摘要
本文根据排队理论和实际应用对带有优先级的信元调度问题进行了定量分析研究,推导出了每个级别的信元平均时延以及丢包率的计算公式,并进行了计算机仿真。将理论公式与仿真结果进行了对照分析,得出了要降低平均时延、减少丢包率就必须提高服务率(在缓冲区大小既定的情况下)的结论。由已得到的公式,我们可以确定将平均时延及丢包率控制在所允许的范围内的服务率的具体数值。本文所采用的分析方法克服了以往计算平均时延和丢包率的概率母函数法以及数值叠代计算方法的复杂性。
According to the queueing theory and practical applications, this paper conducts the quantitative analysis and study on cell scheduling strategies of priority class. After these , we derive the formula of the average cell delay and the cell loss ratio. Through computer simulation and contrastive analysis, we conclude that we must enhance the cell service ratio when we want to reduce the average cell delay and the cell loss ratio. Based on the formula,we can also determine the cell service ratio which curbs the average cell delay and the cell loss ratio in the permissive range. Furthermore, we deduce the situation with more priority classes. We can say that, this analytical approach in this paper overcomes the calculation complexity of the probability generation function approach and numerical methods.
出处
《计算机工程与科学》
CSCD
2005年第9期91-92,96,共3页
Computer Engineering & Science
关键词
排队
优先级
丢包率
QOS
queue
priority class
cell loss mtio
QoS