Adaptive Memory Differential Evolutionary Algorithm for Network Management Using Variants of Dominating Set Models

In wireless sensor networks, where multiple sensors are typically concentrated in a confined area, determining the optimal size of wireless sensors to use for communication and coordination over the network is essential due to concerns about the preservation of energy and battery. In this paper, we introduce new competitive memory-based differential evolution methods to solve wireless sensor network management problems using variants of dominating set models. The problems considered are formulated as constrained optimization problems that are transformed into unconstrained optimization problems using the penalty principle. Then, differential evolution with several memory-based enhanced procedures is designed to find solutions to the reformed optimization problems. Two versions of the proposed Adaptive Memory Differential Evolutionary Algorithm (AMDEA) are designed to deal with the minimum dominating set problem and its variants. The results obtained from the proposed methods show that they are promising compared to some benchmark methods.