Robust Group Key Management with Revocation and Collusion Resistance for SCADA in Smart Grid

1. Robust Group Key Management with Revocation and Collusion Resistance for SCADA in Smart Grid Rong Jiang 2013.07.31

2. Agenda • Introduction • The proposed group key management • Security analysis and performance evaluation • Conclusion

3. SCADA • Supervisory Control And Data Acquisition (SCADA) systems are used to monitor and manage the electric distribution, transmission and generation environments. • the group communication is an essential part • a status scan request • measured value scan request • an emergent shutdown message or a set-the-clock-time message • Requirement: availability

4. SCADA system architecture HMI: Human-Machine Interface MTU: Master Terminal Unit RTU: Remote Terminal Unit

5. Security Model and Design Goal • Security Model • Group confidentiality • Backward secrecy • Forward secrecy • t-collusion-resistant • Design Goal • Availability • Efficiency

6. preliminary knowledge • Session key • DDHC(dual directional hash chain) • bivariate polynomial

7. Description of LiSH • Initialization • Re-keying • Self-healing mechanism • Adding new member nodes • re-initialization mechanism

8. Initialization

9. Initialization For 1 ≤ j ≤ m, each user whose lifetime is from s1 to s2 (1 ≤ s1 < s2 ≤ m) is assigned the set Hi, set Di, two key seeds, buffer length, rekeying period lis the length of key buffer; Trefreshis the rekeying period : random numbers

10. Re-keying the KDC periodically discloses the next secret number C and constructs a self-healing set T and a revocation set R to expel some illegal nodes. :the set of users all revoked in and before sj :a set of irrelevant users :self-healing set The secret in is concealed in this way: :broadcast polynomial

11. Re-keying When a non-revoked node receives the jth session broadcast message , it checks whether the revoked set is changed. If not, it switches the session key from the front of the key queuecalculatesand resets the time of refresh key. Otherwise it will calculate the new session key and update the key buffer. It evaluates revocation polynomial, computes the forward key and backward key, obtains and then recovers

12. After that,can iteratively obtain all before in its legal lifetime by self-healing set

13. Self-healing mechanism • When a sub-MTU breaks down because of attacks or natural disasters, the RTUs can keep on working for at most sessions. • When the timer of refresh key is triggered and the node does not received the re-keying message, it will switch the session key automatically from the front of the key queue and reset the timer of refresh key. • After the broken sub-MTU is repaired or replaced, a legal node with lifetime from to can update all of its session keys in the buffer.

14. Adding new member nodes • When a node (lifetime from to ) tries to join the existing group, it firstly requests the KDC's authentication. After verifying its identification, KDC encrypts the following items via private channel between KDC and the new node and then sends them back to :

15. Determination of self-healing period • We define a utility function f to find the period where SI and RI stand for security index and robustness index. we can find the optimal self-healing period to maximize the utility function.

16. Security Analysis • Theorem 1: LiSH is a session key distribution with privacy and achieves self-healing with time-limited t-revocation capability. • Theorem 2: LiSH achieves t-wise forward and backward secrecy. • Theorem 4: LiSH is resistant to attacks to Dutta [13] and Du [17]’s schemes mentioned in section II.

17. Performance Evaluation

18. Conclusion • In this paper, we have proposed a robust and efficient group key management, named LiSH, to secure SCADA system in smart grid. • The proposed LiSH scheme is characterized by adopting self-healing key to tolerant failures of the sub-MTUs. Security analysis has shown that the proposed LiSH is a collusion-free and self-healing key distribution scheme with t-wise forward and backward security. • In addition, performance evaluation has also demonstrated its efficiency.

19. Thank you!