基于M/G/1排队理论的VOIP网络延时分析  

Delay Analysis of M/G/1 Queuing Theory-based VoIP Network

作　　者：肖龙 李云峰 罗祖川 Xiao Long;Li Yunfengg;Luo Zuchuan(Ning Xia Branch of Civil Aviation Air Traffic Management Bureau,Yinchuan 750001 Ningxia China)

机构地区：[1]中国民用航空西北地区空中交通管理局宁夏分局,宁夏银川750001

出　　处：《中国民航飞行学院学报》2023年第6期22-26,共5页Journal of Civil Aviation Flight University of China

摘　　要：为了解决VoIP通信网络中语音包传输延时导致网络拥塞降低话音通信质量的问题,针对延时分量的特点对Vo IP通信网络中语音包和数据包两类客户提出了一种具有非强占优先权的M/G/1排队模型。首先阐述了网络传输中延时产生;其次提出了具有非强占优先权的M/G/1排队模型;最后给出了排队模型的数学表达式并进行了数值分析和图形分析,结果表明提出的非强占优先权排队模型中语音包(Ⅰ类客户)延时相对数据包(Ⅱ类客户)较小,解决了传统强占型优先权排队模型中Ⅰ类客户延时减小导致Ⅱ类客户延时大幅增加的情况,避免了计算机通信网络中网络拥塞的风险。In order to solve the problem that the transmission delay of voice packets in the VoIP communication network causes network congestion and reduces the quality of voice communication,according to the characteristics of the delay component,the M/G/1 queuing model with a non-preemptive priority is proposed for the two types of voice packets and data packets in the VolP communication network.Firstly,it explains the delay generation in network transmission;secondly,it proposes the M/G/1 queuing model with non-preemptive priority;finally,it gives the mathematical expression of the queuing model and carries out the numerical analysis and graphical analysis.The results show that with the proposed queuing model with non-preemptive priority,the delay of voice packets(type Ⅰ customers)is smaller than that of data packets(type Ⅱ customers),which solves the problem of the reduced delay of type Ⅰ customers in the traditional preemptive priority queuing model leading to the delay of type Ⅱ customers.The substantially increased situation avoids the risk of network congestion in the computer communication network.

关 键 词：基于IP的语音传输 端到端延时 广义随机PETRI网 重试排队系统 M/G/1 

分 类 号：TN912.3[电子电信—通信与信息系统] O226[电子电信—信息与通信工程]