Randomness Evaluation of PP-1 and PP-2 Block Ciphers Round Keys Generators

Abstract

Round keys in block ciphers are generated from a relatively short (64-, 128-, 256-, and more bits) master key and are used in encryption and decryption process. The statistical quality of round keys impact difficulty of block cipher cryptanalysis. If round keys are independent (not-related) then cryptanalysis need more resources. To evaluate key schedule’s statistical quality we can use NIST 800-22 battery test. PP-1 key schedule with 64 bits block size and 128-bit master key generates 22 64-bits round keys that gives cryptographic material length of 1408 bits. PP-2 with 64-bits block size generates in single run from 128-bits master key only 13 round keys, which give 832-bits sample from single master key. Having such short single samples we can perform only couple of NIST 800-22 tests. To perform all NIST 800-22 tests at least 106 bits length samples are required. In this paper we present results of randomness evaluation including all NIST 800-22 tests for expanded PP-1 and PP-2 round key generators.