Carbon-Aware Reinforcement Learning for Sustainable Task Scheduling in Edge Computing: A Review

This review examined recent research on reinforcement-learning-based and carbon-aware task scheduling in edge computing. The analysis identified key challenges related to carbon-intensity awareness, scalability, multi-objective optimization, and SLA protection. To address these gaps, the proposed CRTSE framework integrates K-means-based task clustering, Q-learning, carbon-aware scheduling, and SLA-aware task placement within a unified framework. The reward formulation jointly considers latency, energy consumption, carbon emissions, and SLA compliance. Since CRTSE is presented as a conceptual framework, its effectiveness requires validation through simulation and real-world experiments. Future research can investigate deep reinforcement learning, federated learning, predictive carbon-intensity models, and large-scale real-world validation to improve the practical applicability of carbon-aware edge computing.

[1] M. Cheong, H. Lee, I. Yeom, and H. Woo, "SCARL: Attentive Reinforcement Learning-Based Scheduling in a Multi-Resource Heterogeneous Cluster," IEEE Access, vol. 7, pp. 153432–153444, 2019. [2] X. Lu, C. Liu, S. Zhu, Y. Mao, P. Lio, and P. Hui, "RLPTO: A Reinforcement Learning-Based Performance-Time Optimized Task and Resource Scheduling Mechanism for Distributed Machine Learning," IEEE Trans. Parallel Distrib. Syst., vol. 34, no. 12, 2023. [3] Y. Wang, K. Deng, Y. Ren, J. Chan, G. Tian, and C. S. Jensen, "Reinforcement Learning for Tackling Energy-Saving and Energy-Balance Dilemma of Cluster-Based Routing Protocols in WSNs," IEEE Access, vol. 12, 2024. [4] A. Martin, I. de-la-Bandera, A. Mendo, J. Outes, J. Ramiro, and R. Barco, "Federated Deep Reinforcement Learning for ENDC Optimization," IEEE Trans. Mobile Comput., vol. 24, no. 6, 2025. [5] H. Liu, L. Tian, and M. Guo, “CRTSE: Clustering and Reinforcement Learning Based Task Scheduling Algorithm for Edge Computing,” IEEE Trans. Cloud Comput., vol. 14, no. 1, pp. 129–145, 2025, doi: 10.1109/TCC.2025.3640958. [6] L. Wang et al., "Carbon-Aware Task Offloading for Sustainable Edge Computing," IEEE Internet Things J., 2024. [7] G. Gao et al., "Carbon-Aware Task Scheduling for Cloud-Edge IoT Systems," IEEE Internet Things J., 2026. [8] C. P. Leong, C. S. Liew, C. S. Chan, and M. H. U. Rehman, “Optimizing Workflow Task Clustering Using Reinforcement Learning,” IEEE Access, vol. 9, pp. 110614–110626, 2021. [9] A. Radovanović, R. Koningstein, I. Schneider, B. Chen, A. Duarte, B. Roy, D. Xiao, M. Haridasan, P. Hung, N. Care, and S. Talukdar, "Carbon-Aware Computing for Datacenters," IEEE Trans. Power Syst., vol. 38, no. 2, pp. 1270–1280, 2023. [10] A. Souza, S. Jasoria, B. Chakrabarty, A. Bridgwater, A. Lundberg, F. Skogh, A. Ali-Eldin, D. Irwin, and P. Shenoy, "Casper: Carbon-Aware Scheduling and Provisioning for Distributed Web Services," in Proc. 14th ACM Int. Green and Sustainable Computing Conf., 2023, pp. 67–73.