ASU, Founder and Director of Blockchain Research Lab
ASU, Technical Director, Center for Assured and Scalable Data Engineering
ASU, Research Professor, School of Computing (CIDSE)
Stanford University, Distinguished Visiting Scholar, mediaX
Back in 2005, Boscovic defined a cognitive network as a network that can dynamically alter its topology and/or operational parameters to respond in the most optimal way to the needs of a particular user or a group of users while at the same time enforcing its conformity to operating and regulatory policies.
An essential attribute of cognitive networks is self-configuration - its capability to respond and dynamically adapt to the operational and contextual changes by deploying a distributed management logic in accordance with the autonomic computing paradigm. Over past couple of years this vision has been realized through Software Defined Network (SDN) instantiations which exploit machine learning techniques to make sense of contextual data collected by the participating network nodes and rely on graph theory to abstract and parametrize network's state and topology. SDN's control unit processes the data, takes adequate network management decisions and translates them into network reconfiguration commands that are understandable by the constituting networking elements.
In his latest thinking Boscovic has upgraded his views on Cognitive Networks to include not only the network nodes but also people brought together by the services and applications served by a given cognitive network. This symbiotic relation between the network and its users produces the concept of social collective intelligence blending technology with the human world made up of individuals, groups, organisations and their thoughts , actions and emotions. In such concept the cognitive network's control logic can no longer be defined in isolation from the socio-economic assumptions and incentives that will allow it to operate in harmony with the structure of society itself.