Bandwidth for internet traffic can be utilized more effectively with queueing models. Pascal Lieshout researched this at the Centrum Wiskunde & Informatica in Amsterdam. He took his PhD on Friday 5 September 2008 at the Universiteit van Amsterdam with his thesis Queueing Models for Bandwidth-Sharing Disciplines.
Modern communication networks like the internet are able to simultaneously support several applications, like data (e-mail), speech (VoIP), and video applications (YouTube). The network traffic that is generated places different demands on the handling. In data traffic only a few details may be missed. In speech, however, some noise (loss of data) is not too bad but speed reduction is.
As the different kinds of traffic have different requirements, it is important to distinguish the bandwidth allocated to each traffic class. Lieshout examined methods to achieve this. To map their performance he applied queueing theory. With the results network providers can utilize their capacity in so-called multi-service networks better, and save money. In 2006 Lieshout received the prestigious ACM Sigmetrics/Performance best student paper award for his research.
