Hybrid Approach of Genetic Algorithm and Differential Evolution in WSN Localization

Abstract

Sensor node localization refers to the knowledge of position information and is a procedural technique for estimating sensor node location. In wireless sensor networks, localization refers to the estimation of sensor node location information. Optimization algorithms are used to determine the position of sensor nodes. Traditional algorithms rely on analytical methods, which increase in computational complexity as the number of sensor nodes grows. Due to resource constraints, cost constraints, and sensor node energy constraints, an algorithm with reduced computational complexity is needed, one that does not need external hardware, needs less run time and memory, is scalable and easy to implement without losing performance, and has improved location estimation accuracy with better convergence. The proposed research work uses Genetic algorithm and differential evolution algorithm are combined in the hybrid Genetic Algorithm - Differential Evolution localization algorithm. Differential Evolution has a higher mutational rate, whereas the Genetic Algorithm has a better range and crossover operator. In wireless sensor networks, the Hybrid Genetic Algorithm - Differential Evolution uses the selection with crossover operators of Genetic Algorithms and the mutation operator of Differential Evolution to approximate the positioning information of sensor nodes.