Posts Tagged ‘Self-Management’

Neurotransmitters for … Future Networks (2 of 2)

Monday, February 27th, 2012 by Antonio Manzalini

Computing is becoming low cost and pervasive, embedded in any node, Users’ device, everyday object and in the physical environment (sensors, actuators, etc) as well. An open and dynamic networked world will be the arena of next services and applications. Traditional control and management approaches will be ill-suited to face such environments: the question is how effectively exploiting coordination in huge ensembles of distributed autonomous entities (against strict requirements such as dynamism and complexity). Forget also traditional middleware: communication and computation costs would be too high, and solutions brittle and fragile. Feasible approaches to control and manage myriad of interacting nodes and devices are still unknown…but do we really need them? Let’s change the perspective: did Nature invent a way to control the behavior of any single neuron ? Not indeed! Part 1 of this post (drawing inspiration from neurotransmitter functioning in neuron networks) has proposed a vision of future networks where each node, device, machine, smart object (like a neuron) is capable of autonomous local self-adaptation reactions to the context (through neurons’ interconnections) and where a global harmonization is made through viral propagation of coordination and context information (as neurotransmitters do). This Part 2 is proposing a simple proof-of-concept aiming at demonstrating that this is feasible even today.

Imagine Users (cars, kiosks, lamp streets…) having a sort of communication halo around them: Edge Networks (see a previous post) will emerge spontaneously (as flocks of birds flying around) through these halos overlapping cross-interactions. Imagine each node having perception of its local context, the environment inside its halo. Each node diffuses its context information hop-by-hop accordingly to certain propagation rules. Any context can be accessed locally but at the same time it takes account of the influences of the context and coordination information propagated from other nodes (also the fixed ones). We’ll have a sort of global coordination field, injected by nodes in the network and autonomously propagating … like neutransmitters. In other words, nodes are interacting with each other and with the environment by simply generating, receiving and propagating distributed data structures abstracting context information. This field is providing nodes with a global representation of the situation of the overall network, which is immediately usable, like an object moving in a “gravitational” field. Environmental dynamics and nodes local decisions will determine changes in the field closing a feedback cycle. This is enabling a distributed the overall self-organization.

In real proof-of-concept nodes’ halos can be easily implemented with a smart phone (acting as Wi-Fi Hot Spot), one (or more) cheap, tiny PC (e.g. a Raspberry Pi for $ 25) and one (or more) microcontroller (e.g. based on Arduino). Coordination field can be made of “tuples” of data which can be injected and diffused by each node. Local reading of these “tuples of data” (e.g. through pattern matching) can trigger local self-adaptation behaviors. Plenty of open source applications are available on the web to implement nodes primitives and local autonomic behaviors. It’s simpler than expected.

Future Networks: local reactions and global self-organization

Surprisingly, if we look at the network dynamics as a “many body problem” we can even define the Hamiltonian (a sort of energy function describing the state) of the network. Just following Nature.

I’ll elaborate that in a next post.

Edge Networks “at the Edge of Chaos”

Wednesday, July 27th, 2011 by Antonio Manzalini

It is widely shared vision that in about five-ten years networks will become like complex adaptive system of systems interconnecting an incredible number of heterogeneous nodes, devices, machines, smart things and objects. Even today we are witnessing the progressive migration of “intelligence” towards the edges of the network, where actually Users’ devices are becoming more and more powerful, literally similar to network nodes. Technologies advances in virtualization are already allowing multiple virtual networks to run simultaneously over a same physical infrastructure, decoupling network functionalities from the physical fabric.

Then, as like large scale complex adaptive system of systems, future networks might eventually exhibit behaviour at “edge of chaos”. This term was made popular by Stuart Kauffman and refers to the idea that many complex adaptive systems (including brain) seem to naturally evolve towards a regime that is delicately poised between order and chaos.

Example of chaotic behavior

Basically it means that pervasive future networks will be naturally robust and resilient against attacks, but on the other hand, their dynamics might be influenced by phase transitions (i.e. an abrupt change in some operating characteristics may take place with a relatively small variation of certain parameters). For example, this phenomenon would mean a dependence of the network’s global characteristics (e.g., connectivity and average delay rate) on some local parameters (e.g., communication path and transmission probability). This might bring to instabilities, but at the same time chaotic behaviour can be used to exploit several opportunities (I’ll drop a next post on this).

