Rational Agent-Based Decision Algorithm for Strategic Converged Network Migration Planning

To keep up with constantly growing user demands for services with higher quality and bandwidth requirements, telecommunication operators are forced to upgrade their networks. This upgrade, or migration of the network to a new technology, is a complex strategic network planning problem that involves...

Full description

Saved in:
Bibliographic Details
Published in:Journal of optical communications and networking 2019-07, Vol.11 (7), p.371-382
Main Authors: Patri, Sai Kireet, Grigoreva, Elena, Kellerer, Wolfgang, Mas Machuca, Carmen
Format: Article
Language:English
Subjects:
Algorithms
Artificial intelligence (AI)
Biological system modeling
Converged network planning
Convergence
Migration
Migrations
Net present value (NPV)
Network planning
New technology
Optical communication
Optical fiber devices
Passive optical network (PON)
Passive optical networks
Planning
Profitability
Rational agents
State of the art
Strategic network planning
Uncertainty
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:To keep up with constantly growing user demands for services with higher quality and bandwidth requirements, telecommunication operators are forced to upgrade their networks. This upgrade, or migration of the network to a new technology, is a complex strategic network planning problem that involves technoeconomic evaluations over multiple periods of time. The state-of-the-art approaches consider migrations to a concrete architecture and do not take uncertainties, such as user churn, into account. This results in migration cost underestimations and profitability overestimations. In this paper, we propose a generic migration algorithm derived from a search-based rational agent decision process that can deal with uncertainties and provides the migration path using a maximized utility function. The algorithm maximizes the migration project profitability, measured as the accumulated net present value. This flexible and generic methodology has been evaluated on the example of migration from existing copper networks to the future-proof passive optical network architectures. Our proposed flexible migration algorithm is validated over pure residential and converged scenarios in a fully reproducible case study. The results affirm that migration flexibility is key to profit maximization.
ISSN:1943-0620
1943-0639
DOI:10.1364/JOCN.11.000371