Bounds, Approximations And Applications For A Two-Queue GPS System

In this paper we study Generalized Processor Sharing (GPS) scheduling with Markov Modulated Fluid Sources (MMFSs). We focus on a GPS system serving two classes of customers. By using a combination of the bounding approach and the approximation approach to performance analysis and by taking advantage of the specific structure of MMFSs, we are able to derive lower bound and upper bound approximation on queue length distributions for each class of the GPS system. Numerical investigation shows that the lower bound and the upper bound approximation are very tight. Hence our work greatly improve the earlier results on GPS scheduling in [YaSi94, ZTK95] which are obtained for a more general stochastic model. Applications of our performance bounds for GPS scheduling to call admission control and bandwidth sharing are also illustrated, and comparison with FIFO and strict priority in different scenarios is made and discussed. We show that the flexibility provided by GPS does not provide much bett..