Skip to content

Deep Reinforcement Learning Microservice Deployment Algorithm Based on Simulated Annealing Optimization in Edge Computing Scenarios

Sep 2026 · Concurrency and Computation · Vol 38 · 0 citations · 13 references

TL;DR

The proposed SA‐DRL hybrid method monitors resource usage and network conditions in real time and dynamically adjusts deployment decisions to optimize service quality and resource utilization and achieves better overall performance in terms of convergence speed, resource balance, and robustness.

Abstract

With the rapid development of microservice architectures, edge computing has become an important paradigm for service deployment. However, microservice deployment in edge environments still faces challenges such as resource constraints, network latency, bandwidth limitations, and load balancing. Achieving efficient and scalable deployment requires effective coordination of distributed resources under dynamic and heterogeneous conditions. Existing studies often overlook the time‐varying characteristics and resource heterogeneity of edge environments and lack a unified optimization framework for service quality, system stability, and migration cost. To address these issues, this paper proposes an efficient microservice deployment algorithm that combines Simulated Annealing (SA) and Deep Reinforcement Learning (DRL). The proposed SA‐DRL hybrid method monitors resource usage and network conditions in real time and dynamically adjusts deployment decisions to optimize service quality and resource utilization. Compared with three representative deployment baselines, the proposed SA‐DRL method achieves better overall performance in terms of convergence speed, resource balance, and robustness. It also maintains high efficiency and adaptability under scenarios of node‐scale expansion and workload increase. Experimental results demonstrate that the proposed method provides an effective solution for microservice deployment optimization in edge computing environments.

View source

Similar papers

#reinforcement learning Open access Oct 2026

Reinforcement Learning-Based Resource Allocation and Request Routing Optimization Algorithm for Edge AIGC Services

With the explosive growth of artificial intelligence generated content (AIGC) applications, deploying lightweight tiny neural networks (TNNs) at the edge to mitigate the high latency and cost of large cloud models has become an important trend in mobile computing. However, resource heterogeneity in edge environments, t...

Qian-Yu Cao, Chang Xu, Li Zhang et al. · 0 citations
Open access 2026

Deep Learning for Optimizing Virtual Resources Allocation in MEC-based UAV Environments

A virtualization-aware MEC framework that integrates GRU-based CPU and memory demand forecasting with dynamic virtual resource allocation in UAV-assisted environments and shows the effectiveness of proactive virtualization-aware orchestration for the UAV-assisted MEC systems is proposed.

Khadidja Zairi, Bouziane Brik, Younes Guellouma et al. · 0 citations
#reinforcement learning Open access Dec 2026

Optimizing Resource Allocation in Cloud Computing Environments using Reinforcement Learning

A deep reinforcement learning framework for intelligent cloud resource allocation that jointly optimizes resource utilization, Service Level Agreement compliance, infrastructure cost, and energy efficiency, and adapts to workload distribution shifts within 200 episodes without manual retuning is presented.

Msr Prasad · 0 citations
#edge computing Nov 2026

Hierarchical attention-based reinforcement learning for intelligent network slicing in mobile edge computing

A Hierarchical Attention-based Soft Actor–Critic (HAN-SAC) framework for adaptive network slicing that improves slice admission rates and decreases resource overhead while lowering service level agreement (SLA) violations and transmission latency in emerging 5G/6G edge networks.

Mohammad Alimorad, Babak Nouri-Moghaddam, Abbas Mirzaei et al. · 0 citations
Open access Oct 2026

Energy efficient resource allocation using deep reinforcement learning with starling murmuration optimization in cloud

Cloud computing is one of the widely used technologies due to its advanced features like pay -per-use, scalability, flexibility, and many more. It initially aims at fulfilling users’ needs for accessing computing resources and purchasing cloud services as necessary. This is all done within the framework of on-demand re...

Yogeetha Bangalore Rajanna, Anandaraj Shanthi Pichandi · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.