MD.20.008 – Complex Causal IoT Networks

Route: Measuring and detecting: anything, anytime, anywhere

Cluster question: 123 How can we manage the unpredictability of complex networks and chaotic systems?

The Internet of Things is composed by a vast number of devices, heterogeneous in function and connectivity, applications range from telemetry to smart city, traffic, health actuation and disruptive smart grid applications. Ensuring that the currently designed communication infrastructure will be capable to collect and transport the massive amount of data generated by the distributed devices and, at the same time, ensure a reliable and robust feedback to the actuators is no small task. Current networks are not meant to inter-operate and have been designed to be part of separate communications systems, rather than allowing seamless and heterogeneous data transfer. Network management solutions, based on the interactions between all the stack layers, need to consider the interactions between all the communication agents, in time and space, to determine causal relationships in a heterogeneous network and allow dynamic and distributed management.
The first focus of this project is to develop a functional model of hierarchical communication systems for the Internet of Things. The basis of this work is rooted in a new rediscovery of complexity theory as a powerful framework for the understanding of the interactions within large, heterogeneous systems in which a central controller is either impossible or undesirable.
Secondly, the novel taxonomy developed will be used in cross-layer multi-agent simulators developed for the purpose to identify causality links within the communication networks. Techniques borrowed from climate science, econometrics and machine learning will be employed to determine how information spreads over a network, from node to node and across the layers, and which elements are relevant for a dynamic and proactive management of complex, large and dense IoT networks.
Finally, novel network management protocols will be developed to enable the steerage of local IoT nodes towards optimal global behaviour by utilizing only local information and allow interoperability across different networks.

Keywords

causality, Complex Networks, Emergence, Heterogeneous Internet of Things, Interoperability, Multi-agent

Other organisations

Universiteit Twente (UT)

Submitter

Organisation University of Groningen (RUG)
Name prof. dr. ir. M. (Ming) Cao
E-mail m.cao@rug.nl
Website https://www.rug.nl/staff/m.cao/