基于弹性定额值的分组轮询调度算法

摘要

提出了一种新的适用于变长分组的调度算法——弹性定额值轮询调度算法(Resilient Quantum Round Rob-in,RQRR),与现有算法不同,该算法中每个数据流的定额值不是固定不变的,定额值的生成依赖于前一个轮次中各个数据流的发送情况.理论分析表明,RQRR可以保证数据流之间具有较好的公平性,它的公平性度量具有上界值7Max-1,其中Max为分组的最大长度.RQRR对每个分组的处理复杂度为O(1),易于实现、适用于高速网络.%Aiming at variable length packet queues,a new packet scheduling algorithm named Resilient Quantum Round Robin(RQRR)was presented.Different from existent algorithms,the quantum given to each of the flows in a round is not fixed and is calculated depending on the transmission situation of all the flows in the previous round.The theoretical analyses show that the relative fairness measure of RQRR has an upper bound of 7Max—1,where Max is the largest size of the packets.RQRR takes O(1)processing complexity per packet and is simple to implement at high-speed networks.