The pervasiveness of Internet, the progresses of communication, processing and storage technologies and the evolution of social media are transforming human communication. We are communicating with other people via phone calls, messages, e-mails, and we are updating each other through social networks like Twitter, Facebook or Linkedin.  We are also sharing photos on Flickr and videos on YouTube.  We are more and more consuming, but also creating, contents that are shared with friends and other people.

There are also several tools allowing us to customize the content we create and consume (searches, alerts, RSS readers, etc.). And maybe, we will spend more and more time on watching, listening (e.g., multi media contents), speaking and hopefully reading (e.g. e-books). I imagine that all of this will allow each of us to create and customize sort of Personal Knowledge Ecosystems supported, and interconnected with each other, by future networks. Indeed information, knowledge, multi media contents will be seeds for said ecosystems.

This evolution will impact future networks, which should be able to transport and manage huge amounts of short digital bursts of multimedia contents and information (e.g., audio and video elements). Also, future networks should also provide multi-scale reaction times in handling said huge amounts of data. This reminds me once more the brain networks.

I’m wondering whether we can look at progresses in neuroscience when designing future networks. Cognitive scientists are just beginning to understand human brain mechanisms that underlie complex cognitive processes in managing images. They are arguing that the small world and exponentially truncated power law characteristics of brain network might be the best structural basis for the rapid generation and transfer of information through the coexistence of both informational processing segregation (in specialized regions) and integration (by coherent oscillations in wider regions).

Communications, storage and computing services are becoming more and more pervasive and, in the future, will be embedded in all everyday objects and in the physical environment. This will determine an increasing complexity of future networks and the consequent challenge of managing large amounts of devices and nodes, moreover based on heterogeneous technologies. This cannot be handled by a traditional centralized approach, but there will be the need to evolve towards distributed approaches. In particular, concerning the traditional areas of resource management (such as fault, configuration and performance), a promising approach is the distribution and automation of the decision-making control-loops and the actuation processes (thus hiding complexities and reducing human efforts and mistakes). Even more, these control loops may include also sophisticated cognitive functions (such as planning and learning).

In order to reach such objective, it is necessary to embed in future objects/devices/nodes autonomic components (with self-* capabilities) virtualising physical resources and interacting with each other to realize services and applications. Architectures based on such distributed autonomic components are intrinsically robust, can dynamically self-configure, evolve their plans and self-organize even to adapt to unexpected situations. Eventually future autonomic systems and networks will consist of different nested and linked control loops exhibiting learning and reasoning capabilities to support a sort of cognitive cooperation based on a shared (network) knowledge.

The key requirement leading to such robust, scalable and adaptable behaviours is coordination and cooperation between components. This is posing an even more critical question: how representing the information shared by (and necessary to) the components? How information is updated and transformed into knowledge? How selecting a relevant part of the global information/knowledge for certain goals?

A possible way to approach this issue for distributed architectures is to consider a sort of information/knowledge field (adopting metaphor of field in Physics), spatially distributed across components and as such resources. Information/knowledge field looks like a sort of gravitational field emitted by each component itself. This information/knowledge field could be locally created and accessible by all autonomic components depending on their location, thus providing them with a local context of the global situation of the network. By following the local shape of this field, a component itself can intrinsically perform plan evolution, decision-making and actuation processes. In other words components sense the information/knowledge field and act accordingly to evolutionary behavioural patterns or plans.

Information/knowledge field models can potentially be implemented, as an overlay network, on any middleware, providing basic support for data storing, communication and event-notification. What is required is to provide simple storage mechanisms (to store field values), communication protocols and primitives (to propagate field values to peers), and basic pub-sub mechanisms. In this context it seems promising also to consider the application of bio-inspired mechanisms and algorithms (i.e. gossiping, reaction-diffusion, gradient, self-organization) for enabling creation and development of information/knowledge field across all (communication, storage, computing) pervasive resources.

Neuroscientists argue that brain networks show small world and exponentially truncated power law characteristics. These are considered, by the way, the best structural basis for coexistence of both informational processing segregation (in specialized regions) and integration (by coherent oscillations in wider regions).

