(T) At SuperComm 2005 in Chicago, most vendors of optical CPE and transport products have now adopted the new Ethernet over SONET/SDH (EoS) family of protocols. The first EoS chips and products were launched in 2003.
Enterprise networks are demanding Ethernet services to their service providers.
New services for service providers usually drive the requirements for new equipment. While the demand increases, service providers optimize their networks to increase their profitability until a point where new technology is required again.
In a time when capital is constrained, service providers want to use their present SONET/SDH infrastructure to transport Ethernet in their access and metro networks.
Three new technologies are emerging, whose combination can optimize SONET/SDH for data transport: the Generic Framing Protocol (GFP), Virtual Concatenation (VCAT) and LCAS (Link Capacity Adjustment).
EoS by combining GFP, VCAT, and LCAS provides improved bandwidth efficiently for data transport while allowing the Service Provider to operate its SONET/SDH transport network. Before EoS, equipment vendors have used a number of proprietary encapsulation techniques to transport IP/Ethernet over SONET/SDH.
The first method has been to use ATM AAL 5 over SONET/SDH. ATM is a very efficient switching and multiplexing technology which speeds and feeds scale with SONET/SDH but requires a high overhead with the ATM “cell tax” of 5-byte header and heavy software burden because mainly of its connection-oriented capability.
Other methods have been focused mainly on using PPP. The IP traffic coming to an Ethernet port is encapsulated over a PPP link and multiple ports can be encapsulated over ML-PPP links. By using an HDLC framing, the PPP traffic is transporting over the SONET/SDH payload.
These methods have been standardized within the IETF through RFC 1662, RFC 1990 and RFC 2615. The ITU-T expanded this work by specifying the use of LAPS (very similar protocol to PPP/HDLC) and specifying IP over LAPS in X.85/Y.1321 and Ethernet over LAPS in X.86/Y1323.
All these encapsulation mechanisms suffer from the inherent deficiencies of HDLC framing which introduces variable packet sizes because of its trailer and suffers from limited protection from corruption of flag, address etc…
Now to better optimize the transport of Ethernet and other data services over SONET, GFP has been standardized taking into account both the pros and cons of ATM and PPP/HDLC and leveraging two new emerging SONET/SDH capabilities: VCAT and LCAS.
Generic Framing Procedure (GFP)
GFP defines a mapping of client data signals into SONET/SDH payloads in order to allow SONET/SDH to transport none-TDM traffic more efficiently. GFP defines two types of client signals:
- Frame-mapped GFP for PDU-oriented signals such as IP/ PPP or Ethernet MAC;
- Transparent-mapped GFP for block-oriented signals such as Fiber Channel and ESCON.
GFP provides a flexible and robust encapsulation technology that supports both fixed and variable length frame. Unlike HDLC, GFP does not use any special character for frame delineation. GFP provides a more deterministic encapsulation scheme than HDLC whose overhead is data dependent. GFP generalizes the ATM frame delineation mechanism to encapsulate variable length frames. Its frame delineation is based on the length of the current payload and an error control check.
GFP provides two major benefits. First, it gives one uniform mechanism to transport any data type over SONET/SDH. Second, its encapsulation mechanism is superior to HDLC without the layer processing of ATM.
Virtual Concatenation (VCAT)
Two approaches exist for concatenation: contiguous and virtual. Both solutions provide concatenated bandwidth of X-times Container-N at the path termination.
However, contiguous concatenation keeps the concatenated SONET payload through the whole SONET/SDH transport. Therefore, network elements must support contiguous concatenation from the source to the destination and, at every intermediate node.
Virtual Concatenation relaxes the “rigidity” of SONET/SDH payloads originally designed for TDM traffic. VCAT allows the concatenation of multiple payload frames from VT1.5 to STS-3c SPE to better scale the requirements for incremental data bandwidth.
VCAT breaks the initial SONET payload at the source into individuals Virtual Containers (VC). Each VC is part logically of a Virtual Concatenated Group (VCG). Each VCG member is routed and transported individually across the SONET/SDH transport network and is recombined with the other VCs at the destination node to form the whole VCG.
Therefore, network elements must support virtual concatenation at the source and at the destination. But intermediate nodes do not need to be aware of the virtual concatenation.
VCAT can be provided in two different ways:
- High-Order: for STS-M-Nv where n indicates the number of STS-m virtually concatenated.
- M can be equal to 1 (STS-1) or 3 (STS-3c);
- N can vary from 0 to 255;
- Low-Order: for VT-M-Nv where n indicates the number of VT-m virtually concatenated.
- M can be equal to 1.5, 2, 3 and 6;
- N can vary from 0 to 64.
VCAT provides much more efficient use of the transport bandwidth for data user interfaces. With VCAT, an OC-48 link can carry two full Gb Ethernet with 95% of the link used through 7 virtual STS-3c instead of one Gb Ethernet with 42% of the link used through an STS-48c.
Furthermore, since VCG members are not constrained to the same path, VCAT allows more efficient use of the capacity of the different routes in the network.
Link Capacity Adjustment Scheme (LCAS)
Standardized in ITU-T G.7042/Y.1305, LCAS is a signaling protocol for sizing virtually concatenated paths. With LCAS, VCG can be resized at any time without disturbing network traffic.
LCAS signaling messages are exchanged to change the number of VC between the source and the destination of the path. The number of VC can be increased or decreased without any frames lost therefore increasing or decreasing the capacity of the VCG link.
LCAS provides as well a means of removing links that have experienced failures. The VC in fault is detected and removed automatically from the VCG.
LCAS has been designed to operate both with management systems for the set-up and release of VC but can also operate with emerging GMPLS-based control plane responsible for network path set-up and teardown.
The use of LCAS provides an effective way for the Service Provider to change the bandwidth allocated. Provisioning quickly the right bandwidth at any time is a major operation management goal of Service Providers.
Combining GFP, VCAT, and LCAS
By combining GFP, VCAT, and LCAS, Service Providers have a more efficient way to optimize their SONET/SDH transport network for Ethernet services. GFP, VCAT, and LCAS do not require end-to-end upgrades to the existing SONET/SDH network. EoS network elements using GFP, VCAT, and LCAS can be deployed at the ingress and the egress of the Service Provider’s transport network.
Copyright © 2005-2005 by Serge-Paul Carrasco. All rights reserved.
Contact Us: asvinsider at gmail dot com.