Interpretable Attention-Based Multi-Agent PPO for Latency Spike Resolution in 6G RAN Slicing

2026-02-11

AI Summary

This work addresses the challenge of diagnosing sporadic latency spikes in 6G radio access networks and the lack of interpretability in conventional deep reinforcement learning approaches, which hinders compliance with stringent service-level agreements (SLAs). To this end, the authors propose Attention-Enhanced Multi-Agent Proximal Policy Optimization (AE-MAPPO), which integrates interpretable attention mechanisms directly into a multi-agent PPO framework. The method employs six specialized attention modules to enable three-stage cross-time-scale control—prediction, response, and inter-slice optimization—within the O-RAN architecture. AE-MAPPO delivers high-fidelity decision explanations without incurring additional overhead, ensuring SLA adherence for eMBB, mMTC, and URLLC network slices. In URLLC scenarios, it eliminates latency spikes within 18 ms, restoring latency to 0.98 ms while achieving 99.9999% reliability and reducing fault diagnosis time by 93%.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Multiagent Systems: Multiagent Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Sixth-generation (6G) radio access networks (RANs) must enforce strict service-level agreements (SLAs) for heterogeneous slices, yet sudden latency spikes remain difficult to diagnose and resolve with conventional deep reinforcement learning (DRL) or explainable RL (XRL). We propose \emph{Attention-Enhanced Multi-Agent Proximal Policy Optimization (AE-MAPPO)}, which integrates six specialized attention mechanisms into multi-agent slice control and surfaces them as zero-cost, faithful explanations. The framework operates across O-RAN timescales with a three-phase strategy: predictive, reactive, and inter-slice optimization. A URLLC case study shows AE-MAPPO resolves a latency spike in $18$ms, restores latency to $0.98$ms with $99.9999\%$ reliability, and reduces troubleshooting time by $93\%$ while maintaining eMBB and mMTC continuity. These results confirm AE-MAPPO's ability to combine SLA compliance with inherent interpretability, enabling trustworthy and real-time automation for 6G RAN slicing.

Problem

• latency spike
• 6G RAN slicing
• service-level agreements
• multi-agent control
• interpretability

Innovation

• Attention Mechanism
• Multi-Agent PPO
• Interpretable RL
• 6G RAN Slicing
• Latency Spike Resolution