• DocumentCode
    3135627
  • Title

    GOSSIB vs. IP traceback rumors

  • Author

    Waldvogel, Marcel

  • Author_Institution
    Zurich Res. Lab., IBM Res., Zurich, Switzerland
  • fYear
    2002
  • fDate
    2002
  • Firstpage
    5
  • Lastpage
    13
  • Abstract
    To identify sources of distributed denial-of-service attacks, path traceback mechanisms have been proposed. Traceback mechanisms relying on probabilistic packet marking (PPM) have received most attention, as they are easy to implement and deploy incrementally. We introduce a new concept, namely Groups Of Strongly SImilar Birthdays (GOSSIB), that can be used by to obtain effects similar to a successful birthday attack on PPM schemes. The original and most widely known IP traceback mechanism, compressed edge fragment sampling (CEFS), was developed by Savage et al. (2000). We analyze the effects of an attacker using GOSSIB against CEFS and show that the attacker can seed misinformation much more efficiently than the network is able to contribute real traceback information. Thus, GOSSIB will render PPM effectively useless. It can be expected that GOSSIB has similar effects on other PPM traceback schemes and that standard modifications to the systems will not solve the problem.
  • Keywords
    Internet; security of data; telecommunication security; transport protocols; GOSSIB; Groups Of Strongly Similar Birthdays; IP traceback rumors; Internet; PPM traceback schemes; compressed edge fragment sampling; distributed denial-of-service attacks; misinformation; path traceback mechanisms; probabilistic packet marking; Broadcasting; Computer crime; Computer networks; Information analysis; Laboratories; Proposals; Protection; Sampling methods; Web and internet services; Web server;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Security Applications Conference, 2002. Proceedings. 18th Annual
  • ISSN
    1063-9527
  • Print_ISBN
    0-7695-1828-1
  • Type

    conf

  • DOI
    10.1109/CSAC.2002.1176273
  • Filename
    1176273