Alocação dinâmica de recursos em redes H-CRAN 5G e além baseada em offloading

Abstract

The exponential growth of data traffic and the diversity of mobile applications pose new challenges to wireless networks, especially in densely populated urban areas. In this context, Heterogeneous Cloud Radio Access Network (H-CRAN) architectures emerge as a promising alternative due to their flexibility, scalability, and centralized resource management. However, the efficient orchestration of users and resources remains a chal lenge in environments with high load variability and user mobility. This thesis proposes an intelligent orchestration framework for H-CRAN networks, performing the dynamic allocation of users to RRHs and their distribution across BBU sectors, based on factors such as traffic demand, QoS profile, mobility, and network topology. The proposed solu tion includes adaptive offloading, dynamic redistribution, and continuous network state assessment. The approach was evaluated through simulations encompassing 24 realistic scenarios with varying traffic loads (tidal effect). The results demonstrate significant im provements compared to the traditional TLA model, including a higher number of users served, better load balancing, reduced blocking, and improved computational stability. This research contributes a replicable and adaptive strategy for next-generation mobile networks, aligned with QoS and scalability requirements.