Subspace Techniques for Obfuscating Matrix-based Algorithms (STOMA)

Scientific Systems Co., Inc. (SSCI) and its academic partner, the University of Arizona, have devised a comprehensive and effective methodology for offering software anti-tamper protection via Algorithm Obfuscation. A primary goal going into Phase I was to apply our techniques in the Kalman Filter setting; this example was completed successfully. Thus, we now have a collection of obfuscation techniques that can be used to render an algorithm unrecognizable. In Phase II we will assemble these powerful obfuscation techniques and construct what we have called an "Obfuscating Compiler." Like a more conventional compiler, it will accept algorithms that have been expressed in a "High Level Language" and produce code in a more primitive form. This code is designed so as to make the underlying algorithm completely obscured while remaining functionally equivalent to the original. In Phase III, we will expand this prototype into a robust tool capable of widespread use. This will include, but may not be limited to combining matrix-level algorithm obfuscation with more traditional software obfuscation techniques. Note that we have developed a "meta-algorithm" that can be applied to large classes of matrix-based algorithms, and do not require detailed knowledge of the particular algorithm for which obfuscation is desired.