As todays architectures based on switching evolves into massive distributed data centers interconnected by an oversized mesh of optical fibers, with the last meters covered by wireless, control “vanishes”. We have been seeing control moving progressively from the center to the edges but in the next decade those edges are likely to be autonomous systems and at the core of autonomous systems there is the concept of a fading overall control replaced by local decisions that take into account the context and that in turns change the context.

Brain Connectivity

We can look at future infrastructures more in terms of a fabric, like the ones we have in our brain. In our brain we can recognize many networks but they are so intertwined and they influence one another well beyond direct communications (in addition to axons spikes processing is influenced by serotonin and dopamine and it is difficult to separate communications from processing and from “data”)  that it makes more sense to talk about a communications fabric.

One might even say that we will see a dynamically evolving communications fabric moving from one state to the next, although it may be impossible to describe it in terms of a finite state machine. What will be more and more apparent is that as users and as owners of network resources we will see emergent behavior rising from the way the fabric is being used.

This is due to the evolution in the three pillars of the fabric: communication links capacity, processing nodes (capacity and number) and data (volume, variety and variability). Once a thresholds is reached we see that the overall system becomes an eco-system, better modeled with the mathematics of the small worlds than with the one of finite state machines. In this new game a major role is played by devices, as small as an embedded chip in the Internet of thing or as complex as a connected human being.

I participated today in a discussion on what future networks will look like. Of course one has to qualify the “future”: are we talking about end of this decade, of the next or end of this century? What really matters, beyond the pleasure to speculate about possible “futures” is what could be the implication on actions to be taken tomorrow morning.

In this sense, an horizon at the end of this century is useless, as it is an horizon for the end of this decade (that is if you are interested in radically new architectures and paradigms) since the story for this decade, in terms of deployment is already written and it is about a linear evolution of what we have: more fibre, competition with wireless.

So let’s focus on a fifteen-twenty years horizon. It looks far away but for the king of investment involved in creating a new network Europe-wide (this is the scale we are looking at) it means mobilizing some hundred billion euros and this doesn’t happen overnight.

So what is my vision?

Looking at network technology what I see is an embedding of optical connectivity in chips. This will enable lower power consumption at the chip level and at the board level. Once you have an optical interconnection between two connected chips distance becomes irrelevant (that is not true in terms of processing, since two chips on the same board are separated by one nanosec, whilst two chip one in Italy and another in UK are separated by  0.5 ms but in a telecommunications framework that distance can be considered as 0).

Hence, I claim that in the future we can imagine the communications network as a set of interconnected chips, in other words as a massive distributed data base.

We can therefore use as a blueprint for a European-wide network the interconnection paradigm used today within a supercomputer (that in a few years will be used in data centers).

Now, if you look at communications within a supercomputer you don’s see circuits (or packet switching) rather you see buses and applications spawning. It is like saying that a new call is about creating a processing instance on two (or more) processors connected by a bus. This is what I feel networks in the next decade will start to look like.

Companies managing large distributed data bases (like Google, Apple, Facebook and Amazon) are the ones that are more likely to initiate this evolution. And Telecom Operators will follow suit as this kind of communications infrastructures is cheaper than the one we have today.

Today there are plenty of open source software solutions that can be used to implement a fully open Cloud Computing environment; just to mention some of them: libcloud, OpenStack, NiftyName, Juju, appscale, SlapOS, buildout, supervisord, PyOCNI etc. Imagine using such solutions to create an ICT environment exploiting end-Users’ idle resources (instead of the servers in traditional data centers) for providing computing and storage services. This is Fog Computing at the Edge !

Fog Computing is about extending the Cloud Computing paradigm up to the edge of the network, by using a sheer number of unused ICT resources. It is not just a new tech buzzword, but it is about the migration trend of processing power, storage capability and embedded communications towards the edge of the network, which means in the hands of the Users. Fog computing could be also about storage for disaster recovery.

This is going to become a reality today. Symform is an example of start-up offering disaster resilience as a “decentralized, distributed, virtual, and crowd-sourced” fog. Let’s see how it works. Some Symform’s Users act also as hosts by allocating some amount of their on-site unused storage for use by Symform: pricing is 15 cents per gigabyte per month but if they provide as much storage resource as twice the data they are uploading, then their fog storage is free. When a User uploads a file to Symform’s fog, the system replicates it for redundancy, shreds it into tiny pieces, encrypts each piece, and then distributes it to other Symform Users. The system splits each 64 megabyte block of data into 96 fragments; only 64 of those fragments are necessary to recreate the entire block.

