UBUNYE: um serviço de orquestração mec baseado em qoe, qos e classificação de serviços usando algoritmo de leilão e aprendizado de máquina

Abstract

The growing adoption of devices and applications in the context of smart cities has led to a significant demand for cloud services, where latency and bandwidth play a crucial role in shaping users' perceptions of network service quality. However, not all users have access to cloud services of the desired level of quality. In addition, uneven network coverage between urban and rural areas has created ``digital deserts'', which are characterized by a lack of connectivity resources, making access to cloud services even more challenging. In this context, edge computing has emerged as a promising alternative for providing services in smart cities. Edge computing leverages data processing at the source or on devices near it, rather than relying exclusively on cloud processing. This approach offers several advantages, such as reduced latency and lower bandwidth consumption. This work addresses the need to directly guarantee Quality of Service (QoS) requirements, with indirect effects on Quality of Experience (QoE), in a consistent manner within dynamic network environments, particularly in remote regions with limited infrastructure. To this end, we propose an orchestration service called Ubunye, which operates at the network edge and selects the most suitable edge node to serve a given application request while satisfying QoS requirements, with indirect impact on QoE. Ubunye is based on an auction-based algorithm and considers factors such as latency and available bandwidth when selecting a node to execute the requested service. To support service differentiation according to application requirements, Ubunye also incorporates a service classification mechanism based on machine learning techniques. We determined the optimal edge node through a multifaceted evaluation that includes current CPU load, memory availability, and other relevant system parameters. Experimental results demonstrate that Ubunye efficiently orchestrates resources at the network edge, improving QoS for services that require low latency and high bandwidth, with indirect positive effects on QoE. The results show that Ubunye achieves more than a 20\% improvement in available bandwidth compared to classical resource allocation algorithms (First Fit and Round Robin) and a double auction-based approach (Double Price Double Auction -- DPDA).