The impact of directional antenna models on simulation accuracy

Increasingly, directional antennas are being used in wireless networks. Such antennas can improve the quality of individual links and decrease overall interference. However, the interaction of environmental effects with signal directionality is not well understood. We observe that state of the art simulators make simplifying assumptions which are often unrealistic and can give a misleading picture of application layer performance. Because simulators are often used for prototyping and validating new ideas, their realism and accuracy are of primary importance. In this paper, we apply a new empirical simulation method for directional antennas and study how well this models reality. We show that not only is our model easy to implement, but is also more accurate and thus better able to predict the performance of propagation-sensitive applications.