This paper investigates the impact of Radio Link Control (RLC) protocol modes on the performance characteristics of a 5G network based on Open Radio Access Network (O-RAN) architecture. We examine how these protocol modes influence key network metrics and operational parameters through experimental evaluation of a realistic testbed setup. It is observed that the RLC Unacknowledged Mode (UM) achieves the highest throughput levels for the considered configuration, with gains of up to $13 \%$ compared with RLC Acknowledged Mode (AM). Furthermore, RLC UM outperforms RLC AM with $\mathbf{3 0 \%}$ lower retransmissions and $\mathbf{1 2 \%}$ lower latency.

Energy Efficiency (EE) has been identified as one of the main requirements within next-generation mobile networks. Regarding energy consumption of network infrastructures, the Radio Access Network accounts for most of the energy use in a cellular mobile network. On the other hand, the Open RAN architecture, focused on disaggregation and interoperability, provides a flexible architecture, facilitating the control and optimization of wireless network energy consumption. This study investigates energy efficiency optimization in Open Radio Access Network (ORAN) through transmission power control mechanisms. Using a comprehensive testbed with commercial components, including Radisys O-RAN stack and FOXCONN RPQN Open Radio Unit (O-RU), we evaluate the impact of varying transmission power levels on network performance. Our experimental results demonstrate that reducing transmission power from $\mathbf{2 4 ~ d B m}$ to 16 dBm maintains excellent RSRP levels (-74 to -84 dBm) and consistent throughput without compromising quality of service. This approach offers a more balanced alternative to traditional cell on/off mechanisms, avoiding coverage gaps while achieving energy savings. The findings suggest that intelligent power control through O-RAN’s RAN Intelligent Controller (RIC) framework can effectively optimize network energy efficiency while preserving performance quality.

This paper investigates the impact of data packet processing modes and data packet payload size on the performance characteristics of a 5 G network based on Open Radio Access Network (O-RAN) architecture. We examine how these configurations influence key network metrics and operational parameters through experimental evaluation of a realistic testbed setup. It is observed that the Data Plane Development Kit (DPDK) achieves highest performance in terms of throughput and average delay. Result shows that the Data Plane Development Kit (DPDK) mode gains are up to $220 \%$ and $43 \%$ for throughput and average latency, respectively.

This paper aims to evaluate the performance of a Fifth Generation (5G) Open Radio Access Network (RAN) by varying the distance (in kilometers) from the fronthaul interface. The objective is to analyze the robustness of the link and the Open RAN network itself, observing whether there is a loss in the stability of the connection between the User Equipment (UE) and the network and degradation of the Quality of Service (QoS) parameters (latency and throughput) after a certain distance from the link that interconnects the Radio Unit (O-RU) to the Distributed Unit (O-DU). According to the results, it was possible to observe that exceeding the theoretical fronthaul limit of 20 km, the network maintained reliable performance with minimal degradation in throughput and latency.

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.

A testbed with open-source tools and commercial equipment is designed to investigate the timing of ML-based xApp operations in DDoS attack detection during the Random Access procedure. Results show high-precision and high-recall ML classifiers operating faster than the Medium Access Control (MAC) layer's Contention Resolution timer. For classifiers with over 97% precision and recall, XGBoost has the fastest execution time and the narrowest estimated run time probability density function.

In order to expand its market share and decrease churn, cellular network operators want to achieve the maximum possible number of satisfied clients. In this paper, we propose to solve the problem of user satisfaction maximization using an utility- based dynamic resource assignment algorithm that uses a sigmoidal utility function combined with either Orthogonal Random Beamforming (ORB) or Fixed Switching Beamforming (FSB). System-level evaluations of the downlink of an Orthogonal Frequency Division Multiple Access (OFDMA) system with real-time services demonstrate that there is an optimal number of spatial beams to be used by ORB and FSB, that FSB outperforms ORB with lower complexity, and also that the proposed joint resource assignment and beamforming outperforms other classical techniques found in the literature.