A MDP approach to fault-tolerant routing

Authors:
Antonio Pietrabissa;Marco Castrucci
Affiliations:
Computer and System Science Dept. of the University of Rome Sapienza;Computer and System Science Dept. of the University of Rome Sapienza
Venue:
WD'09 Proceedings of the 2nd IFIP conference on Wireless days
Year:
2009

Citing 4
Cited 0

MDP routing for multi-rate loss networks

Computer Networks: The International Journal of Computer and Telecommunications Networking
Introduction to Reinforcement Learning

Introduction to Reinforcement Learning
Providing Fault-Tolerant Ad hoc Routing Service in Adversarial Environments

Wireless Personal Communications: An International Journal
A Fault-Tolerant Routing Algorithm for Mobile Ad Hoc Networks Using a Stochastic Learning-Based Weak Estimation Procedure

WIMOB '06 Proceedings of the 2006 IEEE International Conference on Wireless and Mobile Computing, Networking and Communications

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper defines a theoretical framework based on Markov Decision Processes (MDP) to deal with fault-tolerant routing algorithms in heterogeneous home networks, which are realized through the integration of different wired and wireless telecommunication technologies. Such networks are characterized by fast dynamics of link availability, mainly due to the wide use of wireless technologies. The novelty of this paper is the MDP approach to the fault-tolerant routing problem, which is addressed by introducing a re-routing policy: when a path becomes unavailable, the flows transmitted over that path are re-routed on another available path; the new path is selected taking into consideration the probability that also the new path can become unavailable in the future, in order to minimize re-routing occurrences. Numerical simulations show the effectiveness of the proposed approach.