Abstract: Aiming at the problem that traditional encryption methods for  sensitive information in  communication network terminals often failed to balance the high sensitivity of signals and robustness in complex environments, making encrypted information vulnerable to cracking or tampering, we proposed a chaotic encryption solution based on parallel interference cancellation technology and evolutionary hypernetwork model. Firstly, by constructing a weight vector matrix with minimum mean square error, the estimation and ranking of interference signals from each user could be achieved, and interference could be subtracted from them to improve the quality of the target signal. Secondly, an evolutionary hypernetwork model was introduced to determine the sensitivity of information samples by using hyperedge association rules, construct a decision classification model, and achieve mining of sensitive data. Finally, by using Lyapunov exponent to measure the sensitivity of chaotic sequences, combined with feedforward 
stream cipher mechanism and scrambling processing, a high security sensitive information key stream was generated to ensure high sensitivity and robustness in complex environments during the encryption process, thereby improving the security and anti cracking ability of encrypted information. The experimental results show that the proposed method can improve the anti-interference ability of encryption methods, and enhance the transmission efficiency of information while ensuring information security.

Key words: communication network security, sensitive information, parallel interference cancellation, chaotic motion, chaotic mapping, Lyapunov exponent