DATUM: distributed applications for the IoT continuum

• Grant number: 22/03621-5
• Support Opportunities: Scholarships abroad - Research
• Start date: May 01, 2023
• End date: April 30, 2024
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Carlos Alberto Kamienski
• Grantee: Carlos Alberto Kamienski
• Host Investigator: Marco di Felice
• Host Institution: Centro de Matemática, Computação e Cognição (CMCC). Universidade Federal do ABC (UFABC). Ministério da Educação (Brasil). Santo André , SP, Brazil
• Institution abroad: Università di Bologna, Italy

Abstract

The Internet of Things (IoT) has been increasingly popular in the last few years, using the massive amount of data generated by sensors and processed by artificial intelligence techniques. IoT Smart Applications have to deal with the inherently distributed nature of infrastructures that span over network edge and core. This IoT Computing Continuum, or IoTinuum, comprises different geographical computing stages, such as thing, mist, fog, cloudlet, cloud, and app. The IoTinuum brings additional complexity to software development due to its inherent distributed nature, the massive number of heterogeneous connected devices, and mobility patterns. In this new programmable world, software development and deployment must consider that computation can occur anywhere, with different requirements, such as privacy, latency, efficiency, cost, sustainability, dynamicity, and mobility. Current approaches to streamline software development for the IoTinuum do not adequately address the challenges of coping with the multiple computing, communication, and storage stages. The DATUM (Distributed Applications for the IoT Continuum) project aims to propose, model, and evaluate an architectural framework, use case scenarios, algorithms, and applications involving different application models, such as federated learning. Like datum is a single piece of data, a DATUM code or model is a single and logically centralized entity that can be deployed and physically distributed over various nodes in different stages of the IoTinuum. These codes or models make it possible to develop and deploy Smart Applications with diverse requirements and use cases belonging to different categories. The proposal will be evaluated by both a qualitative and quantitative analysis. The qualitative evaluation will focus on the modeling use cases: a) smart drone delivery, b) smart irrigation, and c) smart structural monitoring. The quantitative evaluation will rely on a simulation-based performance analysis using the smart drone delivery service. These evaluations will provide a deeper understanding of the tradeoffs involving the development of applications for the IoTinuum. (AU)