Sustainable Graph Analytics Workload Scheduling with Evolutionary Reinforcement Learning in Edge-Cloud Systems

TL;DR

This paper introduces MERSEM, a novel multi-objective evolutionary reinforcement learning framework that optimizes graph workload scheduling in edge-cloud systems to reduce SLA violations and carbon emissions.

Contribution

It presents a new hybrid approach combining evolutionary search and reinforcement learning for sustainable workload management in heterogeneous edge-cloud environments.

Findings

MERSEM reduces SLA violations by up to 45%.

MERSEM cuts carbon emissions by up to 12%.

The framework effectively balances performance and sustainability objectives.

Abstract

Graph analytics powers modern intelligent systems such as smart cities, cyber-physical infrastructure, IoT security, and large-scale social networks. As these workloads scale in complexity, their execution in heterogeneous edge-cloud environments results in higher energy use and carbon emission footprint. To address this challenge, we propose MERSEM, a multi-objective evolutionary reinforcement learning framework for sustainable edge-cloud system management. MERSEM integrates evolutionary search with reinforcement learning (RL) to solve the problem of graph workload allocation and scheduling. The evolutionary component explores diverse global solutions, while the RL agent refines decisions through adaptive local optimization. The framework is designed to jointly minimize service-level agreement (SLA) violations and carbon emissions by considering dynamic carbon intensity, resource heterogeneity, and workload characteristics. Experimental results demonstrate that MERSEM outperforms the state-of-the-art with up to 45% SLA violation reductions and up to 12% carbon emission reductions.