Block cipher identification method based on deep learning MLP algorithm

The invention provides a block cipher identification method based on a deep learning MLP algorithm. The method comprises the following steps of: 1, acquiring a ciphertext file set F of a known cryptosystem, and extracting ciphertext characteristics of each ciphertext file in the ciphertext file set F; step 2, taking the ciphertext feature of each ciphertext file and the cryptographic algorithm adopted by the ciphertext file as a training sample to obtain a training sample set T; 3, taking the training sample set T as training data, and training by using a deep learning MLP algorithm to obtain a block cipher recognition model; and 4, extracting ciphertext features of the to-be-recognized ciphertext file CT, and taking the extracted ciphertext features as input of the trained block cipher recognition model to obtain a cipher algorithm adopted by the to-be-recognized ciphertext file CT. 本发明提供一种基于深度学习MLP算法的分组密码识别方法。该方法包括：步骤1：采集已知密码体制的密文文件集F，并提取所述密文文件集F中各密文文件的密文特征；步骤2：将每个密文文件的密文特征及该密文文件所采用的密码算法作为一个训练样本，从而得到训练样本集T；