Toward Self-Healing Networks-on-Chip: RL-Driven Routing in 2D Torus Architectures

This article likely explores the application of Reinforcement Learning (RL) to improve the resilience and efficiency of Networks-on-Chip (NoC). The focus on 2D torus architectures suggests a specific hardware context. The term "self-healing" implies the system can automatically adapt to and recover from faults or performance degradation. The use of RL suggests an attempt to optimize routing dynamically based on observed network conditions.