Proportional Bandwidth Guaranteed Probabilistic Priority Scheduling for Assured Forwarding

This paper studies the design and performance of the Probabilistic Priority (PP) [9] packet scheduling algorithm to schedule packets. Unlike an earlier design that uses fractional arithmetic and prohibits large number of classes, we present an integer PP algorithm and show that PP turns out to be a special scheme of applying lottery scheduling [16] to bandwidth allocation in a strict priority sense. We then propose a Multi-winner PP (MPP) scheduler using multi-winner lottery scheduling to improve the throughput and response time accuracy. We also propose a technique to reduce the deadline violation probability in probabilistic scheduling. Simulations show that our algorithms provide lower delays for most classes and improved bandwidth guarantee than lottery scheduling and strict prioritization. Finally, we investigate the issue of parameter assignment for an MPP scheduler and use our techniques to implement a prototype Assured Forwarding (AF) mechanism in a high-speed network processor with memory and computing constraint.