throughput of the system and reduces bursty traffic to congestion of the satellite terminal network.