This is an interesting position paper addressing the so-called Cognition-Based Networks, where optimization is achieved by using learning and distributed intelligence. Specifically, the paper proposes to combine a learning architecture with the emerging “softwatization” paradigms (SDN, NFV), eventually providing concrete application cases.
What I like most of this paper, apart the vision, is the technical approach, deeply interdisciplinary and calling for the synergic combination of know-hows of networking engineers, IT experts, cognitive scientists. Have a look at this post “Designing AI Infrastructures that Sustain the Human Race”: isn’t this the same story ? A.I. Infrastructures, Cognition-based Infrastructures, Telecommunications Infrastructures will all converge into one pervasive and distributed “Nervous System” of the Digital Society and Economy. Distinctions without differences.
This will be achieved by: pervasively sensing and collecting massive data (by sensors, terminals, smart things, machines, robots, drones, etc); exchanging data (via fixed and mobile high bandwidth and low latency links); elaborating quickly big data (with Cloud/Edge/Fog Computing) in order to make decisions (e.g., with A.I., cognition methods, algorithms, heuristics) for actuating local actions (by any actuators)…The global picture also includes 5G and IoT.
Let's overcome distinctions creating fragmentations. At the end of the day, it's about exploiting the overall control loop of “collective intelligence”, common to a bee-hive and any nervous system of a living entity! Extracting value from the information embedded in multiple interactions to achieve optimization, or, if you prefer, adaptation.
We shouldn't consider this bio-inspired perspective simply a speculation: look at it, it may open new industry roles and new service paradigms: for example the role of enabling “collective intelligence” services for the Digital Society and Economy, for the human race sustainability.
At the end of the day, any surviving open ecosystem is pursuing (and always reaching, otherwise it will disappear) “optimization”, as the best possible adaptation to the environment.
It’s a simple law of Nature.