Technology Category
Planning, Routing, and Scheduling: Learning for Planning and SchedulingSearch and Optimization: Learning to SearchMachine Learning: Adversarial Learning & Robustness
Deep Reinforcement Learning for Fault-Adaptive Routing in Eisenstein-Jacobi Interconnection Topologies
🤖 AI Summary
This work addresses the limitations of traditional routing in Eisenstein-Jacobi (EJ) interconnection networks, which are prone to deadlock under faults and struggle to balance fault tolerance with performance. To overcome these challenges, the paper introduces, for the first time, a deep reinforcement learning (DRL)-based adaptive fault-tolerant routing scheme that operates without requiring global topological information. The approach employs a multi-objective reward function to guide the agent in avoiding fault clusters while optimizing path efficiency, thereby implicitly achieving load balancing. Experimental results demonstrate that, under nine faulty nodes, the proposed method achieves 94% effective reachability, 91% packet delivery ratio, and normalized throughput exceeding 90%, significantly outperforming greedy routing and even approaching or surpassing Dijkstra’s algorithm in congested scenarios.
Technology Category
Application Category
📝 Abstract
The increasing density of many-core architectures necessitates interconnection networks that are both high-performance and fault-resilient. Eisenstein-Jacobi (EJ) networks, with their symmetric 6-regular topology, offer superior topological properties but challenge traditional routing heuristics under fault conditions. This paper evaluates three routing paradigms in faulty EJ environments: deterministic Greedy Adaptive Routing, theoretically optimal Dijkstra's algorithm, and a reinforcement learning (RL)-based approach. Using a multi-objective reward function to penalize fault proximity and reward path efficiency, the RL agent learns to navigate around clustered failures that typically induce dead-ends in greedy geometric routing. Dijkstra's algorithm establishes the theoretical performance ceiling by computing globally optimal paths with complete topology knowledge, revealing the true connectivity limits of faulty networks. Quantitative analysis at nine faulty nodes shows greedy routing catastrophically degrades to 10% effective reachability and packet delivery, while Dijkstra proves 52-54% represents the topological optimum. The RL agent achieves 94% effective reachability and 91% packet delivery, making it suitable for distributed deployment. Furthermore, throughput evaluations demonstrate that RL sustains over 90% normalized throughput across all loads, actually outperforming Dijkstra under congestion through implicit load balancing strategies. These results establish RL-based adaptive policies as a practical solution that bridges the gap between greedy's efficiency and Dijkstra's optimality, providing robust, self-healing communication in fault-prone interconnection networks without requiring the global topology knowledge or computational overhead of optimal algorithms.
Problem
Research questions and friction points this paper is trying to address.
fault-adaptive routing
Eisenstein-Jacobi networks
interconnection networks
many-core architectures
routing under faults
Innovation
Methods, ideas, or system contributions that make the work stand out.
Deep Reinforcement Learning
Fault-Adaptive Routing
Eisenstein-Jacobi Networks
Multi-objective Reward
Distributed Self-healing
🔎 Similar Papers
No similar papers found.
