Wireless Mesh Network (WMNs) offer a cost-effective and scalable wireless communication solution but finding the optimal mesh router placement remains an NP-hard problem. To deal with this issue, in our previous work, we implemented a hybrid intelligent system based on Particle Swarm Optimization (PSO), Hill Climbing (HC), and a Distributed Genetic Algorithm (DGA). In Genetic Algorithms (GAs), Blend Crossover (BLX \(\alpha \) ) is widely used method because of the simplicity and exploration capability. However, BLX \(\alpha \) generates each gene independently which may lose correlation between genes. In this paper, we implement in our simulation system Parallelotope-Shaped Blend Crossover (psBLX), which extends BLX \(\alpha \) by sampling offsprings from a parallelotope-shaped region. We compare the performance of psBLX with BLX \(\alpha \) using two router replacement methods: Random Inertia Weight Method (RIWM) and Fast Convergence Rational Decreasing \(V_{max}\) Method (FC-RDVM). The simulation results show that psBLX achieves improved performance compared to BLX \(\alpha \) .