A comparison study of Weibull, normal and Boulevard distributions for wireless mesh networks considering different router replacement methods by a hybrid intelligent simulation system

The number of connected devices is increasing exponentially, calling for more and more dynamic networks that need little maintenance and are able to optimize end users experience in the process. Wireless mesh networks (WMNs) are viewed as a solution to keep end users satisfied with reliable connectivity while minimizing the maintenance and upfront costs. However, designing a robust WMN at low cost requires the use of the least possible mesh routers but still interconnected and able to offer full coverage. Therefore, the placement of mesh routers over the area of interest is a problem that entails thorough planning. To deal with this problem, we have previously implemented a simulation system that combines particle swarm optimization (PSO) and distributed genetic algorithm (DGA), in a hybrid intelligent system called WMN-PSODGA. In this work, we implement three distributions of mesh clients: Weibull, normal and Boulevard, and evaluate the performance of the network for different router replacement methods. The router replacement methods considered are constriction method, random inertia weight method, linearly decreasing inertia weight method (LDIWM), linearly decreasing Vmax method and rational decrement of Vmax method. By comparing all the simulated scenarios, we conclude that the best scenario in terms of both coverage and load balancing is normal distribution with LDIWM as a router replacement method.