Using Wireless Channel Characterizations for Secure Secret Key Generation

This year the ISSNET workshop took place in Toronto (26-29 April). I presented a poster and gave a talk. The poster and slides are attached to the bottom of this post.

Using Wireless Channel Characterizations for Secure Secret Key Generation

Masoud Ghoreishi Madiseh, Stephen W. Neville, Micheal L. McGuire
Department of Electrical and Computer Engineering
University of Victoria
Victoria, B.C. V8W 3P6

Wireless key generation solutions have been widely proposed to address the need for secure communication in domains where existing solutions such as PKI can be problematic, (i.e., wireless keyboards, RFID tags, on the first contact to an access point, in wireless sensor networks, etc. ) [1], [2]. Several wireless channel characterization-based key generation solutions have been reported in the literature [3], [4]. The critical aspect of real-world security solutions based on these techniques not addressed in the existing literature is the quantification of the security that they produce. The existing works have only assessed the security of the generated secret key in the context of eavesdroppers located far from the transmitting antennas with poor knowledge of the propagation environment. This leads to the core problem that two parties, nominally Alice and Bob, may engage in wireless channelcharacterization-based key generation under the belief that their resulting key is secure, when in reality this key is partially or completely known to an eavesdropper, Eve. It should be noted that the intention of key generation is not to solve the authentication problem, instead, if properly constructed, it provides a mechanism by which Alice and Bob can engage in an authentication process over a secured channel which requires zero pre-shared secret information to construct, (i.e., once the secure channel is in place Alice and Bob can engage in a standard challenge-response process for authentication).

This work, using experimental measurements, assesses the security of UWB key generation against capable adversaries who: a) have surrounded one of the parties engaged in key generation with their own eavesdropping antennas, b) have a reasonable understanding of communications and signal processing theory, to the extent where they are able to utilize known optimal estimation techniques to deduce their measurable key bits, and c) are idealized, such that they can perfectly synchronize their measurements with Alice and Bob’s key generation process.

This work provides experimental validation of the secret key rate that is pragmatically achievable in real-world implementations of radio channel characterization-based key generation solutions. This is done by performing rigorous signal processing analysis of a significant set of measured UWB channel characterizations [5]–[7]. It is shown that significant care must be taken if the generated key bits are: a) not to be deducible or measurable by capable eavesdroppers who may exist within the environment, and b) to be sufficiently numerous to result

in a non-trivially searchable key space. Moreover, the work shows that critical interdependencies exist between Alice and Bob’s achievable secret key rate and a number of design issues within the key generation systems, (e.g., error correction code power, communication’s channel bandwidth, Doppler frequency, line-of-sight (LOS) versus non-line-of-sight (NLOS) channels, etc. ). The conclusion is that although secret key generation is pragmatically achievable, ensuring the secrecy of the resulting key is non-trivial unless sufficient care is taken in the design of the key generation process.

This work addresses the fundamental issue: ”What can a capable eavesdropper, Eve, deduce about the generated key?” The previous analyses of this question have: a) not allowed Eve access to multiple antennae that surround Alice (or Bob), and b) not allowed Eve to make use of the appropriate optimal signal prediction methodologies. This work addresses this deficiency by analyzing the secrecy of an existing UWB key generation approach. Through this approach: a) empirical bounds on the achievable secure secret key rate are developed, b) the importance of ensuring that key bits only come form NLOS channels is demonstrated, c) the criticality of using sufficiently weak error correction codes is highlighted, and d) the impacts of channel bandwidth and Doppler frequency on the achievable secret key rate are detailed.

REFERENCES

[1] “Report of fraudulently issued certificates,” March 2011. [Online]. Available: http://comodo.com/Comodo-Fraud-Incident-2011-03-23.html
[2] Microsoft Inc. (2001, March) Verisign digital certificates spoofing hazard. Windows Security Update. [Online]. Available: http://www.microsoft.com/downloads/details.aspx?displaylang=en\ &FamilyID=43fd979a-03c1-4008-b38d-70e9bcd67454
[3] C. Ye, S. Mathur, A. Reznik, Y. Shah, W. Trappe, and N. Mandayam, “Information-theoretically secret key generation for fading wireless channels,” Information Forensics and Security, IEEE Transactions on, vol. 5, no. 2, pp. 240 –254, jun. 2010.
[4] N. Patwari, J. Croft, S. Jana, and S. Kasera, “High rate uncorrelated bit extraction for shared secret key generation from channel measurements,” Mobile Computing, IEEE Transactions on: Accepted for future publication, vol. V, no. forthcomming, p. PP, 2009.
[5] M. Ghoreishi Madiseh, M. McGuire, S. Neville, L. Cai, and M. Horie, “Secret key generation and agreement in uwb communication channels,” Global Telecommunications Conference, 2008. IEEE GLOBECOM 2008. IEEE, pp. 1–5, 30 2008-Dec. 4 2008.
[6] M. Ghoreishi Madiseh, S. He, M. McGuire, S. Neville, and S. Dong, “Verification of secret key generation from uwb channel observations,” International Conference on Communications, 2009. IEEE ICC 2009. IEEE, pp. 1–5, June 2009.
[7] M. G. Madiseh, S. W. Neville, and M. L. McGuire, “Time correlation analysis of secret key generation via uwb channels,” in GLOBECOM 2010, 2010 IEEE Global Telecommunications Conference, Dec. 2010, pp. 1 –6.
--------------------------------------------------
Poster    Slides