Abstract
The term “autonomic networking” refers to a recently emerged communications paradigm, that uses distributed techniques (swarm intelligence based methods) and distributed hash tables to implement traditional functionalities of networking, including among others routing, service and resource discovery, addressing, load balancing etc. The shift of interest towards the autonomic networking approach has been mainly motivated by the pervasion of novel algorithms into relevant fields of computer communication science. These innovative techniques have been greatly alleviated by their fusion with already established solutions and applications (i.e. agent-based systems, network middleware development etc) so as to form a new ever-evolving landscape in networking.
Key Terms in this Chapter
Peer to Peer (P2P): Networking concept that contradicts the traditional client-server model by assigning to participating nodes equal roles. P2P based applications include among others file sharing systems, streaming platforms, and so forth.
Distributed Hash Tables (DHT): Distributed platforms that assigns keys (usually representing real life objects such as files) onto nodes of a network, providing the basic infrastructure for the implementation of basic networking functionalities such as routing and resource discovery
Overlay Networks: Networks created by creating virtual interconnections between nodes, on top of the physical structure of the original network.
Epidemic Protocols: Protocols that imitate the spread of a disease among a (usually homogeneous) population in order to provide the algorithmic substructure for reliable information dissemination in large and transient groups of users processes
Ant Colony Optimization (ACO): Recently emerged optimization algorithm that is inspired by techniques used by swarms of ants. At each iteration of an ACO algorithm, each agent (ant) of the swarm builds a solution of the problem based on both it is own estimation and information the indirect form of communication (stigmergy) he attains with the other agents, through an artificial hormone (pheromone).
Evolutionary Algorithms: Optimization algorithms that imitate processes of life such as mutation, selection, and reproduction. Genetic algorithms, evolutionary programming, evolution strategies, classifier systems, and genetic programming are considered part of this category of optimization techniques.
Ad-Hoc Network: A term characterizing mostly wireless networks in which nodes become members just during the process of a specific communication session, or while they are in the physical range of the network.