边缘计算结构比传统云计算更具复杂性，因此数据完整性问题变得尤为重要。针对目前完整性审计方案中仍然存在的隐私保护能力差、计算存储开销大、系统复杂度高、不支持动态操作等问题，本文提出基于动态变色龙认证树的完整性审计方案。该方案在数据加密阶段，引入无证书公钥密码体制，在用户端生成数据加解密钥，保证了数据在系统传输过程中的完整性、机密性；在数据上传阶段，采用动态变色龙认证树存储结构，保证数据存取过程全动态操作的同时，减少通信计算开销；并利用边缘节点的计算能力进行完整性审计，避免了因第三方审计机构而产生的单点失效问题。在随机预言模型下，基于计算性Diffie-Hellman困难问题（Computationa Diffie-Hellman Problem，CDH）和离散对数困难问题证明了本方案的机密性。实验证明，相较于其他完整性审计方案，本方案的计算效率更优，能节省约50%的计算存储开销。

Edge computing structures are more complex than traditional cloud computing, so data integrity issues become even more important. In order to solve the problems of poor privacy protection ability, high computing and storage overhead, high system complexity, and lack of support for dynamic operation in the current integrity audit schemes, this paper proposes an integrity audit scheme based on dynamic chameleon authentication tree. In the data encryption stage, the scheme introduces a certificateless public key cryptography system to generate a data encryption and decoding key at the user end to ensure the integrity and confidentiality of the data in the process of system transmission; in the data upload stage, the dynamic chameleon authentication tree storage structure is adopted to ensure the full dynamic operation of the data access process and reduce the communication computing overhead; and the computing power of the edge node is used to carry out integrity audit, which avoids the problem of single point failure caused by the third-party audit agency. Under the stochastic oracle model, the confidentiality of the scheme is proved based on the computational DH difficulty problem and the discrete logarithmic difficulty problem. The experiments show that compared with other integrity audit schemes, the proposed scheme has better computing efficiency and can save about 50% of computing and storage overhead.