Dynamic Load Impact on Protocols in Mesh: An ANOVA Test Evaluation

Abstract

The presented paper offers an in-depth look at mesh routing protocols, focusing on how well they perform under different conditions of node density and mobility. To make accurate comparisons, the study uses robust non-parametric statistical methods, including the Kruskal-Wallis, Mann-Whitney, correlation and covariance tests, to pinpoint which protocols stand out across crucial Quality of Service (QoS) metrics. These metrics cover energy consumption, end-to-end delay, packet delivery ratio, throughput, and network overhead. The research simulates several mesh network scenarios to collect important data for a comprehensive assessment of the reliability and effectiveness of each protocol. The Kruskal-Wallis test identifies large differences in performance across different circumstances. Following that, the Mann-Whitney test offers a closer look, comparing protocols in detail to identify those that work best under specific conditions. The study finds that protocol performance can vary dramatically depending on network density and how mobile the nodes are. It outlines which protocols excel in more stable setups and which remain dependable in fast-changing, dynamic networks. For network designers and operators, these insights are incredibly valuable, offering practical guidance on optimizing performance in real-world deployments. Furthermore, the research underscores why non-parametric statistical methods are so crucial. Network data often show high variability and don't always follow normal distribution patterns, making traditional parametric techniques less reliable. By choosing a more flexible and accurate approach, this study makes a meaningful contribution to wireless communication, providing a clearer picture of how different routing protocols hold up under real-world conditions.