Effectiveness of modified evolution algorithm to improve network lifetime in WSN

Abstract

Enhancing network lifetime is the primary goal of wireless sensor network researchers. Among the various options for reducing energy consumption,

the energy invested in steering and group head selection is one of the most effective mechanisms. Both tasks were deemed too difficult and the effective solution difficult to obtain. Since it is difficult for traditional methods to meet the requirements and difficulties, the solution according to meta-algorithms focusing on natural arithmetic methods has provided simplicity. The work proposes to face these challenges by using computational intelligence, especially DE Differential Evolution and GA Genetic Algorithm. An energy-efficient method discovery was designed for dynamic network with variations in DE, and fast and adaptive methods for networks undergoing changes were designed. Knowledge-based DE is designed drawing on prior knowledge to redefine new ways to change the network. The strategy of hybrid mutagenesis under standard DE is designed to select the head of the cluster that provides faster convergence characteristics. The proposed solutions were implemented under the MATLAB environment and the results showed that the proposed solutions work best for different network configurations. Dynamic path detection using KDE achieved energy savings of 9.83 to 49.2 percent compared to RDE and 6.7 to 29.5 percent compared to PDE. The energy savings achieved in group head selection with the proposed HMDE are 10 to 33 percent better compared to dyPSO and 5 to 10 percent better compared to SDE.