Assessing Combinatorial Design for Analyzing System Performance of a Computer Network

TL;DR

This paper demonstrates how combinatorial design can efficiently generate test cases to analyze the impact of various factors on computer network performance, reducing experimental complexity while ensuring comprehensive coverage.

Contribution

It introduces the application of combinatorial design methods for network performance analysis, enabling cost-effective and thorough testing of multiple factors affecting communication systems.

Findings

Effective factor analysis on network performance

Reduced number of experiments needed

Validated approach for communication network testing

Abstract

Generally, combinatorial design concerns with the arrangement of a finite set of elements into patterns (subsets, words, arrays) according to specified rules. The usefulness of this design method is that the number of input combination can be reduced dramatically but all of them are covered by the combinatorial set. This paper presents the application of this design method in communication networks. Communication engineers can use this novel method to generate test cases for producing a cost-effective set of experiments to recognize the factors that have the least and most impact on the system's performance. A well-known scenario is used for the purpose of the experiment and five factors with different values are chosen to qualify their effect on the network performance. The experimental set is generated using combinatorial design method and then it is been used to analyze the effect of each factor. The experiments showed the effectiveness of the method to be used for analyzing the effect of factors on the communication network.