The development of factory noise prediction models over the past thirty years has parallelled the development of computer technology. There are currently two main approaches to this problem both of which have been extensively investigated, they are the image-source method and the ray-tracing technique. Presented are brief details of the background to the specific problems of factory noise prediction together with the various approaches used to solve the problem. An outline of how the two types of mathematical models work, together with details of their representational development is presented. A comparison of the current potential of each type of model, enabling an insight into when and where each type of model may be effectively used, is given. Three independent studies comparing various prediction models were considered. All three reviews drew the conclusion that the Ondet and Barbry, and the Lindqvist models were the most accurate ray-tracing and image-source models, respectively. Finally, a review of barrier prediction models is presented.
Identification of image source using serial-number-based watermarking under Compressive Sensing conditions