One may wonder about the performance of Fog Computing. Well, this brings me back to folding@home, a crowdsourcing initiative about computing intensive simulations of protein folding and other types of molecular dynamics. Folding@home uses the idle processing resources of thousands of personal computers owned by volunteers. As of November 28, 2012, folding@home has 208,622 active CPU cores, 10,206 active GPUs, and 4,583 active PS3s, for a total of about 5 petaFlops! (a petaFlop is a quadrillion calculations – 10¹⁵ – per second). Titan, first supercomputer in the world, has reached today a speed of 17.59 petaFlops.

Just imagine the variety of ICT services that could be executed and provided by orchestrating the idle computing and storage local resources of millions of smart nodes at the edge…one may argue that not all types of services and applications can run entirely on the edge, however, there are several examples like disaster resilience, content aggregation and transformation, data collection and analytics, static data bases, and many others (even at lower OSI layers) which can benefit from the fog.

I still remember reading as a boy the Jules Verne adventure of Nemo and its submarine. And that is what came to my mind reading today of the 550,000 miles of undersea cables, each one about five cm thick, zipping terabytes of data across the Oceans. There are also satellites providing connectivity across the globe but their contribution to the global connectivity is about 1% (and decreasing…).

Internet seems to most of us as something that just exists, and often we feel it as decoupled from the telecommunications networks.

In fact, Internet could not exist without the telecommunications network!

In the map you see these submarine cables linking the continents (and you can also get a very detailed map if you fancy).

It is interesting to notice that a number of these cables are future proof: in the past once a cable reached its maximum load the only option was to lay another one. Nowadays, optical fibre can increase their capacity by replacing the optoelectronics at the ends. This is much cheaper than laying a new cable and keeps the pace of optoelectronics evolution.

What we are now starting to see is that the telecommunications network, at the logical layer (protocols) is relying more and more on Internet so that by the end of this decade the two of them will probably become synonyms.

Fog Computing vision goes beyond Cloud Computing by arguing the use a sheer number of resources distributed at the edge of the network. Apparently another buzzword. On the other hand, today we are already witnessing a progressive migration of processing power, storage capability and embedded communications towards the edge of the network; this trend, coupled with devices miniaturization and costs reductions, will create the conditions whereby Users will literally “decide and drive”  future ICT networks and services. This will have big impacts. This floating “fog” of ICT resources at the edge will give rise to new biz models based on new forms of competition and cooperation between existing Providers, and new  ones entering the arena, including utilities, car manufacturers, consumers’ electronics, public administrations, communities, etc.

In these dynamical games we’ll see innovative proposals, rewarded directly by the market itself, which will be essential encouragement for further investments. So, ideally, in the edge ICT fabric it will be possible creating, programming, instantiating or migrating dynamically different types of virtual functionalities and services as well as alternatives of the same. No more ossified architectures. In other words sort of ephemeral networks of resources will plastically self-adapt to humans’ dynamics. And it shouldn’t be a surprise discovering that this follows the laws of emergence in “complex systems”!

Emergence of flocks of birds: each individual responds to local conditions with a similar rules set.

Emergence is topic already acquiring a growing interest in social networks: there are interests in modeling and predicting the dynamics of groups of people, the viral diffusions of certain ideas or concepts, use of resources, or even the potential adoption of product and services.  Think about the the convergence of Internet and the social attitude of humans: beyond sites such as Facebook, LinkedIn, MySpace, Wikipedia, YouTube there is a broader process to form connections with others, build groups and to engage communications.  A political message, or a piece of news or a meme are examples of information that can spread from person to person, in an epidemic way. This can catch the attention of millions of people creating ephemeral human dynamics, made visible by on-line expressions.

 Modeling ICT social epidemics provides the opportunity to identify influential behaviors, or ways to predict, trigger or incentive mass adoptions of products or ICT services. Several mathematical approaches have been proposed for modeling these dynamics: a diffused idea is modeling the state of each person as a member of a lattice and updating it by using simple rules depending on the states of neighboring members. Each state could be represented by a set of variables such as cultural skill, preference, beliefs, etc. and each of them associated with a certain flipping energy, which is the cost of changing the state given its connection with other members, which are in other variables-states. At the end of the day, it’s about how environment conditions, or messages, will influence people and vice-versa how individuals influence each other and the environment.  In a next post I’ll provide some simulations examples.

 In summary, as in any complex system, also in the Edge ICT Fabric, it will a matter of taming complexity and extracting simplicity out of “local-to-global” dynamics.

With today’s computer you are unlikely to reach with your hand inside your jacket, take out a roll and unroll it saying “Let me unroll my computer”.

The Plastic Logic Paper Tablet

We have seen, also discussed in this blog, a number of inventions of ways to use plastic, or plastic support, to design electronic circuits as well as foldable screens.

But now Plastic Logic is demonstrating a Paper Tablet, shown in the photo on the left.

