Distributed asynchronous algorithms with stochastic delays for constrained optimization problems with conditions of time drift

A distributed asynchronous algorithm for minimizing a function with a nonstationary minimum over a constraint set is considered. The communication delays among the processors are assumed to be stochastic with Markovian character. Conditions which guarantee the mean square and almost sure convergence...

Full description

Saved in:
Bibliographic Details
Published in:Parallel computing Vol. 21; no. 9; pp. 1431 - 1450
Main Authors: Beidas, Bassem F., Papavassilopoulos, George P.
Format: Journal Article
Language:English
Published: Amsterdam Elsevier B.V 01.09.1995
Elsevier
Subjects:
Algorithmics. Computability. Computer arithmetics
Applied sciences
Asynchronous iteration
Computer science; control theory; systems
Computer systems and distributed systems. User interface
Convergence analysis
Exact sciences and technology
Multiprocessor environment
Optimization problem
Software
Stochastic communication delay
Telecommunications
Telecommunications and information theory
Teleprocessing networks. Isdn
Theoretical computing
Asynchronous iteration
Stochastic communication delay
Multiprocessor environment
Optimization problem
Convergence analysis
Multiprocessor
Asynchronous
Simulation
Programming environment
Distributed algorithm
Routing
Stochastic approximation
Iterative method
Implementation
Delay
Convergence
Constrained optimization
ISSN:0167-8191, 1872-7336
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A distributed asynchronous algorithm for minimizing a function with a nonstationary minimum over a constraint set is considered. The communication delays among the processors are assumed to be stochastic with Markovian character. Conditions which guarantee the mean square and almost sure convergence to the sought solution are presented. We also present an optimal routing application for a network that connects various U.S. cities. Results of the extensive simulation that we implemented assert the practical applicability of distributed asynchronous algorithms with stochastic delays. Comparison results for varying the probability distribution of these delays are provided. The impact of varying the communication delay bound and the stepsize is also assessed.
ISSN:0167-8191
1872-7336
DOI:10.1016/0167-8191(95)00005-9