Network OAM requirements for the New York City transit network

Abstract

New York City transit's subway system is one of the largest and most complex mass transportation systems in the world. Because of the extensive subway system, an extensive communications network infrastructure is needed to allow for communications of the various subway services with the control centers to support mission-critical (and safety-critical) applications. The current network utilizes multilayer network technologies, including SONET, ATM, and IP (layer 2/3 device) layer networks. As a consequence of using these various technologies, extensive challenges are faced in trying to consolidate these separate networks into a single management view, simplifying (and automating) not only the provisioning aspects but also troubleshooting/fault management aspects of the network. Such automation will help to simplify network operations and allow for better handling and management of the reliability and availability of the network. To achieve this goal, various technologies were evaluated. Consideration is being given to multiprotocol label switching technology. MPLS offers the potential to help converge to a simpler networking model. In order to achieve this, certain capabilities must be available (e.g., the ability to quickly and automatically identify defects/failures and subsequently reroute around these failures). This article discusses the operational requirements for the MPLS network from the point of view of backbone networks that support a mass transportation system operator.