Marcos Falcão’s research while affiliated with Federal University of Pernambuco and other places

In Ultrareliable and Low Latency Communications (URLLC), balancing trade-offs between energy consumption, service availability, and strict reliability and latency requirements is a significant challenge, especially in unmanned aerial vehicle (UAV)-enabled multi-access edge computing (MEC) environments. The constraints imposed by the size, weight and power limitations of UAVs further complicate this task. This study addresses optimizing resource allocation in such environments to meet URLLC demands while minimizing power consumption and maximizing service availability. We explore the virtualization layer of the network function virtualization (NFV)-MEC architecture, incorporating node availability and power consumption alongside conflicting URLLC reliability and latency demands. We introduce an energy-aware model based on continuous-time Markov chain (CTMC) with an embedded virtual resource scaling scheme for Dynamic Resource Allocation (DRA). To solve the optimization problem related to MEC-enabled UAV node dimensioning, we propose a genetic algorithm (GA)-based solution. Our results demonstrate that the proposed GA-based approach achieves a superior balance, with up to a 44% reduction in power consumption compared to the first fit with maximum resources strategy, while also improving service availability and meeting URLLC requirements. This work provides a comprehensive analysis of key virtualization parameters and their impact on critical services within a single NFV-MEC over a UAV node, offering a robust framework for future 6 G network applications.

A Computação de Borda de Acesso Múltiplo (MEC) e a Virtualização de Funções de Rede (NFV) são tecnologias-chave da Quinta Geração de Redes Móveis (5G) para suportar serviços como o de Comunicação Ultra Confiável e com Baixa Latência (URLLC) e a Banda Larga Móvel Melhorada (eMBB). Entretanto, garantir a coexistência desses serviços é desafiador, especialmente na alocação dinâmica de recursos no domínio MEC-NFV. Este artigo apresenta um modelo baseado em Cadeia de Markov de Tempo Contínuo (CTMC) para analisar o impacto da alocação dinâmica de recursos em ambos os serviços em um ambiente MEC-NFV, considerando a sobrecarga de virtualização, falhas nos recursos virtuais e diferentes número de contêineres e tamanho de buffer. Resultados mostram que a disponibilidade, o tempo de resposta e o consumo de energia são fortemente impactados pelo número de contêineres, enquanto o tamanho de buffer afeta principalmente os tempos de resposta.

Unmanned Aerial Vehicle (UAV) communication networks and Multiaccess Edge Computing (MEC) will occupy an important position in the future wireless communication system. Unlike regular datacenter environments, MEC can help mobile devices improve computing and communication capabilities, and its combination with UAVs helps to deal with the Line of Sight (LoS) issues, besides allowing node mobility. This work addresses the dynamic resource provisioning in a UAV equipped with MEC resources (MEC-enabled UAV) that provides on demand communication capabilities to Ultra-reliable and Low-latency Communication (URLLC) services. We adopt a Continuous Time Markov Chain (CTMC) to analyze the overall node availability and reliability, while taking into account virtual host setup (repair) delays and failure events for Virtual Network Functions (VNFs) hosted on MEC-enabled UAVs. Our results show that the containerized VNF setup delays critically impact the admission process, whereas reliability is more prone towards VNF failures.

The conventional mobile architecture is unlikely to cope with Ultra-Reliable Low-Latency Communications (URLLC) constraints, being a major cause for its fundamentals to remain elusive. Multi-access Edge Computing (MEC) and Network Function Virtualization (NFV) emerge as complementary solutions, offering fine-grained on-demand distributed resources closer to the User Equipment (UE). This work proposes a multipurpose analytical framework that evaluates a hybrid virtual MEC environment that combines VMs and Containers strengths to concomitantly meet URLLC constraints and cloud-like Virtual Network Functions (VNF) elasticity.