Maykon Renan Pereira da Silva

Resumo-Este trabalho propõe uma abordagem com Deep Reinforcement Learning (DRL) para otimizar a alocação de recursos para nós inteligentes em redes Open RAN. O foco está na seleção eficiente de nós near-RT-RICs considerando tipos distintos de fatias da rede. O agente DDPG decide quais nós utilizar, como alocar recursos computacionais para o processamento das tarefas nos near-RT

Resumo-As redes Open RAN, com sua proposta de desagregação e interfaces abertas, são fundamentais para viabilizar a inovação e a interoperabilidade nas gerações pós-5G. Dentre suas interfaces, o fronthaul (O-FH) se destaca por sua sensibilidade à latência, sendo crítico para o desempenho global da rede. No entanto, há uma lacuna na literatura quanto a avaliações experimentais dessa interface em cenários realistas

This paper presents the OpenRAN@Brasil testbed, designed to address the challenges of Open RAN deployments in Brazil. It is composed by commercial off-the-shelf 5G Open RAN components, distributed over two sites, and interconnected by a 400 km midhaul link. We use this to evaluate end-to-end network performance (latency and throughput) under diverse signal conditions. Results show that the long-haul link increases latency to up to 28 ms, degrading downlink throughput by up to 400 Mbps, while the lower uplink throughput is less affected. These findings highlight the importance of strategic RAN component placement and robust signal quality for optimizing Open RAN performance in challenging environments, particularly within the Brazilian ecosystem.

Research has shown that analysis and modeling techniques that provide a better understanding of the behavior of network traffic flows are very important in the design and optimization of communication networks. For this reason, this work proposes a multifractal model based on a multiplicative cascade in the wavelet domain, to synthesize network traffic samples. For this purpose, in the proposed model, a parametric modeling based on an exponential function is used for the variance of the multipliers along the stages of the cascade. The exponential function parameters are obtained through the solution of a non-linear system, for this purpose, the Levenberg-Marquardt method is used. The main contribution of the proposed algorithm is to use a fixed and reduced number of parameters to generate network traffic samples that have characteristics such as self-similarity and wide Multifractal Spectrum Width (MSW) similar to the real network traffic traces and without the need for prior adjustment of these parameters.