Open Access
Article
SciPap-1918
Dynamic Load Impact on Protocols in mesh: An ANOVA Test Evaluation
by
Ibrahim Alameri 1,* 
, Jitka Komárková 2,* and Tawfik Al-Hadhrami 3
1 computer Science, Jabir ibn hayyan University For Medical and pharmaceutical Sciences, Alkufa, Najaf 54001, Iraq
2 Faculty of Economics and Administration, Institute of System Engineering and Informatics, University of Pardubice, Studenská 84, Pardubice 53210, Czechia
3 Department of Computer Science, Nottingham Trent university, 50 Shakespeare Street, Nottingham NG118NS, United Kingdom of Great Britain and Northern Ireland
* Authors to whom correspondence should be addressed.
Abstract: This paper takes a deep dive into mesh routing protocols, unraveling how they hold up under the pressures of varying node densities and the hustle and bustle of mobility. This paper included robust and advanced non-parametric statistical tests—think Kruskal-Wallis and Mann-Whitney—to figure out which routing protocol stands out in terms of Quality of Service (QoS) metrics for instance how long it takes a packet to travel end-to-end, the ratio of packets delivered successfully, throughput, and the amount of network overhead. Kicking things off, this study simulated a bunch of mesh environments. It was like setting up different conditions for the protocols to see how well they affect under various conditions. This part was crucial—it gave the work the raw data to put these protocols through their paces. With the Kruskal-Wallis test, this study aims to look for significant differences in how the protocols are performed across different scenarios. Moreover, when it came to the core, the Mann-Whitney test helped us conduct some head-to-head comparisons to spot the top performer under specific conditions. The performance of these protocols can vary wildly depending on how crowded the network is or how fast nodes are moving. The current investigation highlighted which protocols can keep optimal results when things get stable, and which can stay tough when the network feels more like a rollercoaster. This kind of investigation is gold for network designers and operators. Furthermore, this study brings significant attention to using non-parametric statistical methods for this analysis. The usual network parametric techniques often assume everything is normal (statistically) and that the variance is consistent across the board.
Keywords: Mesh, Aodv, Kruskal – Wallis, One Way Anova, And Mann-Whitney.
JEL classification: C15 - Statistical Simulation Methods: General, D85 - Network Formation and Analysis: Theory, L14 - Transactional Relationships • Contracts and Reputation • Networks, L63 - Microelectronics • Computers • Communications Equipment, L86 - Information and Internet Services • Computer Software
Received: 1 March 2024 / Revised: 21 November 2024 / Accepted: 5 December 2024 / Published: 1 January 2025