Most of neuroscience investigations, as far as I know, are currently based on the assumption that thoughts, feelings, consciousness emerge from the electrical and chemical communications between brain cells. On one side there are experimental studies, using different anatomical technologies (e.g. magnetic resonance imaging) to observe and elaborate about brain patterns. On the other side, there are studies and simulations on various computational models of complex networks neurons: for example how neurons fire electrically in response to inputs of other neurons, or release neurotransmitters to communicate with each other.

Anyway, how such network of networks of cells can produce the subtlety of mind, or can perform the body autonomic control, is still a big open question.

An interesting approach [1] proposes alternatively to focus on the relevance of abstractions (and their relationships to the natural evolution, or the human design).  Being an electronic engineer (fond of Physics) let me resume the example of the Ohm’s law, V = IR, where V, I and R denote abstract entities respectively known as voltage across a resistor, current through it and resistance. So, the movement of electrons, in a physical resistor, results in the realisation of an abstract scheme satisfying the Ohm’s law. Engineers make use of such (and many others) abstract schemes to design electronic systems performing particular functions (but, basically ignoring the laws behind the movement of electrons). The same ideas probably apply to bio-systems. Bio-systems contain several components of various kinds, conforming to some scheme of abstractions. On the other hand, they differ from machines in that the entities concerned often lack a formal definition, their properties being inferred from investigations (at different levels of observation) of several instances encountered in nature.

 
This reasoning equally applies to autonomic systems functioning.  The human autonomic nervous system, in its environment, is probably characterised by a hierarchy of systems conforming to a range of abstract schemes; typically these schemes relate to particular neural circuits or biological sub-systems and their adaptive behaviour in a given environment.

Natural design and evolution are driving (indirectly) the human autonomic nervous system to conform to such abstract schemes, through development and cognition (learning, at least). I think this is an important point.

Following this reasoning, to engineer autonomic systems requires defining, not only control-loop circuitry (at the proper level), but also hierarchies of abstract schemes and the processes of evolutionary development and cognition in dynamic environments.

In a next post, we’ll elaborate about how cognition may control phase-transitions in autonomic networks.

[1] http://www.tcm.phy.cam.ac.uk/~bdj10/papers/messina2001.html

Indeed looking at nature is fascinating and amazing.

Let me make the example of the study of brain anatomy. It seems that there are empirical and theoretical reasons in favor of considering “small world networking” as a model of brain anatomical and functional architecture. Some papers that I’ve recently read, e.g. [1], [2], provided first experimental demonstrations of small-world properties in brain functional networks as derived from fMRI (functional Magnetic Resonance Imaging) data analysis.

Moreover, other publications, e.g. [3], showed that the degree distribution of brain networks in cortical regions can be described as an exponentially truncated power law (truncated power law degree distributions are typical of complex systems that are physically embedded or constrained, such as transport or infrastructural networks, and in those systems in which nodes have a finite life span).

Scientists starts arguing that small world and exponentially truncated power law characteristics of brain network might be the best structural basis for the coexistence of both informational processing segregation (in specialized regions) and integration (by coherent oscillations in wider regions), for the rapid generation and transfer of information, and for the robustness of brain networks [4].

This brings to my mind some questions. Can the natural evolution of small-world brain networks be really read in terms of maximizing brain complexity, robustness (to local pathological attack), effectiveness (of cognitive capabilities) while, at the same time, minimizing some cost parameters ?

Small-world networks can be created in a multitude of ways: by randomly rewiring links of a regular graph, adding new links to it with a certain probability, or letting sites connect locally to geographically nearby sites: which of these (or other) possible (small-world networks) growth rules are most strategic biologically?

How the cognitive capacities are associated with (emerge from) configurations of brain functional networks ?

References

[1] Salvador R, Suckling J, Coleman MR, Pickard JD, Menon D, Bullmore E. “Neurophysiological architecture of functional magnetic resonance images of human brain” Cereb Cortex 15:1332–42, 2005.;

[2] Salvador R, Suckling J, Schwarzbauer C, Bullmore E. 2005. „Undirected graphs of frequency-dependent functional connectivity in whole brain networks” Philos Trans R Soc Lond B Biol Sci 360:937;

[3] Achard S, Salvador R, Whitcher B, Suckling J, Bullmore E. “A resilient, low-frequency, small-world human brain functional network with highly connected association cortical hubs” J Neurosci 26:63–72, 2006; 

[4] Smith Bassett D., Bullmore E., “Small-World Brain Networks” The Neuroscientist volume 12, N. 6, 2006.