Algorithms of Load Balancing in Next-Generation Mobile Netwoks: A systematic literature review

This work addresses a critical gap in the current literature: the lack of a unified framework to classify and evaluate load-balancing algorithms in next-generation mobile networks (e.g., 5G/6G). Through a systematic review following the PRISMA methodology, we analyzed 45 studies from five scientific databases (IEEE Xplore, Scopus, ScienceDirect, SpringerLink, and ACM Digital Library), identifying key patterns:

1. The correlation between optimization parameters (latency, throughput, user mobility) and the performance of automatic/hybrid algorithms.
2. The critical need to reduce latency for sensitive applications (e.g., Industry 4.0, autonomous vehicles).
3. The emergence of hybrid schemes combining classical machine learning with policy-based techniques (e.g., policy-based RL).