摘要
准确、高效的业务流识别和分类是保障多媒体通信端到端服务质量(Quality of Service,QoS)和执行相关网络操作的前提。但多媒体通信业务构成复杂,具有较严格的QoS约束,且存在包/流水平统计特征多样性,因此业务统计特征有效选取直接关系到识别和分类方法的有效性。针对流行的多媒体业务,分析了典型的业务特征,从业务QoS保证角度,选取区分特征,基于隐马尔可夫模型(Hidden Markov Model,HMM),对多媒体业务在QoS类上进行区分,实现简单,能以较小的空间复杂度较快地识别出多媒体业务流,有利于提高分类准确度。通过仿真验证了该方法的有效性。
The accurate and efficient identification/categorization of multimedia traffic is the premise of end-to-end QoS guarantees and corresponding network operations.The traffic structure of multimedia communications is very complex,and the most traffic of multimedia communications require strict QoS.Meanwhile,the statistical characteristics of multimedia traffic are diverse at packet/flow level.Therefore,it is vital to select appropriate traffic characteristics in packet/flow level for efficiently identifying multimedia traffic.This paper analyzes some typical flow characteristics for prevalent multimedia traffics,selects some differentiating characteristics from the point of view of QoS requirements,and designs a multimedia traffic QoS classification method based on HMM(Hidden Markov Model)to differentiate multimedia traffics according to QoS class.Finally,the simulation results are given to demonstrate the effectiveness of the proposed method.
作者
王再见
张有健
WANG Zaijian;ZHANG Youjian(College of Physics and Electronic Information,Anhui Normal University,Wuhu 241000,China)
出处
《无线电通信技术》
2019年第1期35-41,共7页
Radio Communications Technology
基金
安徽省高校领军人才引进与培育计划(gxfx ZD2016013)
安徽师范大学博士科研启动金项目(2016XJJ129)
安徽省自然科学基金项目(1608085QF138)
安徽省高等学校省级自然科学研究项目(KJ2015A105
KJ2018A0311)
关键词
多媒体通信
流识别
业务区分
隐马尔可夫模型
multimedia communications
flow identification
traffic classification
hidden Markov model