Hybrid Harris Hawks With Sine Cosine for Optimal Node Placement and Congestion Reduction in an Industrial Wireless Mesh Network

Abstract

The optimal performance of a wireless mesh network (WMN) can be greatly improved by strategically placing wireless mesh routers. As a result, it is crucial to optimally locate the WMN routers for better coverage and connectivity. Besides the optimal placement, the network congestion due to overlaying routers has to be taken into consideration. These issues have become a motivation for researchers to identify a variety of approaches to optimize WMN performance. Multiple metaheuristic algorithms have been employed for identifying the trade-offs between coverage and connectivity in WMN. Consequently, a novel hybrid Harris Hawks optimization with the sine cosine algorithm (HHOSCA) is presented in this work to tackle the aforementioned WMN optimization problems. The proposed HHOSCA seeks optimal router placement that leads to significantly increased network coverage and achieves full connectivity between the mesh routers. In addition, the proposed HHOSCA produces a cost-effective WMN by reducing the congestion in the network to the minimum number of routers whilst ensuring maximum coverage and connectivity. The superiority of the proposed HHOSCA in comparison to the other algorithm was validated by using 33 benchmark functions. It was compared against four well-known algorithms including Sine Cosine Algorithm (SCA), Harris Hawks optimization (HHO), Gray Wolf Optimization (GWO), and Particle Swarm Optimization (PSO). These algorithms are statistically analyzed and compared to the simulated results of the proposed method. In addition, the performance of HHOSCA is compared to the state-of-the-art to highlight the efficacy of the proposed algorithm. The statistical analyses and simulation findings confirm that the HHOSCA outperforms the other algorithms in terms of network connectivity, coverage, network reduction, and convergence. The experimental results reveal that the proposed HHOSCA method achieves favourable optimization results compared with other relevant methods.