Parametric Dimensioning and Enhancements of a Path Prediction Algorithm
Miltiadis Kyriakakos, Stathes P. Hadjiefthymiades, Lazaros F. Merakos · 2007
To meet the growing requirements for resource management schemes in the context of mobile and wireless computing, path prediction algorithms are gaining more attention and are being extensively examined. Path prediction allows the network to enhance the user quality of service. A parametric dimensioning of a path prediction algorithm based on learning automata is presented. The algorithmic parameter values are optimised for maximum prediction accuracy. Furthermore, two add-on mechanisms of the path prediction algorithm are introduced, and their performance is evaluated through simulations.