Talks and Poster Presentations (with Proceedings-Entry):
R. Casadei, M. Viroli, C. Tsigkanos, S. Dustdar:
"Engineering Resilient Collaborative Edge-enabled IoT";
Talk: IEEE International Conference on Services Computing (IEEE SCC 2019) part of the 2019 IEEE World Congress on Services,
Milan, Italy;
2019-07-08
- 2019-07-13; in: "Proceedings of the IEEE International Conference on Services Computing (IEEE SCC 2019) part of the 2019 IEEE World Congress on Services",
E. Bertino, C. Chang, P. Chen, E. Damiani, M. Goul, K. Oyama (ed.);
IEEE,
(2019),
ISBN: 978-1-7281-2721-7;
36
- 45.
English abstract:
Novel scenarios like IoT and smart cities promote a vision of computational ecosystems whereby heterogeneous collectives of humans, devices and computing infrastructure interact to provide various services. There, autonomous agents with different capabilities are expected to cooperate towards global goals in dependable ways. This is challenging, as deployments are within unknown, changing and loosely connected environments characterized by lack of centralized control, where components may come and go, or disruption may be caused by failures. Key issues include (i) how to leverage, functionally and non-functionally, forms of opportunistic computing and locality that often underlie IoT scenarios; (ii) how to design and operate large-scale, resilient ecosystems through suitable assumptions, decentralized control, and adaptive mechanisms; and (iii) how to capture and enact "global" behaviors and properties, when the system consists of heterogeneous, autonomous entities. In this paper, we propose a model for resilient, collaborative edge-enabled IoT that leverages spatial locality, opportunistic agents, and coordinator nodes at the edge. The engineering approach is declarative and configurable, and works by dynamically dividing the environment into collaboration areas coordinated by edge devices. We provide an implementation as a collective, self-organizing workflow based on Aggregate Computing, provide evaluation by means of simulation, and finally discuss properties and general applicability of the approach.
Keywords:
self-organization; situated problem solving; decentralized coordination; collective intelligence; edge computing.
"Official" electronic version of the publication (accessed through its Digital Object Identifier - DOI)
http://dx.doi.org/10.1109/SCC.2019.00019
Created from the Publication Database of the Vienna University of Technology.