Understanding Challenges and their impact on network resilience - - PowerPoint PPT Presentation

understanding challenges and their impact on network
SMART_READER_LITE
LIVE PREVIEW

Understanding Challenges and their impact on network resilience - - PowerPoint PPT Presentation

Understanding Challenges and their impact on network resilience Marcus Schller Paul Smith NEC Europe Laboratories Lancaster University Datum Motivation and Overview Goal: Determine high impact challenges Add defensive and


slide-1
SLIDE 1

Datum

Understanding Challenges and their impact

  • n network resilience

Marcus Schöller Paul Smith NEC Europe Laboratories Lancaster University

slide-2
SLIDE 2
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

Motivation and Overview

  • Goal: Determine high impact challenges
  • Add defensive and remediation mechanisms
  • Meet cost constraints
  • Challenges
  • Component Faults
  • Hardware destruction
  • Challenges to the communication

environment

  • Human Mistakes
  • Malicious attacks from intelligent

adversaries

  • Unusual but Legitimate Demand for Service
  • Failure of a Provider Service
  • Case Study to exemplify

2

slide-3
SLIDE 3
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

The community mesh network in Wray

3

  • Single back-haul link
  • Mesh gateway at school
  • Upstream provider: Uni Lancaster
  • 802.11b mesh
  • Currently hard coded routes
  • Network services
  • DNS, DHCP, NAT and FW
  • Secondary DNS at Uni Lancaster
slide-4
SLIDE 4
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

Identify critical assets and their value

  • End Users
  • Internet Connectivity
  • Safety: of less computer literate members,

personal identity, reputation, protection of children

  • Security: documents, photos, …
  • Operator
  • Physical Infrastructure
  • Security of services incl. charging
  • Connectivity: SLAs, QoS, …
  • Researchers
  • Monitoring of network
  • Measurements on network

4

  • Value
  • Loss depends on state
  • f impairment
  • Three-level state space
  • Acceptable
  • Impaired
  • Unacceptable
  • Assign impact value

[0:1]

slide-5
SLIDE 5
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

System, challenges, and faults

  • 1. System understanding
  • Software engineering
  • 2. Identify potential challenges to the

system

  • Learning from past events
  • Advisories on past resilience

incidence

  • System analysis
  • Fault Tree
  • Event and attack tree
  • 3. System faults
  • Connecting challenges and system

components

  • Taxonomy

5

  • Challenges in Wray
  • Milk Truck
  • Extreme Weather
  • Device mis-configuration
  • Broken hardware
  • Mis-use of resources
  • Malware
slide-6
SLIDE 6
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

Challenge Sheets for documentation

Challenge Name Frequency Jammer Classification Category Malicious Attack Scenario Wireless Communication Characteristics Description The frequency used for communication is jammed by a) constant, b) periodic, c) interactive, d) arbitrary transmissions of the attacker. Scope MAC Potential Impact Communication among nodes in the vicinity is prevented or severely degraded Details Parameters Duration of interference, period of jamming signal, output signal strength Symptoms MAC layer protocol violation, disrupted link frames, reduced link bandwidth

6

slide-7
SLIDE 7
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

Determine a measure of exposure

  • 1. Likelihood of challenge

degrading a service

  • Three-level state space
  • 2. Impact costs
  • Derived from asset values
  • 3. Calculate exposure
  • Impact Cost * likelihood
  • Order challenges based on

exposure

  • 4. Adding defensive and

remediation mechanisms

  • Cost constrainted

7

slide-8
SLIDE 8
  • M. Schöller, P. Smith

Dagstuhl Seminar, 14-17 April 2009

Summary and Open Questions

  • Reliable measures for challenge occurrence probabilities
  • Analyzing monitoring data
  • Online documentation of adaptation cycles
  • Machine learning on dataset
  • Information sharing to system engineers is difficult
  • Quantifying the impact of a challenge
  • Understand challenges parameters
  • Build challenge simulation models to run against system models incl.

fault models

  • Advise to system monitoring engineering
  • Monitoring of network metrics based on symptoms of most probable

challenges

8

slide-9
SLIDE 9

Datum

Thank you for your attention

9