Indoor Wireless Local Area Network (WLAN) : Measurement and Modeling from a user perspective

Abstract: If we had detailed wireless local area network (WLAN) coverage maps, both staff and equipment could be used more efficiently, for example, less time would be spent searching for connectivity. In addition, system administrators could understand their WLAN's utilization better, thus enabling better planning for where to install new access points, where to remove access points, where to change the type of antenna, etc. This thesis concerns creating detailed indoor coverage models by using measured network performance - in order to enable both users and administrators to visualize the network coverage. Today a user can only easily know about the access points that they currently hear – in the location where they currently are. Giving the users and administrators access to a model of the entire campus coverage will allow them to understand not only the _local_ coverage, but the patterns of coverage (or lack there of). However, no efficient modeling techniques are currently available for those deploying and operating indoor WLANs. The thesis begins with some general background information and then examines a number of WLAN survey tools; in terms of both their performance and cost. Following this a number of related projects are presented. This background provides the motivation for why a new tool is needed and what functions such a tool should have. Next a site survey of the KTH campus in Kista was conducted using a newly developed survey application. This application was developed to better meet the requirements derived from the missing functionality of existing tools. In addition, developing this application gave the author an opportunity to learn a new objectoriented programming language, i.e. C# and the .NET environment. Learning to use this new environment was essential to building both an easy to use application and collecting the data from the system - the later was often not straight-forward. ne of the key issues after data collection is how to present the collected data to the user and how this varies depending upon the user’s interests and task. An initial representation of the experimental data is presented as a manually painted coverage map overlayed onto a map. Next the thesis examines how to integrate the experimental data using Google’s SketchUp in order to build a 3D model of the WLAN coverage on this campus. Future work related to this thesis should focus on how to automate the collection of data and how to automate the presentation of the resulting experimental data.