Edge computing technology has led to the emergence of a variety of emerging businesses, but it also introduces more uncertain security threats and demand for differentiated security services. This paper focuses on the security protection technology of the English education network and proposes a statistically based incremental network traffic feature extraction strategy. The strategy aims to extract and tailor the intruding network features, ensuring that the feature vectors accurately represent the real network behaviors while maintaining simplicity. Then, from the perspective of endogenous security, based on the idea of dynamic heterogeneous redundancy, the virtualized resource security construction using the edge computing network’s architectural characteristics is utilized. Aiming at the security needs of the English education network, an endogenous edge computing network security scheme based on network function virtualization is proposed, and a dynamic heterogeneous redundancy DHR security model is established. The model has the ability to provide anomaly information to the scheduling module, which in turn lowers the ranking of executives under attack in the pending set. This reduces the likelihood of penetration attacks targeting system vulnerabilities, thereby enhancing the system’s security reliability. The failure rate of the method in this paper is consistently maintained below 0.52 overall, highlighting excellent security and realizing the design of a secure access method for English education networks.