6533b82ffe1ef96bd1296538

RESEARCH PRODUCT

Gradient Scheduling Algorithm for Fair Delay Guarantee in Logarithmic Pricing Scenario

Pete RäsänenTimo HämäläinenSimo LintunenRiku KuismanenJyrki Joutsensalo

subject

Service (business)Mathematical optimizationOptimization problemLogarithmNetwork schedulerComputer scienceQuality of serviceBandwidth (signal processing)RevenueProportionally fair

description

In this paper we propose a packet scheduling scheme for ensuring delay as a Quality of Service (QoS) requirement. For customers, fair service is given while optimizing revenue of the network service provider. Gradient type algorithm for updating the weights of a packet scheduler is derived from a revenue-based optimization problem in the logarithmic pricing scenario. Algorithm is simple to implement. We compared algorithm with optimal brute-force method. The weight updating procedure is independent on the assumption of the connection's statistical behavior, and therefore it is robust against erroneous estimates of statistics.

https://doi.org/10.4108/icst.simutools2008.3046