Research‎ > ‎

Consensus in Self-organised Wireless Networks

In order to engineer truly pervasive wireless systems, there is the need to ensure a significant level of system autonomy. In this context, self-organisation properties of complex systems allow the coordination of decentralised systems to arise out of the local interactions among system components. This process is spontaneous: it is not necessarily controlled by any agent inside or outside of the system.

However, it is still unclear how to characterise system-level behaviour that emerges from the interactions of decentralised software agents.  The emergence of a coordinated system behaviour is dependent upon the capability of the system components to reach a consensus on a single value to represent the behaviour of the complete system.

Most research in consensus problems assumes that the final agreed value is inherently constant. This may not be the case in pervasive wireless systems, since the information state of each agent may be dynamically evolving in time according to some inherent dynamics. Hence, it is important to study dynamic consensus problems where the final consensus value evolves over time as a function of environmental dynamics

Hence, I'm investigating consensus solutions applicable to pervasive wireless networks. Research effort is focus on the analysis of fundamental tradeoffs of applying swarm intelligence to consensus, taking into account the tradeoff between scalability and controllability of the network, as well as between availability, consistency and reliability of control data.