Investigation of Serverless Consumption and Performance in Multi-Access Edge Computing

Serverless computing is a promising next-gen deployment paradigm that is resource-efficient and energy-usage friendly. However, due to its current confinement to cloud-based environments, a comprehensive exploration of its applicability within heterogeneous paradigms, such as Multi-Access Edge Computing (MEC), becomes imperative. To address this gap, this paper evaluates the suitability of MEC to host serverless functions. We thoroughly assess the performance, resource utilization, and energy consumption of serverless computing through a realistic implementation. Our findings reveal that, although the current serverless design possesses certain limitations, the serverless model holds the promise of enhancing resource and energy efficiency in the context of a MEC environment.

I. Introduction

The increasing demand for automation and data-driven use cases fosters the development of the Internet of Things (loT). The number of connected devices is increasing drastically, and the IoT finds applicability in various domains, including healthcare, security, sustainability, and digital twins [1]. As a result, IoT applications generate substantial data traffic [2], which contributes to congestion, increasing resource and energy demands, and leads to an impaired Quality of Service (QoS) within conventional cloud computing environments.