具有不可预测性的物理混沌随机序列的产生与随机性统计测试分析

Test and Comparison Analysis on Properties of Random Sequences Generated by Unpredictable Physical Chaos

给出了通过几种经典的物理混沌系统,包括Chua’s电路、Lorenz系统、Chen系统,产生混沌随机序列的两种离散化方法.用NIST制定的测试标准对产生的随机序列进行了随机性统计测试分析.分别对不同的物理混沌系统产生的随机序列进行单比特频数测试、游程测试、离散傅里叶变换测试、近似熵测试、累积和测试,并进行了比较分析.实验结果表明:物理混沌序列要比算法混沌序列随机性更强,不同序列之间的随机性差异可以通过NIST测试得到体现.在三种经典电路中,建议选择Chen混沌系统的电路做为混沌随机序列发生器.

Two methods of generating random sequences by several typical physical chaotic systems,including Chua’s circuit,Lorenz system and Chen’s system,were given.And the randomness properties of the random sequences generated by the chaos systems were analyzed by using NIST’s STS randomness test suite.Six tests,which are the monobit test,the runs test,the discrete Fourier transform test,the approximate entropy test,and the cumulative sums test,were conducted to compare the pseudorandom properties of these chaotic sequences.The experiment results show that the randomness properties of the sequences generated by physical chaos are better than those by simulation chaos,and NIST’s STS randomness test suite can distinguish the randomness of different sequences.Among three typical physical chaotic circuits,we recommend the circuit of Chen’s system to generate random sequences.