1. Denial of Service in Sensor Networks Szymon Olesiak

2. Outline Denial of Service (DoS)- description Underwater Sensor Networks (UWSN) - special features Review of DoS attacks Important DoS attacks in UWSN Conclusions

3. DoS in Sensor Networks Disrupt communication and cooperation between nodes Decrease availability of the whole network Waste precious resources (power)

4. Why Denial of Service so dangerous Cheap Hard to detect Often precede the “real” attack Deadly for sensor nodes

5. Why care about it now? Security integrated at every layer Have to ensure security at design time UWSN at its initial stage Adding security after the protocol is “stable”

6. UWSN constraints Not able to secure their communication medium Low speed of the transmission (1500 m/s) Low data rate (5-20 kb/s) Limited bandwidth (around 30 KHz) High mobility (3-6 km/h) Environment 3D vs. 2D Physical constraints (pressure, corrosion)

7. Types of DoS Physical Layer Jamming Tampering Link Layer Collision Exhaustion

8. Types of DoS (Cont.) Network Layer Neglect and greed Homing Misdirection Black holes Wormhole Transport Layer Flooding Desynchronization

9. DoS in Physical Layer 1) Jamming Constantly sending noise into channel Even if temporary, can be dangerous Workaround: Switch to lower duty cycle Defense: - Spread spectrum - Try to alert other nodes about the attack

10. DoS in Physical Layer 2) Tampering Physically compromising sensor node Destruction or turning into a malicious node Workaround: Complete failure of the node, once compromised Defense: Hide the node

11. DoS in Link Layer 1) Collision Inducing a collision in section of the transmitted packet Defense: No fully-effective found, because of the initial assumption that nodes should cooperate to avoid corruption of others’ packets

12. DoS in Link Layer 2) Exhaustion Induce retransmissions Induce redundant traffic Defense: Authentication

13. DoS in Network Layer 1) Neglect and greed Refuse to forward packets Give priority to its own traffic Defense: Redundant paths

14. Dos in Network Layer 2) Homing Finding a privileged node and compromising it Defense: Encrypt message headers

15. Dos in Network Layer 3) Misdirection Send packets in different direction then intended by the source node Defense: Routing table updates, allowed only by authorized nodes.

16. Dos in Network Layer 4) Black holes Advertising zero-cost links to itself Defense: - Redundant paths - Authorization of updates of routing tables

17. Dos in Network Layer 5) Wormholes Eavesdrop a packet and then release it in remote location in the network Defense: Geographic forwarding

18. DoS in Transport Layer 1) Flooding Flood a node with connection requests Defense: Require authentication to create a connection

19. DoS in Transport Layer 2) Desynchronization Forge message with sequence numbers or flags that cause retransmission Defense: Authenticate all the messages

20. Attacks Strongly Influencing UWSN Jamming - Limited bandwidth - Simple and cheap to perform Collision - The environment itself causes enough errors

21. Attacks Strongly Influencing UWSN Homing - We will need to have some privileged node, with greater computation power Wormhole - The difference in propagation delay between radio waves and sound

22. Conclusions Security has to be considered at the design time of protocols on every layer We will need to have a solution that will attempt to provide defense from multiple DoS attacks The security vs. energy efficiency trade-off

23. References (1) Anthony D. Wood, John A. Stankovic Denial of Service in Sensor Networks Anthony D. Wood, John A. Stankovic A Taxonomy for Denial-of-Service Attacks in Wireless Sensor Networks John Heidemann, Wei Ye, Jack Wills, Affan Syed and Yuan Li Research Challenges and Applications for Underwater Sensor Networking

24. References (2) Ian F. Akyildiz, Dario Pompili, Tommaso Melodia State of the Art in Protocol Research for Underwater Acoustic Sensor Networks Zhong Zhou, Jun-Hong Cui, and Amvrossios Bagtzoglou Scalable Localization with Mobility Prediction for Underwater Sensor Networks John A. Stankovic Research Challenges for Wireless Sensor Networks

25. Thank you!