بهینه سازی ژنتیکی الگوریتم خوشه بندی ترکیبی در شبکه های حسگر بی سیم موبایل Genetic optimization of hybrid clustering algorithm in mobile wireless sensor networks

1. Introduction Genetic algorithm (GA) was initially suggested as a search algorithm. GA is also known as a global heuristic algorithm. A GA estimates an optimal elucidation through generating different individuals. Focused fitness function is one of the main procedures of GA. It is taken from natural evolution of new species in natural environment. The natural assessment has the accompanying features:
the individual characteristics are implied on a chromosome;
every chromosome has a specific fitness function value as indicated by the location in which it is present;
individual chromosomes with best fitness value can survive and produce next generations of better individual chromosomes. In this work, Base Station (BS) is a centralized, high-capacity node capable of co-coordinating the entire communication among nodes in an environment (Gherbi et al., 2017). GAbased cluster heads (CHs) selection is proposed, for better CHselections. On every round, new populations of CHs are selected by BS on evaluating fitness function. In dynamic clustering, i.e. re-clustering, it is significant to avoid early death of CH s. Hence, such efficient CHs can be obtained by applying selection, crossover and mutation processes. This proposed centralized approach is projected to overcome the main limitations of Low Energy Adaptive Clustering Hierarchy (LEACH) protocol (Heinzelman et al., 2002). In our proposed GA, binary representations are used to represent each sensor node bit by bit. The representation of a node in mobile wireless sensor network (MWSN) is called a Chromosome or Genome, a collection of bits. In this network, cluster members (CMs) are represented as “0” and CHs are represented as “1”. Initially, GA starts with any initial random population, a predefined number of chromosomes; each chromosome consists of a potential solution. Then, GA calculates fitness value of each chromosome by using fitness function. After evaluation, GA selects the best fit chromosomes from the population, by using a specific selection method based on their fitness values and then applies crossover and mutation operator, respectively. These procedures are repeated iteratively to obtain a new population better than the previous one.