Scalability Analysis: Reconfiguration of Overlay Networks Using Nature-Inspired Algorithms

Overlay networks are virtual networks of nodes and logical links built on top of the existing network infrastructure,with the purpose of contributing new functionality. There are many different solutions proposed to tackle a range of specific needs such as content distribution and caching,file sharing, improved routing,multicast and streaming, ordered message delivery, and enhanced security and privacy. In this chapter, the focus lies on the optimization of overlay networks in terms of cost, performance, and reliability. In particular, the main objective is the optimization of data mirroring. Three different optimization approaches are introduced. The first approach is based on a ”related work” implementation using Genetic algorithms, the second makes use of artificial immune systems, and the third approach uses the Particle swarm optimization approach. The three algorithms are implemented and experiments are conducted to measure the overall performance, the behavior and feasibility of network and link failures, as well as a scalability analysis is performed.