|
In this technique, we apply multi-objective optimization to the problem at hand.
Specifically, we use an existing algorithm called NSGA-II [13]. Multi-objective
optimization is one of several techniques in evolutionary computing. Evolutionary
computing is the branch of science that takes randomness as amean of problem
solving; it also considers solutions of the problem as chromosomes. Mating
between different chromosomes could yield a better breed or better solutions.
Using evolutionary computing techniques is very helpful in situation where the
search space of a problem is huge; searching this huge space in sequential search
techniques takes exponential times. Small business network configuration jumps in the search
space in such away that explores areas in which a potential good solution can
be found. Many of evolutionary computing techniques rely on operators such
as crossover operator which is used for mating between chromosomes, mutation
operator which is used to alter genes of the chromosomes, parent selection
operators which is responsible of choosing chromosomes for mating. Doing the
mating process continues over and over until specific conditions are met such as
accuracy of solution or no change over the best solution.
|