My take is that at this level of complexity management of future edge networks will becom highly challenging (not feasible with the traditional approaches, even if these (maybe) will be still applicable at the core). Should an Operator wish to play a “game” in this arena, it should start thinking about “edge network self-management”, a domain of investigations which is at the intersection between automatic control theory and non-linear dynamics, the ideal place for exploiting the enormous potential of the Chaos Theory.

“Strategy is the art of making use of time and space”

Wednesday, July 13th, 2011 by Antonio Manzalini

I’ve found this well-known quote at the beginning of a nice paper “A New Paradigm for Dynamical Modelling of Networked C2 Processes” that I’ve recently read. The paper (from Defence Science & Technology Organisation, part of Australia’s Department of Defence) proposes a new type of mathematical model for studying synchronisation of interacting Command and Control (C2) processes (for example planning, execution of strategic, campaign or tactical activities, actually control-loops) across complex networks. This problem (consensus reaching in interacting C2 processes) seems to me to have several analogies with the problem of coordination of interacting (sub-) networks in self-managed complex networks. Actually, in future networks, embedded communications will determine the same emergence of dynamic games of (sub-)networks (owning to the same, or different Operators), particularly at the edges, which will be able to support any sort of services by leveraging local processing and storage capacities

Interestingly, the approach here distinguishes the time-scales of and interactions (time and space) between individual C2 processes and makes use of the Kuramoto model to simulate the essence of Command and Control dynamics. The main message is that, at the end of the day, we are talking about networks, and nodes of these networks are entities – modelled by Kuramoto as oscillators (see a previous post on Kuramoto) – which undertake a cyclic change of state, or loop. Well, this applies perfectly to future networks, where self-management introduces interacting control-loops, for example for self-configuration. Following the metaphor, the rate of progress of one oscillator (i.e. a node through a self-configuration control-loop) at a particular node depends on the point in the control-loop of another oscillator at a connected node (which is the coupling of controllers). And noise has an important impact in the local to global network dynamics.

In the same avenue, I’ve been even more surprised when I saw capturing said modelling similarities also in this paper simulating the responses of a network of neurons as the product of coupling among hierarchies of neuronal populations. In its essence, the problem is very much the same.

Schematic of interacting neuronal sub-populations (Harrison L et al. Phil. Trans. R. Soc. B 2005;360:1075-1091)

What I’m doing now is using the Kuramoto model for simulating two (sub-) networks (red and blue) aiming at reaching internal consensus whilst competing with each other. Amazing behaviours emerge (I’ll show you some results in a next post). There are plenty of key questions which need to be addressed both at local and global context, but what I’ve found impressive is that certain phenomenon like synchronization and cross-coupling appear to be so ubiquitous in Nature to reveal its mathematical nature !

Autonomic Networks and Epidemic Information

Tuesday, December 29th, 2009 by Antonio Manzalini

 Autonomic and Epidemic are two terms widely used in biology. It is really intriguing to realize that biological processes can provide several ideas for the design of distributed, self-adaptive, robust and scalable networks. An interesting area of study, for instance, is the application of epidemic information dissemination to the communication and cooperation of autonomic network elements pervasively distributed in dynamic environments: for instance, autonomic management and epidemic communications could enable the emergence of global properties from simple local behavior and interaction among independent nodes and devices.

 

There is already a rich prior-art showing that autonomic management of domains of collaborating network nodes and devices can provide efficient and cost effective network access and optimisation. Key principle of autonomic management is decentralization. Several components (e.g., hosted in network elements) detect each others and then start to coordinate their actions thus increasing management efficiency and effectiveness. Network elements can be considered, for example, as full-fledged routers for their delegated IP subnet, able to operate stand-alone.  Central feature is the exchange of network information or knowledge (e.g. that can be global, local, private), among the network elements, supporting the autonomic management processes.

 

Then, for scalability issues, it is advisable to have different kinds of information exchange. Network information that is important on a global dimension (e.g., system-wide parameters) should be disseminated throughout the network using, for example, epidemic communication mechanisms. On the other hand, local information (e.g., radio frequencies, transmit power or link utilization) should be only disseminated locally among the affected neighbouring network elements. Private information (e.g., logs) should probably never be disseminated.

 

What is interesting observing is that simulations of epidemic information dissemination show that the time to disseminate the global state in an autonomic network domain does not grow significantly with the number of network elements. On the other hand, it grows proportional to the topology diameter. This means that the scalability of the autonomic networks with epidemic information dissemination depends mainly on the topology network structure, having this more impact on management performance than the number of network elements.