Staying Young at the Edge: A Software Aging Perspective for Foundation Models as a Service

Nowadays, the emergence of Foundation Models as a Service (FMaaS) enables mobile users11 to access powerful capabilities — such as inference and fine-tuning — on demand, without12 incurring local computational overhead. This paper introduces a software-aware offloading13 framework for FMaaS, allowing edge nodes to forecast software aging and prevent service14 degradation. Each node employs a lightweight Echo State Network to predict its software15 age, and tasks are dynamically assigned based on communication cost, inference delay, and16 forecast reliability. Simulation results, including ablation studies, confirm the effectiveness17 of software age forecasting in reducing task failures and improving session continuity.