Intrusion Detection and Advanced Persistent Threats - PowerPoint PPT Presentation

About This Presentation
Title:

Intrusion Detection and Advanced Persistent Threats

Description:

CS 591 Andrew Bates ... 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700 2800 2900 4000 ... – PowerPoint PPT presentation

Number of Views:196
Avg rating:3.0/5.0
Slides: 11
Provided by: and135
Learn more at: http://cs.uccs.edu
Category:

less

Transcript and Presenter's Notes

Title: Intrusion Detection and Advanced Persistent Threats


1
Intrusion Detection and Advanced Persistent
Threats
  • CS 591
  • Andrew Bates
  • University of Colorado at Colorado Springs

2
Introduction
  • What is the Advanced Persistent Threat
  • Pattern Based Intrusion Detection
  • Proposal
  • Conclusion

3
What is APT
  • Combination of many existing known threats not
    just Phishing or Spear Phishing
  • Social Engineering
  • Zero Day Exploits
  • Botnets
  • Whats different? Persistent!
  • Exploits custom built for a given attack
  • Threat or attack can span many months
  • Very carefully crafted
  • Low Volume

4
APT and Intrusion Detection Systems
  • IDS very good at alerting known exploits and
    vulnerabilities
  • IDS also good at identifying Denial of Service
    (DoS) and Distributed DoS (DDoS) attacks
  • APT can be low volume and may not actually
    exploit any known vulnerability
  • Targeted email that coerces victim to download
    and run some software

5
Pattern Based Intrusion Detection
  • Always one step behind
  • Must know of a vulnerability in order to build
    pattern
  • Can have very high false positive rate in large
    organizations
  • Must know what normal behavior is
  • Very high maintenance

6
Pattern Based Intrusion Detection
  • On small networks can have hundreds of alerts in
    short period of time
  • If the relationship between number of hosts and
    number of alerts/false positives is linear

7
Proposal
  • Push IDS as close to the host as possible
  • Use learning algorithms to determine normal
    activity
  • Trigger on anomalous activity
  • Score sessions based on triggers and then perform
    more strenuous tests
  • Pattern matching, traffic analysis, etc.

8
Proposal
  • Leverage VM technology to place inline IDS/IPS
    with host system
  • Funnel data to central collection/correlation
    infrastructure
  • Alert on anomalous activity based on learned
    normal behaviour

9
Conclusion
  • APT is just like any other threat, but may be
    lower volume and more targeted
  • Pattern based IDS not well suited for APT
    detection in an Enterprise
  • Push IDS towards the host, perhaps even on the
    physical hardware
  • Learn normal behavior and trigger further tests
    when abnormal behavior occurs

10
Questions?
Write a Comment
User Comments (0)
About PowerShow.com