Aplicação de algoritmos genéticos na configuração de parâmetros em técnica bioinspirada para balanceamento de carga em RSSF

Abstract

In reactive Wireless Sensor Networks (WSNs) the necessary redundancy to cover the area of interest that the sensor configuration requires makes it possible for more than one sensor to detect the same event in a same area at the same time. This redundancy motivates the adoption of static and dynamic coordination techniques to achieve a balancing the processing of events by the nodes in order to increase the network lifetime. This work deals with the development of an automated parameter selection strategy in a load-balanced algorithm for bioinspired WSN, called Ant-based. A genetic algorithm was developed and an acceleration technique of this algorithm was implemented. The results showed that for networks in which the nodes are randomly distributed in the space of interest, it was possible to obtain more efficient parameters. The new algorithm configuration parameters obtained better performance than the original parameters for a strongly connected network of 7.4% and in a weakly connected network of 4.7%.