Accelerated Backpressure Algorithm
Optimization and Control
2013-02-07 v1
Abstract
We develop an Accelerated Back Pressure (ABP) algorithm using Accelerated Dual Descent (ADD), a distributed approximate Newton-like algorithm that only uses local information. Our construction is based on writing the backpressure algorithm as the solution to a network feasibility problem solved via stochastic dual subgradient descent. We apply stochastic ADD in place of the stochastic gradient descent algorithm. We prove that the ABP algorithm guarantees stable queues. Our numerical experiments demonstrate a significant improvement in convergence rate, especially when the packet arrival statistics vary over time.
Cite
@article{arxiv.1302.1475,
title = {Accelerated Backpressure Algorithm},
author = {Michael Zargham and Alejandro Ribeiro and Ali Jadbabaie},
journal= {arXiv preprint arXiv:1302.1475},
year = {2013}
}
Comments
9 pages, 4 figures. A version of this work with significantly extended proofs is being submitted for journal publication