|
This research was funded in part by an HP Research Gift (2004) and NSF Award 0830576 (2008-2011).
Some of these publications [2,3] are also listed as part of my applied information theory research. I have worked on two problems in cryptanalysis. The first was an attempt to view cryptanalysis as communication over a noisy channel in the manner of Filiol. A consequence of this view, not investigated by Filiol, is that related keys correspond to the use of error-correcting codes. This enables the application of classical information-theoretic results to upper bound the complexity of an attack in number of queries per bit of entropy [2]. Master's student Darakhshan Mir (now a doctoral student at Rutgers) carried out related key attacks on DES to demonstrate that the model is realistic and that the use of related keys does indeed improve communication efficiency [3]. Undergraduate Jacob Alperin-Sheriff (now a doctoral student at Georgia Tech.) investigated the relationship between distance properties of codes and communication efficiency aspects of linear cryptanalysis; this manuscript is unpublished. The second problem I have worked on has been driven largely by doctoral student Kerry McKay (now graduated and with the Cryptographic Technology Group at NIST). We derived a statistical distinguisher for ARX (addition-rotation-xor) block ciphers, and used it to cryptanalyze a reduced-round version of Skein, a finalist in the SHA-3 competition for the next secure hash standard [1]. The strong diffusion properties of the full version of Skein seem to make it quite resilient to our attack. |
|