Research and Application of APN Network Security Access Management Technology Based on Zero Trust Model for Power Grid IOT Terminal

With the development of smart grid technology, the security of APN network of power grid IOT terminal is facing greater challenges. If outsiders steal accounts or insiders engage in illegal operations, the network security under traditional border protection will be threatened. To address these issues, this paper proposes a secure access management method based on zero trust model ABAC-IntroVAE. This method combines rule matching and behavior analysis technology based on variational self-encoder to judge whether the access request of APN network of power grid IOT terminal meets the security policy. First of all, the requests that do not meet the requirements are filtered out through rule matching. Then, the normal request samples are trained by IntroVAe model to realize dynamic access decision. Because the collection of normal samples is relatively easy and stable, the challenge of frequently changing attack samples is avoided. The experimental results show that ABAC-IntroVAE model performs well in the dynamic access control of APN network of power grid IOT terminal, and its anomaly detection accuracy reaches 97.2%.