Cryptographic Function Location Based on Binary Entropy

To solve the accurate problem of cryptographic function location for fuzz testing of binary program, a location method based on binary entropy was proposed. Firstly, instruction features of binary program were modeled to locate key code. Then binary entropy of data handled before and after key code was calculated. Finally, the cryptographic function was located according to binary entropy relationship before and after key code. Experimental results show that this method can locate cryptographic functions in binary program effectively and the accurate rate is more than 99%. This method can enhance the ability of fuzz testing.