QoS Optimization in 5G Networks Using Network Slicing: A Systematic Review of Algorithms, Efficiency, and Security Issues

Abstract

The emergence of 5G networks has created a growing demand for adaptive Quality of Service (QoS) optimization mechanisms due to the wide range of service requirements, including eMBB, mMTC, and uRLLC. Network slicing offers a strategic solution by logically separating network resources to ensure stable and reliable performance for each service category. This study applies a Systematic Literature Review (SLR) approach to synthesize and evaluate scientific findings related to optimization algorithms, resource efficiency, and security issues in network slicing. A total of 30 studies published between 2020 and 2025 were analyzed through a rigorous selection process using inclusion–exclusion criteria and a Quality Assessment Checklist (QAC). The results indicate that artificial intelligence-based algorithms, particularly Reinforcement Learning, significantly improve throughput and reduce latency but require substantial computational resources. Meanwhile, spectrum and energy efficiency are enhanced by adaptive orchestration techniques, although most studies still focus on single QoS parameters. Security analysis shows that cross-slice attacks remain a major threat, and integrated security frameworks are rarely implemented. This study highlights the need for a unified intelligent orchestration model that simultaneously integrates algorithmic optimization, energy efficiency, and slice-aware security to support reliable 5G and future 6G networks.