They called it a PaperTab (short for Paper Tablet I guess) and they developed it in conjunction with the Human Media Lab at Queen’s University. The tablet is powered by a second generation Intel® CoreTM i5 Processor.

Interestingly, but not surprisingly since the Human Media Lab was involved, they are also proposing a new paradigm of interaction for this tablet that looks and feels like a sheet of paper. Rather than the usual screen with several applications on it that you select and run, here the idea is that every “sheet of paper” (every PaperTab) contains just one application. If you need to change application you get another sheet. May be one with a different colour.

Clearly you might end up with hundreds of sheet but on the other hand how many apps are you actually using most of the time. Just a few, and it might make sense to characterise each of them with a specific sheet!

Something that puzzles me is how thin it is, well like a sheet of paper, and the fact that to plug in a connector you need to have a bulge! But probably it will be just a matter of time and we will get rid of the connectors and of the bulge!

I have to confess that I am not really buying into the idea of one application – one sheet, but on the other hand it is interesting to think that form (media) can really change a paradigm we have been taking for granted for many years.

DNA is the well known macro-molecule with a double helix, encoding all genetic information in a language with 64 three-letter words built from an alphabet with a set of four different letters. The used symbols are A, C, G, T and mean adenine, cytosine, guanine and thymine (thymine T is replaced by uracil U in RNA). Since the discovery of the molecular structure of DNA in 1953, by Watson and Crick, a lot of progresses have been made in studying the ensembles of molecular structures of the genetic code.

Scientists are investigating why this special language has been chosen by Nature.

As we have read in the last post there are effort for mimicking this language in informatics:    an avenue towards DNA-based computing and bioinformatics. On this matter, let me go back again to symmetry.

DNA has two helices, which run anti-parallel to each other: this is an inherent symmetry, which is highly important in the replication process of DNA. Furthermore, they say that the genetic code has an exact A-G permutation symmetry and an almost exact T-C permutation symmetry with respect to the third nucleotide. Given the enormous importance of spontaneous symmetry breaking in several physical phenomena, these symmetries in the genetic code are even more amazing!

In 1993 this paper proposed explaining the degeneracy of the genetic code as the result of a symmetry breaking process. This can be can be compared with explaining of the positioning of the chemical elements in the periodic table as a consequence of an underlying dynamical symmetry (which, in turn, are reflected in the electronic shell structure of atoms). Have a look at this recent paper to read more details about this fascinating perspective. Universal characteristics of symmetry breaking are even here, in the language of Nature.

Since the discovery of DNA huge progresses have been made in unveiling the genetic code, and the rate of discoveries in this field is accelerating day by day thanks also to the growing amounts of processed genetic data: a multi-disciplinary approach capable of integrating Mathematics, Physics, Biology, Informatics and Engineering could bring to a breakthrough, changing profoundly ICT horizons.

How can we model the dynamics of a human community, or a business ecosystem as a complex adaptive system ? I believe the basic understanding and tools needed for deal with these tasks have been already developed in other disciplines (such as Physics and Mathematics). Most probably what is required is a just a critical cross-fertilization and re-interpretation of achievements in order to make them applicable to another context (apparently far away). Let me make an example.

Symmetries in the Alhambra Palace, Granada

Think about symmetry, which is playing a key role in our understanding of Nature. Mathematically speaking, symmetry is characterized by the invariance of some mathematical object under some transformation. For example, a parabola y=x² is symmetrical with respect to the y-axis, since it is invariant under the transformation that takes the variable x and transforms it into –x.

In physics, mathematical symmetries imply conservation laws: for instance, translation invariance implies momentum conservation, while rotational invariance implies angular momentum conservation.

Symmetry breaking is an amazing phenomenon in Nature.

There are two types of symmetry breaking: explicitly and spontaneously. To understand the difference in simple terms, let us imagine that we are watching a group of people in a square downtown in Turin. People will be walking in random directions, with local aggregation patters. Now suppose that someone on the first floor of a building in the square starts to do something spectacular, then people will all look in the same direction, eventually moving in the direction of that building. This is an example of explicit symmetry-breaking: an action external to the behavior of people in the square makes all of them to behave in the same way. Coherency emerges as a result of the symmetry breaking. It’s the same principle of the laser!

Spontaneous symmetry-breaking is more subtle. Imagine that a single person, for example in the center of the square, walking randomly among the people, suddenly stops and starts simply to look up the sky in a very curious manner. When someone else notices this, stops and looks up. This induces others. Symmetry-breaking it emerges out of the interactions of the people in the square. It’s easy to detect similar dynamics in social communities, like Facebook for example. A trend-setter – someone who popularizes a new fashion – is creating a symmetry breaking.

In Physics, spontaneous symmetry breaking is an even more general principle at the basis of a vast number of physical phenomena, ranging for example from ferromagnetism to superconductivity. Also, in the context of the physics of elementary particles, spontaneous symmetry-breaking provides a mechanism by which the  masses of particles are generated.

