Wilderness Survival Skills for Multihop Wireless Networks · Effects of Node Misbehavior in Mobile Ad hoc Networks Survival Skills A Short Guide to Survival Mitigating Misbehavior

1

Prof. Dr.-Ing. Matthias HollickSecurity in Mobile Networks and Wireless Mesh Networks

Universidad Carlos III de MadridDepartamento de Ingeniería Telemática

Phone +34.91.624.8794Fax +34.91.624.8749

[email protected] or [email protected]. de la Universidad, 30

E-28912 Leganés (Madrid), Spainhttp://www.it.uc3m.es/

02-Jul-2009

Wilderness Survival Skills for Multihop Wireless Networks

tubs.CITYTechnische Universität Carolo-Wilhelmina zu Braunschweig

02-Jul-2009

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 2 –

The Wilderness

[Sources: wikimedia commons]

2

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 3 –

Predators

Wild animalsLionsTigersCrocodilesWolves…

[Sources: www.sxc.hu]

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 4 –

Other Threats

Wild animals“Smaller animals actually present more of a threat to the survivor than large animals.” (Source www.wilderness-survivor.net)Insects such as mosquitosArachnids…

[Sources: www.sxc.hu]

3

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 5 –

Outline

The WildernessThreats for Multihop Wireless Networks in the Wild Effects of Node Misbehaviorin Mobile Ad hoc Networks

Survival SkillsA Short Guide to SurvivalMitigating Misbehavior using Geographically Secure Routing

[Sources: www.sxc.hu]

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 6 –

Threats & Attacks against Multihop Wireless Networks

Security goalsAuthenticity, confidentiality, integrity, non-repudiation, availability

Passive and active attacksWireless channel simplifies attacks

Attacks take place on all layers (examples)Jamming on PHY layer, physical securityof small devices typically weakSelfishness on MAC layerMalicious behavior during routing and/or forwarding on NET layerTRANSPORT layer of the Internet has notbeen designed for multihop wireless networks, is weak even without attackAttacks are scenario/application dependent

Layer 1 – PHY

Layer 2 – LINK

Layer 3 – NET

Layer 4 – TRANS

Layer 5 – APP

4

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 7 –

Wireless Multihop Networks in the Wild

Heterogeneous multihop wireless networks, different application needs, etc.

Heart Rate

Building Automation

Location

Temperature

Health Monitoring

Tracking of Goods

Network Data & ServicesData & Things

[Sources: sunsportworld.com, gumstix.com, nokia.com, apple.com, motive.com, xbow.com, meshcube.org, meraki.com, tropos.com]

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 8 –

Heterogeneity of Scenarios, Network Characteristics, Threats, and Security Goals

Industrial WSNHarsh environmental conditionsPhysical protection vs. intruders possibleData integrity to be protectedAvailability to be protectedQoS provisioning under attack

Participatory (Ambient) SensingOpen network susceptible to attackIntegrity & privacy (however protecting privacy might challenge integrity)Confidentiality for some data

MANET/WMNVANET: integrity of warning messages, fast exclusion of errant devices, location privacyMANET: availability of service, protect cooperation of nodes, “classical” network security depending on applicationProvider WMN: closed network (nodes authenticated), QoS provisioning and availability to be protectedCommunity WMN: open network, distributed mechanisms, cooperation necessary

5

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 9 –

Effects of Node Misbehavior in Multihop Wireless Networks

Scenario

ChallengeQualify and quantify the effects of node misbehavior on the overall performance of the routing system

XX

Vulnerability of multihop route

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 10 –

Experimental Analysis: Simulation Study

Performance Evaluation by Extensive Simulation StudyStarting from a well performing MANET in normal operationStudied variants of AODV routingStudied various degrees of mobilityStudied scale of the networkStudied type/degree of misbehavior

Malicious Nodes - Black HolesActively attract routes by injectingfalse routing information Remove packets from network

Selfish Nodes Optimize their own gain, neglect welfare of other nodesDiscard other node‘s packets

6

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 12 –

Simulation Study: Selected Results

Setup250 nodes, low node mobility (1-2m/s), AODV with expanding ring search

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 13 –

Observations for Maliciousness/Black Holes

ResultsSuccessful communication is possible only in close proximityPacket loss is extremely high, even for few black holesPacket loss further increases with node mobility

