Techniques for Improving the Accuracy of 802.11 WLAN-Based Networking Experimentation

Abstract

Wireless networking experimentation research has become highly popular due to both the frequent mismatch between theory and practice and the widespread availability of low-cost WLAN cards. However, current WLAN solutions present a series of performance issues, sometimes difficult to predict in advance, that may compromise the validity of the results gathered. This paper surveys recent literature dealing with such issues and draws attention on the negative results of starting experimental research without properly understanding the tools that are going to be used. Furthermore, the paper details how a conscious assessment strategy can prevent placing wrong assumptions on the hardware. Indeed, there are numerous techniques that have been described throughout the literature that can be used to obtain a deeper understanding of the solutions that have been adopted. The paper surveys these techniques and classifies them in order to provide a handful reference for building experimental setups from which accurate measurements may be obtained.