A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks

Routing data packets in a dynamic network is a difficult and important problem in computer networks. As the network is dynamic, it is subject to frequent topology changes and is subject to variable link costs due to congestion and bandwidth. Existing shortest path algorithms fail to converge to bett...

Full description

Bibliographic Details
Main Authors: Shrirang Ambaji KULKARNI, Raghavendra G . RAO
Format: Article
Language:English
Published: Stefan cel Mare University of Suceava 2017-04-01
Series:Journal of Applied Computer Science & Mathematics
Subjects:
Online Access:http://jacsm.ro/view/?pid=23_2
_version_ 1819127702138912768
author Shrirang Ambaji KULKARNI
Raghavendra G . RAO
author_facet Shrirang Ambaji KULKARNI
Raghavendra G . RAO
author_sort Shrirang Ambaji KULKARNI
collection DOAJ
description Routing data packets in a dynamic network is a difficult and important problem in computer networks. As the network is dynamic, it is subject to frequent topology changes and is subject to variable link costs due to congestion and bandwidth. Existing shortest path algorithms fail to converge to better solutions under dynamic network conditions. Reinforcement learning algorithms posses better adaptation techniques in dynamic environments. In this paper we apply model based Q-Routing technique for routing in dynamic network. To analyze the correctness of Q-Routing algorithms mathematically, we provide a proof and also implement a SPIN based verification model. We also perform simulation based analysis of Q-Routing for given metrics.
first_indexed 2024-12-22T08:16:07Z
format Article
id doaj.art-3172dc72cd4b4a139f5886d6e31ec999
institution Directory Open Access Journal
issn 2066-4273
2066-3129
language English
last_indexed 2024-12-22T08:16:07Z
publishDate 2017-04-01
publisher Stefan cel Mare University of Suceava
record_format Article
series Journal of Applied Computer Science & Mathematics
spelling doaj.art-3172dc72cd4b4a139f5886d6e31ec9992022-12-21T18:32:53ZengStefan cel Mare University of SuceavaJournal of Applied Computer Science & Mathematics2066-42732066-31292017-04-01111131610.4316/JACSM.201701002A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic NetworksShrirang Ambaji KULKARNI0Raghavendra G . RAO1GIT, Belgaum - 590008, IndiaNIE Institute of Technology, Mysore - 57008, IndiaRouting data packets in a dynamic network is a difficult and important problem in computer networks. As the network is dynamic, it is subject to frequent topology changes and is subject to variable link costs due to congestion and bandwidth. Existing shortest path algorithms fail to converge to better solutions under dynamic network conditions. Reinforcement learning algorithms posses better adaptation techniques in dynamic environments. In this paper we apply model based Q-Routing technique for routing in dynamic network. To analyze the correctness of Q-Routing algorithms mathematically, we provide a proof and also implement a SPIN based verification model. We also perform simulation based analysis of Q-Routing for given metrics.http://jacsm.ro/view/?pid=23_2Reinforcement LearningQ - Routing. SPIN model checkingDynamic NetworksQ - Routing. SPIN model checking
spellingShingle Shrirang Ambaji KULKARNI
Raghavendra G . RAO
A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks
Journal of Applied Computer Science & Mathematics
Reinforcement Learning
Q - Routing. SPIN model checking
Dynamic Networks
Q - Routing. SPIN model checking
title A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks
title_full A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks
title_fullStr A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks
title_full_unstemmed A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks
title_short A Formal Verification Model for Performance Analysis of Reinforcement Learning Algorithms Applied t o Dynamic Networks
title_sort formal verification model for performance analysis of reinforcement learning algorithms applied t o dynamic networks
topic Reinforcement Learning
Q - Routing. SPIN model checking
Dynamic Networks
Q - Routing. SPIN model checking
url http://jacsm.ro/view/?pid=23_2
work_keys_str_mv AT shrirangambajikulkarni aformalverificationmodelforperformanceanalysisofreinforcementlearningalgorithmsappliedtodynamicnetworks
AT raghavendragrao aformalverificationmodelforperformanceanalysisofreinforcementlearningalgorithmsappliedtodynamicnetworks
AT shrirangambajikulkarni formalverificationmodelforperformanceanalysisofreinforcementlearningalgorithmsappliedtodynamicnetworks
AT raghavendragrao formalverificationmodelforperformanceanalysisofreinforcementlearningalgorithmsappliedtodynamicnetworks