PRIME: Plasticity-Robust Incremental Model for Encrypted Traffic Classification in Dynamic Network Environments

2025-08-03

AI Summary

Existing encrypted traffic classification (ETC) methods struggle with continual incremental learning in dynamic networks, suffering from progressive plasticity degradation and performance collapse as tasks accumulate. Method: We propose PRIME, a novel framework that jointly monitors parameter effective rank and neuron activation states to dynamically expand model capacity—enabling lightweight, elastic scalability. PRIME integrates content-agnostic packet-sequence modeling, continual learning mechanisms, and plasticity-driven parameter adaptation. Contribution/Results: Evaluated across multiple datasets and class-incremental scenarios, PRIME achieves substantial gains over state-of-the-art methods—improving accuracy by 3.2–7.8% while increasing parameters by less than 5%. It also demonstrates superior long-term stability, establishing a scalable and sustainable learning paradigm for evolving encrypted traffic analysis.

Technology Category

• Machine Learning: Privacy
• Search and Optimization: Learning to Search
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• Security and Privacy: Large-scale security measurements
• Web Mining and Content Analysis: Models for Web evolution
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

With the continuous development of network environments and technologies, ensuring cyber security and governance is increasingly challenging. Network traffic classification(ETC) can analyzes attributes such as application categories and malicious intent, supporting network management services like QoS optimization, intrusion detection, and targeted billing. As the prevalence of traffic encryption increases, deep learning models are relied upon for content-agnostic analysis of packet sequences. However, the emergence of new services and attack variants often leads to incremental tasks for ETC models. To ensure model effectiveness, incremental learning techniques are essential; however, recent studies indicate that neural networks experience declining plasticity as tasks increase. We identified plasticity issues in existing incremental learning methods across diverse traffic samples and proposed the PRIME framework. By observing the effective rank of model parameters and the proportion of inactive neurons, the PRIME architecture can appropriately increase the parameter scale when the model's plasticity deteriorates. Experiments show that in multiple encrypted traffic datasets and different category increment scenarios, the PRIME architecture performs significantly better than other incremental learning algorithms with minimal increase in parameter scale.

Problem

• Addresses declining plasticity in neural networks for encrypted traffic classification
• Proposes PRIME framework to adaptively scale parameters for improved incremental learning
• Enhances model performance across dynamic encrypted traffic datasets and scenarios

Innovation

• Plasticity-robust incremental learning for traffic classification
• Dynamic parameter scaling based on model plasticity
• Enhanced performance in encrypted traffic scenarios