🤖 AI Summary
Existing network simulation models neglect application-layer behavior, leading to traffic distortion and hindering robustness evaluation of monitoring and anomaly detection systems. To address this, we propose the first framework that models application-layer behavior as learnable and composable probabilistic processes. Specifically, it estimates probability density functions from real-world traffic traces and employs behavioral pattern convolution to generate dynamic, scalable, and realistic traffic. We further design a lightweight simulation engine supporting coexistence of multiple applications on a single machine and real-time behavioral modulation. Experimental results demonstrate that our approach significantly improves test coverage while achieving traffic distributions closely aligned with real-world scenarios. The open-source implementation has been validated on large-scale production networks.
📝 Abstract
Existing network simulations often rely on simplistic models that send packets at random intervals, failing to capture the critical role of application-level behaviour. This paper presents a statistical approach that extracts and models application behaviour using probability density functions to generate realistic network simulations. By convolving learned application patterns, the framework produces dynamic, scalable traffic representations that closely mimic real-world networks. The method enables rigorous testing of network monitoring tools and anomaly detection systems by dynamically adjusting application behaviour. It is lightweight, capable of running multiple emulated applications on a single machine, and scalable for analysing large networks where real data collection is impractical. To encourage adoption and further testing, the full code is provided as open-source, allowing researchers and practitioners to replicate and extend the framework for diverse network environments.