流密码应用广泛，对密钥流进行随机性测试在信息安全领域有重要作用。有别于对选定长度的随机序列利用NIST统计包进行随机性检测，本文针对HC256，RC4密钥流多重分段利用NIST提供的测试方法进行系统随机性检测，并对测量结果进行频率分布统计，形成二维可视化特征分布。选定的测试数据显示，该类特征分布图对HC256和RC4生成的密钥流区分明确。
Stream cipher was widely used and the random number tests of the key stream play a key role in information security. Unlike usual NIST random number test for the random sequence with selected length, this article was focused on splitting the random sequence to do the NIST random number test. Then, the frequency distribution was measured through using the test results. Finally, two-dimensional feature map was generated. The selected test data show that such feature maps distinguish the key streams generated by HC256 and RC4 through the feature maps explicitly.