面向用户隐私保护的用电数据压缩加密方法

Power Data Compression and Encryption Method for User Privacy Protection

用户侧电压和电流的录波数据包含大量的负荷和用电信息,伴随着非侵入式用电负荷识别方法的兴起,这些隐私信息可能以监听的形式被非法获取。而由于采集终端资源受限且数据量庞大,同时现有加密算法资源需求高、密钥管理复杂,难以有效部署,给用户隐私安全留下了巨大隐患。该文立足于用户侧用电信息实时安全监测的需求,提出一种基于压缩感知框架的数据压缩同步加密算法。算法基于压缩感知架构,通过三元Logistic-Tent混沌系统建立联合随机模型,生成混沌测量矩阵实现对电压电流录波数据的边压缩边加密,有效解决了由于高频采样造成的数据重载问题,同时保证了用电信息的机密性,实现隐私保护。并且在压缩采样端和重构端部署的联合随机模型可实现同步运算,避免了密钥的传输,有效降低了智能终端的密钥管理成本、通信成本和密钥泄露可能性。另一方面,构造基于伯努利矩阵的动态S盒作为联合随机模型的核心,保证了联合随机模型关键参数的不可观测,提升了算法的抗差分分析安全性。最后,采用PLAID用电公开数据集,测试了所提算法的安全性、可行性、密钥敏感性、加密效率和抗非侵入式分析性能。实验结果表明,算法具有高数据压缩比,有效减少用电信息传输数据量,加密效率较RSA算法提升了82%,并能够抵抗非法负荷辨识,破解成功率仅为3.6%,有效提升了用电信息的机密性。

The recorded data of user-side voltage and current contain a large amount of load and electricity information.With the rise of non-invasive electricity load identification method,these privacy information may be illegally obtained in the form of monitoring.However,due to the limited collection terminal resources and large amount of data,the existing encryption algorithm has high resource requirements and complex key management,which is difficult to effectively deploy,leaving a huge hidden danger to user privacy security.Based on the demand of real-time security monitoring of user-side electricity consumption information,this paper proposed a data compression synchronization encryption algorithm based on the compressed sensing framework.Based on the compressed sensing architecture,the joint random model was established by ternary Logistic-tent chaotic system,and the chaotic measurement matrix was generated to realize the compression and encryption of voltage and current recording data,which effectively solves the problem of data overload caused by high frequency sampling,and ensures the confidentiality of electricity information and realizes privacy protection.The joint random model deployed in the compressed sampling end and the reconstruction end can realize synchronization operation,avoid the transmission of key,and effectively reduce the key management cost,communication cost and key leakage possibility of intelligent terminal.On the other hand,the dynamic S-box based on Bernoulli matrix was constructed as the core of the joint stochastic model,which ensures the unobservability of the key parameters of the joint stochastic model and improves the anti-differential analysis security of the algorithm.Finally,this paper used the PLAID electricity public data set to test the security,feasibility,key sensitivity,encryption efficiency and non-invasive analysis performance of the proposed algorithm.The experimental results show that the algorithm has high data compression ratio and effectively reduces the amount of data transmitted by electricity information.The encryption efficiency is 82%higher than that of RSA algorithm,and can resist illegal load identification.The success rate of cracking is only 3.6%,which effectively improves the confidentiality of electricity information.