Quality of Service (QoS) Routing for Connection Bundles Using Stochastic Knapsack Framework

Abstract

Today's Internet carries different class of traffic which has a distinct set of Quality of Service (QoS) constraints and requires different routes. Traditional routing protocols cannot support the demanded QoS for different traffic classes. Best Effort routing also leads to network congestion and uneven distribution of traffic. These drawbacks have led to the dawn of QoS routing. The objective of this thesis is to determine a path for a connection bundle such that the determined path has the highest probability of satisfying the connection bundle's QoS requirements. This thesis mainly concentrates on stochastic knapsack framework to compute the dropping probabilities in the network and route the best path based on the dropping probability.