Hephaestus: Mixture Generative Modeling with Energy Guidance for Large-scale QoS Degradation
This addresses network and distributed ML system vulnerabilities, offering a novel approach for a previously unsolved nonlinear variant, though it is incremental in applying generative models to this specific domain.
The paper tackles the Quality of Service Degradation (QoSD) problem, where adversaries perturb edge weights to degrade network performance, by proposing a generative framework that synthesizes feasible solutions, and it outperforms classical and ML baselines, especially in nonlinear scenarios.
We study the Quality of Service Degradation (QoSD) problem, in which an adversary perturbs edge weights to degrade network performance. This setting arises in both network infrastructures and distributed ML systems, where communication quality, not just connectivity, determines functionality. While classical methods rely on combinatorial optimization, and recent ML approaches address only restricted linear variants with small-size networks, no prior model directly tackles the QoSD problem under nonlinear edge-weight functions. This work proposes \PIMMA, a self-reinforcing generative framework that synthesizes feasible solutions in latent space, to fill this gap. Our method includes three phases: (1) Forge: a Predictive Path-Stressing (PPS) algorithm that uses graph learning and approximation to produce feasible solutions with performance guarantee, (2) Morph: a new theoretically grounded training paradigm for Mixture of Conditional VAEs guided by an energy-based model to capture solution feature distributions, and (3) Refine: a reinforcement learning agent that explores this space to generate progressively near-optimal solutions using our designed differentiable reward function. Experiments on both synthetic and real-world networks show that our approach consistently outperforms classical and ML baselines, particularly in scenarios with nonlinear cost functions where traditional methods fail to generalize.