본문 바로가기
HOME> 논문 > 논문 검색상세

학위논문 상세정보

Traffic Classification Method Using Statistical Information Extracted from Original Packet Sequence 원문보기

  • 저자

    Hyun-Min An

  • 학위수여기관

    고려대학교 일반대학원

  • 학위구분

    국내석사

  • 학과

    컴퓨터정보학과

  • 지도교수

    조충호

  • 발행년도

    2014

  • 총페이지

    vi, 45 p

  • 키워드

    traffic classification traffic identification statistical signature application traffic network congestion;

  • 언어

    eng

  • 원문 URL

    http://www.riss.kr/link?id=T13541948&outLink=K  

  • 초록

    With the rapid development of the Internet and the increased popularity of multimedia services, Internet traffic has become a big data traffic. The velocity of traffic has increased dramatically as well. These phenomena cause several problems for traffic classifications, such as increased levels of computational complexity and difficulties for real-time control. Recently, traffic classification methods have been proposed for real-time traffic classifications based on the statistical information in the flows. However, abnormal TCP behaviors and instances of cross-order can cause inconsistencies in the statistical flow information and in the analysis results. In this study, we analyze the traffic classification issues that are caused by inconsistencies in the statistical flow information and propose a novel traffic classification method using statistical information that is extracted from the original packet sequence. The proposed method reorders traffic into the original packet sequence in order to achieve consistency in the statistical information and generates statistical signatures that are used to classify application traffic. This method can classify application traffic that uses the same application protocol and encrypted traffic. The feasibility, applicability, and performance levels are evaluated by conducting experiments using several popular applications.


 활용도 분석

  • 상세보기

    amChart 영역
  • 원문보기

    amChart 영역