Title :
Accelerated backpressure algorithm
Author :
Zargham, Michael ; Ribeiro, Alejandro ; Jadbabaie, A.
Author_Institution :
Dept. of Electr. & Syst. Eng., Univ. of Pennsylvania, Philadelphia, PA, USA
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.
Keywords :
Newton method; approximation theory; packet radio networks; queueing theory; scheduling; stochastic processes; telecommunication network routing; ABP algorithm; accelerated backpressure algorithm; accelerated dual descent; distributed approximate Newton-like algorithm; network feasibility problem; packet arrival statistics; stochastic ADD; stochastic dual subgradient descent; stochastic gradient descent algorithm; Acceleration; Approximation algorithms; Approximation methods; Convergence; Next generation networking; Optimization; Routing;
Conference_Titel :
Global Communications Conference (GLOBECOM), 2013 IEEE
Conference_Location :
Atlanta, GA
DOI :
10.1109/GLOCOM.2013.6831412