Ensemble Empirical Mode Decomposition with Approximate Entropy: A Novel Detection Model for QUIC Anomaly Traffic

With the rise of Web3.0 and the continuous development of blockchain technology, traditional transmission protocols such as TCP and UDP are gradually unable to meet the growing demand for network applications. Quick UDP Internet Connection (QUIC) stands out with its multiple features such as low latency and support for multiplexing. However, new network attack methods continue to emerge, and how to accurately identify abnormal network traffic has become an extremely important challenge. Because QUIC traffic has the self-similarity, this research proposes an abnormal traffic detection model based on Ensemble Empirical Mode Decomposition (EEMD) and Approximate Entropy (ApEn). By adaptively decomposing traffic, mining abnormal traffic features, and adding ApEn method to reduce component complexit. Finally, it is determined whether the QUIC transmission network is under attack by comparing the changes in Hurst values before and after the attack. This paper uses the NS-3 platform to implement and evaluate this solution. Experiments have verified that the model proposed in this research can better solve the problems of traditional methods in terms of poor adaptability and low efficiency, and has a high accuracy rate in detecting abnormal traffic, making it suitable for network traffic attack detection.