Scaling protection to the needs of metro optical networking

Abstract

The predominant transmission technology today is still SONET/SDH. Most of the service level agreements service providers have with their customers are based on the characteristics of SONET/SDH protection mechanisms. Therefore significant effort has been spent to transfer the principles of SONET/SDH protection and self healing rings to WDM technology. Mandatory functionality of the protection scheme is protection switching within 50 ms switch-over time, and a permanent monitoring capability for both paths in order to manage switching after path restoration. This leads to DWDM protection schemes like O-UPSR or O-BLSR which are WDM equivalents to the well known UPSR and BLSR schemes from the SONET/SDH world. One major characteristic of those schemes is the protection of the entire end to end path per individual channel. End to end protection in a WDM network means full redundancy in all network sections, i.e. redundant fibre traces and line equipment, redundant filter stages and redundant active WDM channel cards. The switching mechanism takes place either on the application layer or on the tributary side of the channel cards: if the working path fails the protection channel card tributary interface goes active. This means protection is done on the single wavelength level. Looking at service level agreements among numerous service providers it is learned that typical contractually guaranteed availabilities in the metro arena range from 98.000% to 99.990%. O-UPSR or O-BLSR typically outperforms this benchmark at the cost of extremely high capital expenditure (CAPEX). Alternative protection schemes that are characterized by a lower level of redundancy inside the WDM terminals could cut down CAPEX by up to 50%. The logical consequence when reducing the redundancy in a network is an increase of the possible network downtime. Whereas in a path protected WDM terminal channel cards, channel filters and band filter stages are redundant, those alternative protection schemes introduce some sort of path switch in any stage of a WDM terminal such as line, wavelength band, single wavelength or even between line and tributary side of a WDM channel card. Due to this introduction of a single point of failure the probability for service interruption in such an alternatively protected network necessarily increases in some way. But service providers do not leave their network downtime to chance. They are used to plan and design networks according to well defined network availability numbers. The network availability is mainly based on three parameters: Mean time between failure (MTBF) value of transmission equipment, mean time to repair (MTTR) value of transmission equipment, and downtime of fibre per year and kilometer. Based on assumptions that are specific for metro networks the paper gives an overview over different protection schemes for WDM systems. Each protection scheme is explained and characterized by typical configuration, level of redundancy, level of network availability that is achieved, and protection triggering mechanism that is used. With a metro WDM system that is able to reflect such a variability in network protection service providers would be able to scale their level of infrastructure redundancy exactly according to the service level agreements (SLA) they have to offer.