SIA: Symbolic Interpretability for Anticipatory Deep Reinforcement Learning in Network Control
This addresses the barrier to adopting proactive control in next-generation mobile networks by making anticipatory DRL transparent and tunable for network operators.
The paper tackles the problem of interpretability in forecast-augmented deep reinforcement learning (DRL) for network control by proposing SIA, an interpreter that exposes how predictions guide decisions in real time. It achieves sub-millisecond speed, 200x faster than existing methods, and improves agent performance by up to 25% in specific networking tasks.
Deep reinforcement learning (DRL) promises adaptive control for future mobile networks but conventional agents remain reactive: they act on past and current measurements and cannot leverage short-term forecasts of exogenous KPIs such as bandwidth. Augmenting agents with predictions can overcome this temporal myopia, yet uptake in networking is scarce because forecast-aware agents act as closed-boxes; operators cannot tell whether predictions guide decisions or justify the added complexity. We propose SIA, the first interpreter that exposes in real time how forecast-augmented DRL agents operate. SIA fuses Symbolic AI abstractions with per-KPI Knowledge Graphs to produce explanations, and includes a new Influence Score metric. SIA achieves sub-millisecond speed, over 200x faster than existing XAI methods. We evaluate SIA on three diverse networking use cases, uncovering hidden issues, including temporal misalignment in forecast integration and reward-design biases that trigger counter-productive policies. These insights enable targeted fixes: a redesigned agent achieves a 9% higher average bitrate in video streaming, and SIA's online Action-Refinement module improves RAN-slicing reward by 25% without retraining. By making anticipatory DRL transparent and tunable, SIA lowers the barrier to proactive control in next-generation mobile networks.