A hidden Markov Model-based method for virtual machine anomaly detection

Chaochen Shi, Jiangshan Yu

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearch

Abstract

The normal operation of virtual machine is a necessity for supporting cloud service. Motivated by the great desire of automated abmornal operation detection, this paper proposes a Hidden Markov Model-based method to conduct anomaly detection of virtual machine. This model can depict normal outline base of virtual machine operation and detect system outliers through calculating non-match rate. Through verifying the method in a real distributed environment, experiment results indicate that this method has 1.1%–4.9% better detection accuracy compared with two leading benchmarks with a much better efficiency.

Original languageEnglish
Title of host publicationProvable Security - 13th International Conference, ProvSec 2019 Cairns, QLD, Australia, October 1–4, 2019 Proceedings
EditorsRon Steinfeld, Tsz Hon Yuen
Place of PublicationCham Switzerland
PublisherSpringer
Pages372-380
Number of pages9
ISBN (Electronic)9783030319199
ISBN (Print)9783030319182
DOIs
Publication statusPublished - 2019
EventInternational Conference on Provable Security 2019 - Cairns, Australia
Duration: 1 Oct 20194 Oct 2019
Conference number: 13th
https://www.monash.edu/provsec2019

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume11821
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceInternational Conference on Provable Security 2019
Abbreviated titleProvSec 2019
CountryAustralia
CityCairns
Period1/10/194/10/19
Internet address

Keywords

  • Anomaly detection
  • Cloud computing
  • Hidden Markov Model
  • Virtual machine

Cite this