arXiv subjects

Kelvin Lopes Dias

Publications and source records attributed to Kelvin Lopes Dias.

6 recordsLinked to original sources

arXiv2026

Open-Source 5G RAN Platforms: A Dual Perspective on Performance and Capabilities

Open-source implementations of the fifth generation (5G) Radio Access Network (RAN), such as OpenAirInterface (OAI) and srsRAN, have become increasingly relevant for industry and academic research, rapid prototyping, and low-cost private 5G deployments. The state of the art presents different evaluation scenarios for both platforms; however, these are often limited to individual testbed evaluations, end-to-end studies, or partial comparisons between them. In particular, there is a lack of evaluations considering different Software-Defined Radios (SDRs), RAN configurations, user applications, and numbers of real UEs. To address this gap, this paper presents a comparative evaluation of OAI and srsRAN. The study includes a qualitative assessment of supported features and deployment flexibility, followed by a quantitative performance analysis covering radio resource control setup procedures, compliance with theoretical data rates, and real application workloads with diverse quality of service requirements. Specifically, we consider Video on Demand (VoD) for data rate evaluation, Live Streaming (LS) and Cloud Gaming (CG) as latency-sensitive applications. These workloads provide a comprehensive view of each platform capability.

cs.NI↗

arXiv2026

An O-RAN-Assisted MARL Approach for Dynamic Sidelink and Infrastructure Selection in V2X Communications

Future applications in the 6G-based Internet of Vehicles will leverage sidelink (SL) transmissions in Vehicle-to-Everything (V2X) scenarios. However, SL-based direct communication can significantly increase interference among vehicles and between vehicles and other entities of the Intelligent Transportation System. Thus, both Vehicle-to-Vehicle communications and Vulnerable Road Users (VRUs) uplink resources may be degraded or subject to starvation. Existing solutions primarily focus on improving resource allocation and pair selection. Nonetheless, they lack a comprehensive approach to tackle the communication modes and the entire network. To address these challenges, this paper leverages Open RAN to manage V2X communication and proposes a multi-agent reinforcement learning (MARL) resource-aware system. Open RAN provides control loops through a global view of the network and also an open interface-based framework for machine learning models applied to resource decision-making. Meanwhile, the MARL model aims to mitigate interference, optimize resource usage, and enhance quality of service by optimally selecting between sidelink and network transmissions. To reduce system complexity, this work employs a clustering strategy. Each agent manages a group of pairs, rather than assigning one agent to each pair. The solution supports this design by adopting a centralized training with decentralized execution approach, empowered by Open RAN. The strategy uses offline training and an off-policy approach, in which each agent stores experience for fine-tuning. Results indicate that the MARL approach reduces average loss by 21% and latency by 19% in Vehicle-only scenarios. In coexistence VRU scenarios, loss and latency drop by 18% and 30%, respectively, compared to the single-agent approach.

cs.NI↗

arXiv2022

Towards a High Efficiency of Native NDN over Wi-Fi 6 for the Internet of Vehicles

Named Data Networking (NDN) is a top-notched architecture to deal with content distribution over the Internet. With the explosion of video streaming transmission and future massive Internet of Things and Vehicles (IoT/IoV) traffic, evolving Wi-Fi networks will play an essential role in such ecosystems. However, Native NDN deployment over wireless networks may not perform well. Wi-Fi broadcasts/multicasts result in reduced throughput due to the usage of basic service mode. Despite recent initial works addressing that issue, further studies and proposals are required to boost the adoption of Native NDN. We advocate that an initial step towards designing a feasible Native NDN over wireless networks should be understanding the challenges in emerging scenarios and providing a uniform baseline to compare and advance proposals. To this end, first, we highlight some challenges and directions to improve throughput and energy efficiency, reduce processing overhead, and security issues. Next, we propose a variant of NDN that minimizes the problems identified by performing transmission via unicast to avoid storms in wireless networks. Finally, we conducted a performance evaluation to compare Standard Native NDN with our proposal on Wi-Fi 6 vehicular networks. The results show that our proposal outperforms the Standard NDN in the evaluated scenarios, reaching values close to 89% of satisfied requests, achieving more than 200% of data received than Standard NDN.

cs.NI↗

arXiv2021

Native versus Overlay-based NDN over Wi-Fi 6 for the Internet of Vehicles

Internet of Vehicles (IoV) is a cornerstone building block of smart cities to provide better traffic safety and mobile infotainment. Recently, improved efficiency in WLAN-based dense scenarios has become widespread through Wi-Fi 6, a license-free spectrum technology that can complement the cellular-based infrastructure for IoV. In addition, Named Data Networking (NDN) is a promising Internet architecture to accomplish content distribution in dynamic IoV scenarios. However, NDN deployments, i.e., native (clean-slate) and overlay (running on top of IP stack), require further investigation of their performance over wireless networks, particularly regarding the IoV scenario. This paper performs a comparative simulation-based study of these NDN deployments over Wi-Fi 6 for IoV using real vehicular traces. To the best of our knowledge, this is the first effort that extends ndnSIM 2 with an overlay-based NDN implementation and that compares it with the native approach. Results show that the overlay-based NDN consistently outperforms the native one, reaching around 99% of requests satisfied, against only 42.35% in the best case of native deployment.

cs.NI↗

arXiv2020

DQN-AF: Deep Q-Network based Adaptive Forwarding Strategy for Named Data Networking

NDN has gained significant attention due to the appearance of several unforeseen design flaws that became evident with new communication scenarios. Among its many features, the two standard NDN forwarding strategies are not adaptive, causing performance degradation in several scenarios. This paper proposes an adaptive forwarding strategy based on deep reinforcement learning with Deep Q-Network, which analyzes the NDN router interface metrics without creating signaling overhead or harming the design principles from the NDN architecture, besides showing significant performance gains compared to the standard strategies.

cs.NI↗

arXiv2014

Mapping of Secondary Virtual Networks onto Wireless Substrate based on Cognitive Radio: multi-objective formulation and analysis

There is a growing demand for wireless services with different requirements in a dense and heterogeneous wireless environment. Handling this complex ecosystem is becoming a challenging issue, and wireless virtualization emerges as an efficient solution. Although the inclusion of virtualization in wireless networks ensures a better use of resources, current approaches adopted for wireless virtualization can cause an underutilization of resources, since the resource allocated to a virtual wireless network is not shared with other one. This problem can be overcome by combining wireless virtualization with the cognitive radio (CR) technology and dynamic access spectrum (DSA) techniques. Thus, virtual wireless networks with different access priorities to resources (e.g., primary and secondary) can be deployed in an overlay form and share the same substrate wireless network, where the secondary networks use the resources opportunistically. However, challenges emerge in this new scenario, ranging from the mapping to the operation of these networks. This paper is the first to propose the cognitive radio virtual network environment and to formulate the problem of mapping the secondary virtual networks (SVNs) onto the wireless substrate network based on CR. To this end, a multi-objective formulation is designed. Moreover, an analysis of the metrics defined in the problem is performed in order to give the reader useful assistance in designing schemes to solve the problem of mapping SVNs onto substrate networks.

cs.NI↗