Affiliation:
1. University of Waterloo
Abstract
A commonly used technique for evaluating and comparing the performance of systems using 802.11 (WiFi) networks is to conduct experiments. This approach is appealing and important because it inherently captures critical properties of wireless signal transmission that are difficult to analytically model and simulate. Unfortunately, obtaining consistent and statistically meaningful empirical results using 802.11 networks, even in well-controlled environments, can be quite challenging and time consuming because channel conditions can vary over time.
In this paper, we use 2.4 and 5 GHz 802.11n MIMO networks to study different methodologies that could be used to evaluate and compare the performance of different alternatives used in 802.11 systems (e.g., different systems, configurations or algorithms). We first illustrate that some of the more commonly used methods in existing research are flawed and explain why. We then describe a methodology called multiple interleaved trials that, to our knowledge, has not been used for, or studied on, 802.11 networks. We evaluate this methodology and find that it can be used to repeat experiments and to compare the performance of different alternatives. Finally, we discuss other possible applications of this approach for comparative performance evaluations.
Publisher
Association for Computing Machinery (ACM)
Cited by
9 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Reproducibility in Computing Research: An Empirical Study;IEEE Access;2022
2. Demystifying frame aggregation in 802.11 networks: Understanding and approximating optimality;Computer Communications;2021-12
3. NeuRA;Proceedings of the 23rd International ACM Conference on Modeling, Analysis and Simulation of Wireless and Mobile Systems;2020-11-16
4. PNOFA;Proceedings of the 23rd International ACM Conference on Modeling, Analysis and Simulation of Wireless and Mobile Systems;2020-11-16
5. Reproducibility in Scientific Computing;ACM Computing Surveys;2019-05-31