A Field Study of WiMAX Performance

Abstract

WiMAX is a wireless access system that offers fixed, nomadic, portable and soon mobile wireless broadband connectivity. Much hype is spread about system performance and loads of contrary statements are observed. This thesis aims at discovering the real life performance of WiMAX, and will concentrate on both the fixed and mobile WiMAX profiles. Scientific research methods are used to study performance in a fixed WiMAX field trial, a fixed WiMAX deployment and a pre-mobile WiMAX field trial. The performance study is focused on throughput and coverage, where the affecting underlying parameters and relations among them are analyzed. All the measurement results and analyses are presented, and a detailed overview of the two profiles fixed- and mobile WiMAX is given.

WiMAX performance showed to be close to the vendor specifications and the theoretical limits under perfect conditions, but deviated significant in imperfect scenarios, which is the normal case. Maximum throughput obtained was 76% compared to the theoretical limitations, where factors as protocol overhead caused the deviations. Typical imperfect scenarios may be at long distances, without clear sight to the base station in a possible overloaded cell with radio interference. Throughput and coverage was therefore found to be dependent variables in that they rely on underlying independent parameters and the relations between them. For instance, both throughput and coverage depends on the attribute received signal strength, which again depends on the sight capabilities. Analytical expressions and models were derived based on measured field data when determining and modelling WiMAX performance under imperfect conditions. Specific derived expressions determine throughput performance and path loss models determine the expected coverage. An analytical expression that reveals co-channel interference is derived together with a range of expressions that model the gain in performance for different attributes when various diversity orders are used.

To satisfy all the hype and expectations about WiMAX performance, deployment conditions must be perfect with line of sight at longer distances, no co-channel interference and most of the various optional system parameters must be optimal. Because WiMAX performance depends on a range of varying factors, analytical expressions and models should be used to determine the real life performance.