Working on Security in Wireless Sensor Networks - Horizons

Current focus: Self-adaptive software & Security. Application domain: ... Wireless Sensor Network at Camalie Vineyards - Mt. Veeder, Napa Valley, California. 19 ...
5MB taille 1 téléchargements 340 vues
Visiting researcher at NII, Honiden Lab (JSPS funding) Member of the GRACE center of excellence http://grace-center.jp/en/ Ph.D. from Paris 6 University & Sokendai (2007) Research Interests: Engineering of dependable systems Current focus: Self-adaptive software & Security Application domain: Wireless Sensor and Ad Hoc Networks 2

Email: [email protected] Homepage: http://honiden-lab.ex.nii.ac.jp/~eric Ofcial NII page inside http://www.nii.ac.jp Organizer of: The APSLA research track at ACM SAC; Tutorials on Agents and WSN; The future AOC conference on agent computing. 3

Wireless Sensor Networks Security Decentralized Security

6

7

8

© Sentilla™, photo from James Gosling’s blog: http://blogs.sun.com/jag/date/20071016

9

© http://www.schneiderism.com/

10

11

12

13

Limited Performance

14

Limited Resources

Limited Connectivity

15

16

17

© Culler, UC Berkeley

18

© Wireless Sensor Network at Camalie Vineyards - Mt. Veeder, Napa Valley, California.

19

© CodeBlue Project, Harvard

20

© CitySense Project, Harvard

21

Health-care Environment Surveillance Home automation Smart ofces Military support & intelligence

Ambient Intelligence

22

Ubiquitous Computing

Internet

24

Internet

25

Internet

26

Internet

27

? 28

(Fire)wall?

29 (c) Wireless Sensor Network at Camalie Vineyards - Mt. Veeder, Napa Valley, California.

Usual concerns Stringent conditions !

Energy

!

Computing power

!

Communication

!

Scale

!

“Wild” environment 30

? 31

Cryptography Network Topology Routing strategy

32

33

Communication: Cryptography Computation: ECC Energy concern: ECC, Cross-layering Scale: Secure routing Wild environment Node compromise: Little work Key management: Little dynamics 34

Communication: Cryptography Computation: ECC Energy concern: ECC, Cross-layering Scale: Secure routing Wild environment Node compromise: Little work Key management: Little dynamics 35

Detection & connement of node compromise Attacking WSN Key management Secure routing

37

Work from Yuichi Sei Detection of illegitimate data for connement See Yuichi’s work: http://honiden-lab.ex.nii.ac.jp/~sei

38

Properties Faster than related work K-robustness Limitations Fixed infrastructure (no mobility) Overhead data still high 39

Work with Jaewoo Lee, Seoul National University Goal: “Aggressive” attacks of WSN for data collection

Use of common and available security modules Method: Use known weaknesses, from the user doc ! 40

41

!"#$%&'($% Sink

Field Sensor

Attacker

)*+,(-%&'(

%$42

Field Sensor

Condition of attacks: •! Broadcast communications •! Message payload “too long” •! Weak link due to environment

Eavesdropping

Attacker

43

How to dynamically manage keys? Preliminary stage Research issues: Key deployment Re-keying Mitigating key capture Routing 44

Approach: Physical and bio-inspired mechanisms Exploration: Robustness of physical structures Key as digital pheromone “It would be nice if organisms were able to...” 45

Goal (unachieved) Recursive security by decentralized mechanisms

Property “If an attacker captures a node, the attacker must also capture the neighbor to access the network” 46

Current (naive) attempt The “chain-mail” robustness

47

*++$%&'$(&!)($%& ,)-$&'$(.&,/012$3& !"#$%&'$(&!)($%&

48

Advantage of double chain mails: Very (too) simple Limitations: Expensive Attempt with digital pheromones (current)

49

End-to-end security scheme Target features Rely on dynamic key infrastructure Adaptive routing Energy efciency by cross-layering Started this month with Neeraj from NJIT 50

Current work on security in wireless sensor networks Target issues !

Security via decentralized approaches

!

Preserving qualities (energy efciency, adaptiveness)

Theoretical and practical results in Summer! 51

Cyrille Artho & Pierre-Loïc Garoche. Accurate centralization for applying model checking on networked applications, ASE 2006, pp. 177–188. Fok, C.L. et al. Rapid development and exible deployment of adaptive wireless sensor network applications. ICDCS 2005, pp. 653–662. Ghosh, S. K.: On Optimality of key predistribution schemes for distributed sensor networks. ESAS 2006, pp. 121-135. Keromytis, A.D. et al.: A Holistic Approach to Service Survivability. Workshop on Survivable and Self-Regenerative Systems (SSRS),(2003) 11– 22. C. Heath and B. Blakley: Security design patterns. The Open Group, 2004. Hui, J.W et al.: The dynamic behavior of a data dissemination protocol for network programming at scale, ACM SenSys 2004, pp. 81–94. Karlof, C. et al. Tinysec: a link layer security architecture for wireless sensor networks, SenSys ’04, 2004, pp. 162–175.

53

Luk, M et al. Minisec: a secure sensor network communication architecture, IPSN'07, pp. 479– 488. Girão, J. et al. Tinypeds: Tiny persistent encrypted data storage in asynchronous wireless sensor networks, Ad Hoc Networks 5 (2007), no. 7, 1073–1089. Perrig, A. et al. Spins: security protocols for sensor netowrks., MOBICOM, 2001, pp. 189–199. Luk, M. et al.: MiniSec: a secure sensor network communication architecture. IPSN 2007, pp. 479-488. Tanveer Zia, Albert Y. Zomaya: A Secure TripleKey Management Scheme for Wireless Sensor Networks. INFOCOM 2006. Wang, D et al.: Self-Protection for Wireless Sensor Networks. ICDCS, 2006. Zhu, S. et al. Leap: efcient security mechanisms for large-scale distributed sensor networks, CCS ’03, 2003, pp. 62–72.