So, it is amazing recognizing these universal characteristics of symmetry breaking: from classical to quantum physics, from biology to social dynamics analysis, to future Internet architectures (e.g. in IoT and IwT): whenever there is a sheers number of interacting  entities, from particles to objects to people. At the end of the day, a new fashion, or a new business could be seen as symmetry breaking events in markets fluctuating around steady states. Importance of understanding these principles is great.

Let’s go further with this analogy. Think about the Goldstone theorem. The theorem is arguing that massless particles (Goldstone bosons), generated at the spontaneous breakdown of symmetry, which are crossing the quantum system: capturing the nature of these emergent bosons, how many of them are generated in the symmetry breaking, and their dynamic propagation would allow us to understand the behavior of the system during the symmetry breakdown. One may ask “is there an equivalent of the Goldstone boson concept in an ecosystem , in a social community, or in a market” ? I suspect so: symmetry breakings are based on universal principles, so the mathematical laws behind their instances are very much the same!

In the last few years we’ve witnessed a progressing migration of “intelligence” towards the edges of the networks, i.e. towards the Users. Advance in processing and communication technologies (e.g.: higher and higher performance, miniaturization and cost reductions) is already bringing a proliferation of devices, embedding communications and computing, which , in the near future, will be more and more deployed in the environment we live.  In less than a decade, the edge of current infrastructures will become a distributed processing, storage and communication environment made of virtual resources (operated by a multitude of Players, not just Network Operators or OTTs). This transformation will create an ICT fabric capable of interconnecting people, machines and things, where services will be created and accessed through “everything”.

We are going beyond the Internet of Things or Machine to Machine as basically we are transforming “everything” in the ambient we live in a network node, by embedding not processing, storage and communications capabilities, but also a nervous behavior. These networks of networks of edge nodes will be connected to the traditional infrastructures so to expand traditional Telco-ICT networks towards the edge. A web of heterogeneous connections will capillary cover the ambient in which we live. 

We know from biology that spatially continuous networks with heterogeneous connections are ubiquitous in biological living systems, which naturally exhibit self-adaptation and self-control features, empowered by a capillary nervous system. Actually, in living organisms, body and nervous system are adapted to natural environments on many time scales, from evolution to development and learning: any individual organism brings is particular history of behavior and stimulation to any situation in which cognition is acted out. Considering the nervous system, each nervous cell is a very simple autonomous entity but, through the interactions of hundreds of billions of them, body is controlled and intelligence emerge.

The same will be for the future ICT networks: in this metaphor the body will be a dynamical set of physical networks and the nervous system will be a distributed overlay of cognitive software; both will adapt to the service environments, from evolution to development and learning: any dynamic network will bring is particular history of behavior and stimulation to any situation in which its artificial cognition is acted out. This network nervous system will be created by an overlay network of “nervous components” embedded into “everything”, from objects, edge nodes up to core nodes: you may see it as a very capillary management system of a highly pervasive network.

Imagine the impact of transforming each object into an entity that can communicate, that can allow you accessing every services and that is aware of the environment: it will create a huge number of new biz opportunities to be exploited. A Manufacturer or a Consumer electronics Provider would have the chance of transforming their products into means to provide services, and even to remain in contact with the User of the product.

If you think that I’m dreaming, have a look at this link, there is a nice example. The idea they have is creating mathematical models of the way the nervous system and the brain work, so to build products that behave more like animals.

 We are not that far from having a technology capable of developing an artificial nervous system for future ICT networks and services.

We too are active in this field, thanks to the EIT Activity “Smart Networks at the Edge”!

The rendering shows the optical communications within a silicon chip.

IBM has announced the discovery of a method to create optical pathways in silicon chip, effectively integrating optical communication in chips to replace electrical communications.

They have been able to assemble at sub 100 nm scale silicon and optical components side by side within the same chip using what is called silicon nano-photonics.

Silicon nano-photonics increases the speed in data exchange within the chip and can extend its communications to other chips on the same board or on different boards, decreasing connectivity cost and energy requirement.

The research was initiated in 2010 and now has reached maturity, letting researchers to create a CMOS based chip that includes WDM components for optical transmission, the same technology used on large capacity optical fibre for long distance communications.

The growing number of enterprises needing to process huge amount of data )as well as institutions) and to correlate data form different sources (Big Data) is a perfect match to this technology that will provide enormous capabilities to handle data within a chip and across chips distributed all over a network. As a matter of fact processing of data may take place seamlessly within the chip with different parts connected through optical communications or distributed over many chips potentially distributed and connected through an optical network.
This, really, is the future Cloud, at the level of the chip and providing capacity by connecting chips.