Complex adaptive communication network

Complex adaptive system is a means of classifying special type of systems which consist of a large number of components interacting nonlinearly. There are two key factors which imply that a system would be a cas:
#
# The system is made up of a large number of components
# The components interact nonlinearly
#

Modern Communication Networks such as Wireless Sensor Networks, Peer-to-peer, Swarm robotics and multiagent systems are gradually growing in size and complexity. Although there exist a set of modeling and simulation tools which assist in developing models of most small-scale systems, it can get extremely difficult to develop models of very large scale communication networks such as internet scale systems. As such, they are nowadays commonly considered to behave as and are modeled in a similar manner to Complex Adaptive Systems. These communication networks thus require special modeling and simulation skills to comprehend emergent behavior and other attributes.
Background
Although the domain has not been named previously, it has actually been emerging over a period of several years for now. Complex Adaptive Communication Network and Environments (CACOONS) is thus a modeling approach which allows the unification of various types of communication networks based on their similarity to the complex adaptive systems approach common in complexity theory. Examples include the examination of complex network approaches to cascading power failures such as . Other examples include . Similar approaches to networks have been demonstrated in other systems such as and . The web has been regarded recently as a complex adaptive system in
Problems associated with CACOONS
Because of the large scale nature of modern-day CACOONS, traditional means of modeling and simulation fail to model situations such as the Amazon.com web services Cloud Failure. Other examples include the effects of modeling torrent traffic on the internet.
Examples of CACOONS
A recent special issue is being proposed in the SAGE Journal "Simulation". Original ideas of self-organization can be found in the book .
Dressler first describe the various concepts of complex systems. Then he correlated these self-organization concepts and associates them with Wireless Sensor and Actuator Networks. He takes all commonly well-known WSN algorithms and correlates them with self-organization in general and self-* in particular.
In , Beal et al. from MIT present CRF-Gradient, a self-healing gradient algorithm that provably reconfigures in O(diameter) time. Beal et al. in subsequently extend their previous work and propose Flex-Gradient, a new self-healing gradient algorithm which has a tunable trade-off between precision and communication cost.
While there are articles on these topics taken separately, CACOONS is basically an area which serves to assist modeling and simulation of these complex communication networks (as opposed to complex networks in general, which can even be correlated with Biological networks or Social Network analysis).
Another common modeling approach in complex systems is the agent-based modeling approach. Again, while there are Wiki pages dedicated only to agent-based modeling, the specific telecommunication aspect of using agent-based modeling is lacking in general. As a result, it is apparent on mailing lists such as for agent-based modelers (e.g. netlogo-users) that a large number of interested people need ready reference for specifically agent-based modeling in the domain of communication networks.

Fault Tolerance
Fault tolerant computer systems have, at times, been considered to be complex adaptive systems. The basic idea is to develop a joint industry-government initiative for developing mathematical and practical tools for improving the security, performance, reliability and robustness of energy, financial, telecommunications and transportation networks.
A related area is Autonomic Computing which basically differs from CACOONS with its focus only on the self-* aspects of the network components. Whereas CACOONS approach is open to any new ideas that might emerge in the domain of complex adaptive systems, so that they may be applicable for use by practitioners in this domain. In addition, another key difference is the focus on developing simulation and network models in this domain for detecting emergent behavior specifically in contrast to Autonomic Computing, which is focused more on engineering the actual systems.

Main Menu

See Also

Main Menu