Controlling G-AIMD by Index Policy

Abstract : We consider the Generalized Additive Increase Multiplicative Decrease (G-AIMD) dynamics for resource allocation with alpha fairness utility function. This dynamics has a number of important applications such as internet congestion control, charging electric vehicles, and smart grids. We prove indexability for the special case of MIMD model and provide an efficient scheme to compute the index. The use of index policy allows us to avoid the curse of dimensionality. We also demonstrate through simulations for another special case, AIMD, that the index policy is close to optimal and significantly outperforms a natural heuristic which penalizes the strongest user.
Complete list of metadatas

Cited literature [34 references]  Display  Hide  Download

https://hal.inria.fr/hal-01648312
Contributor : Konstantin Avrachenkov <>
Submitted on : Saturday, November 25, 2017 - 12:57:38 PM
Last modification on : Sunday, July 22, 2018 - 9:10:03 AM
Long-term archiving on : Monday, February 26, 2018 - 12:35:29 PM

File

bandshare2.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01648312, version 1

Collections

Citation

Konstantin Avrachenkov, Vivek Borkar, Sarath Pattathil. Controlling G-AIMD by Index Policy. CDC 2017 - 56th IEEE Conference on Decision and Control, Dec 2017, Melbourne, Australia. pp.1-6. ⟨hal-01648312⟩

Share

Metrics

Record views

264

Files downloads

198