49%78%

i.e. 10 out of 250 nodes

Increase in malicious nodes

7

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 14 –

Wireless Survival Skills

[Image Source “www.sxc.hu“]

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 15 –

Survival Actions

[Sources: www.sxc.hu, "Survival FM 3-05.70" US Army Handbook]

8

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 16 –

Exploiting Space & Time: Complementing Contemporary Security Solutions

Challenge:Inherent vulnerability of multihop wireless networksMitigating misbehavior in networks without well-defined boundaries

(I) Intrusion PreventionMostly secure routing protocols (leveraging cryptography)

How to prevent intrusions on other layers than network layerE.g. SAODV and Ariadne as secure versions of AODV and DSR

New attack vectors identified by Acs et al. and Hu et al.

(II) Intrusion ResponseE.g. Watchdog & Pathrater by Marti et al.E.g. CONFIDANT by Buchegger et al.E.g. CORE by Michiardi et al.E.g. OCEAN by Bansal et al.

Today, intrusion response is based mostly on addresses, which can easily be attacked in open networks

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 17 –

Exploiting Space: GeoSecGeographically Secure Routing

Approach: Exclude Misbehaving Nodes Based on LocationSetting up ‘quarantine zones’ void of communicationResults in cross-layer and attack-independent defense

Implementation(1) Establish quarantine zones

Interrupt affected routesExclude attacker from network

(2) Establish new routesBy restricting broadcastEvading quarantine zones

(3) Maintain quarantine zonesReset zones periodicallyTracking of attacker not necessary

9

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 18 –

Address-based IRS GeoSec

Performance Evaluation GeoSec vs.Address-based Intrusion Response

MetricPacket drop rates

Attacker vs. IRS

Setup1000 nodes, low node mobility, AODV with expanding ring search

No IRS

Address-based IRS GeoSec

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 19 –

Exploiting Time: GeoSec+Geographical & Delay-tolerant Secure Routing

Approach: Buffer Packets During Attack and Retransmit LaterTransparent operation, i.e. sender is informed about IRS event Non-transparent operation, i.e. “silent” or “localized” operation of IRS

Late buffering strategy, i.e. packets are buffered as IDS detects misbehaviorHowever, packet loss during time IDS needs to detect misbehavior

Early buffering strategy, i.e. packets also buffered in detection interval

Transparent Non-transparent

10

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 20 –

Summary of Results for GeoSec+Avg. Packet Delivery Ratio

Tran

spar

ent,

Late

Buf

ferin

g

Non

-tran

spar

ent,

Late

Buf

ferin

g

Tran

spar

ent,

Early

buf

ferin

g

Non

-tran

spar

ent,

Early

Buf

ferin

g

In studied scenario up to~80% delivery ratio

Performance independentof number of

retransmissions

In studied scenario up to~75% delivery ratio

Performance dependentof number of

retransmissions

In studied scenario up to~95% delivery ratio

Performance dependentof number and scheduling

of retransmissions

In studied scenario up to~90% delivery ratio

Performance independentof number of retransmissions

(but congestion observed)

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 21 –

Summary of Results for GeoSec+Avg. End-to-End Delay

Tran

spar

ent,

Late

Buf

ferin

g

Non

-tran

spar

ent,

Late

Buf

ferin

g

Tran

spar

ent,

Early

buf

ferin

g

Non

-tran

spar

ent,

Early

Buf

ferin

g

11

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 22 –

Summary & Conclusion

The WildernessThreats for Multihop Wireless Networks in the Wild Effects of Node Misbehaviorin Mobile Ad hoc Networks

Selected Survival SkillsA Short Guide to SurvivalMitigating Misbehavior using Geographically Secure Routing

[Sources: www.sxc.hu]

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 23 –

Thank You Very Much for Your Attention

12

Making Multihop Wireless Networks Secure and Quality of Service Aware

Matthias Hollick, tubs.CITY Workshop, 02-Jul-2009– 24 –

Acknowledgements, Copyright Notice

Copyright NoticeThis document has been distributed by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.

Acknowledgements To André König, Technische Universität Darmstadt, for his work in the area of innovative security mechanisms for Mobile Ad hoc Networks