A Reservation Optimised Advance Resource Reservation Scheme for Deploying RSVP in Mobile Environments

摘要

One of the major problems of deploying Resource ReSerVation Protocol (RSVP) in mobile environments is called the advance resource reservation (ARR) problem. Conventional solutions to this problem waste too many network resources and increase the new Quality of Service (QoS) session blocking probability. In this paper, we propose a reservation optimised ARR scheme which constrains the number of advance reservation paths in a subnet and only allows the most eligible mobile nodes to make advance reservations. Furthermore, to evaluate the performance of the schemes, we build Markovian models of different ARR schemes using a formal performance modelling formalism named Performance Evaluation Process Algebra (PEPA). Our results indicate that the proposed reservation optimised ARR scheme can effectively balance the active and passive reservation blocking probabilities and achieves a better utilisation of the network resources, especially when the traffic intensity is high.