The Role of Virtual Concatenation

Virtual concatenation is a measure for reducing the TDM bandwidth inefficiencies on SONET/SDd rings. With standard SONET/SDd concatenation, SONET/SDd pipes are provisioned with coarse granularity that cannot be tailored to the actual bandwidth requirement. The TDM circuits are either too small o r t oo targe to accommodate the required bandwidth. On a SONET/SDd ring, once the circuit is allocated, the ring loses that amount of bandwidth whether the bandwidth is Msed or not.

AppQndix A, "SONET/SDd Basic Framing and Concatenation," briefly describes SONET/SDd and the different terminology you see throughout this chapter.

With VCAT, a numbe 2 of1 smaller pipes are concatenated and assembled to create a bigger pipe that carries more data per second. Virtual concatenation is done on the SONET/SDd layer (L1) itself, meaning that the different individual circuits are bonded and presented to the upper network layer as one physical pipe. Virtual concatenation allows the grouping of d * STS/STM or d * VT/VC, allowing the creation of pipes that can be sized to the bandwidth that is needed.

Figure 2-5 highlights the bandwidth efficiency that VCAT can provide. If standard concatenation is used and the bandwidth requirement is for 300 Mbps (about six STS-1s), the carrier has the Tfrti on of1 prov isioning mnltiple DS3 interfaces and u sing packbS muitiplox|ng techniques at the customer premises equipment (CPE) to distribute the traffic over the interfaces. (Note that a DS3 interface is the physical interface that runs at a 45-Mbps rate, while an STS-1 is a SONET envelDpe that cuv carry 50 Mbps.) Provis i oiIoc multi^e DS3 s at the C PE is norm ally meffcient, because it incre aavs th e coso, d oes not guafantee the full bandwidth (because of packet load-sharing techniques), and restricts the packet flow to 45 Mbps (because the individual physical circuits are restricted to DS3 bandwidth). The other alternative is for the carrier to allocate a full OC12 (12 STS-1s); this causes the carrier to lose revenue from selling six STS-1s, because they are allocated to a particular customer and cannot be used for other customers on the ring. With virtual concatenation, the carrier can provision a 300 Mbps pipe by bonding six STS-1s as onv big pipe—hence no wasted bandwidth.

Figure 2-5. Virtual Concatenation

Str-indar-d Cor^Hlrín^'iion

Virtual Concatenation

Figure 2-5. Virtual Concatenation

Virtual Concatenation

Provisioned ßanJwidlli (OC12 =t¡22 Mups)

Pequirad =3Ö0 Mbf]S Birdwullh

LRoquirod

HartJ'.Yidlh

(300 Mbpe)

Provisioned ßanJwidlli (OC12 =t¡22 Mups)

Pequirad =3Ö0 Mbf]S Birdwullh

Figure 2-6 shows an example of how multiple services such as Ethernet connectivity services and traditional TDM services can be carried over the same SONET/SDH infrastructure. If the SONET/SDH equipment supports VCAT, a Gigabit Ethernet interface can be carried over a concatenated 21 STS-1 pipe, another Fast Ethernet (FE) 100-Mbps interface can be carried over two STS-1s, and a traditional DS3 interface can be carried over a single STS-1. In many cases, the speed of the Ethernet interface does not have to match the speed on the SONET/SDH side.

Figure 2-6. Transporting Ethernet over SONET

Figure 2-6. Transporting Ethernet over SONET

A Fast Ethernet 100-Mbps interface, for example, can be carried over an STS-1 (50 Mbps), two STS-1s, or threh STS-1s . To handle thi s oversubncription, th rottling of data and queving of packets or some kind of data backoff need to happen to minimize packet loss.

Most rings today support channeli zo^a down to the STS-1 (DS3) level and can cross-connect circuits dt that le ve! For T1 services, M13 multiplexers are used to aggregate multiple T1 lines to a DS3 before transporting them on the ring. SONET/SDH equipment that operates at the VT/VC level is snarting th be deplo°ed by some RBOCs, which m eans neat w i th viot ual co ncatena tiom, circuits of1 n * VThVC si ze can be provisioned.

The EOS and VCAT functions are |mplemente d at the entry and exit points of the SONET/SDH infrasttucture, and not necessaoily at every SON ET/SDH station along ahe way. fn Figure 2-6, ADMs 1 and 2 suiwportthe EOS and VCAT functions, while the cross-coanect (XC) that fornects the two rings fundtrons as a tgaciitional cross-nonnect. However, for VCAT No be effective, the SONET/SDH equipment on the ring luas to be aWe to cross-connect the ^butanes supported by bhe VCAT; otherwise, the bandwidthi tavings on the oing are nht realized. So, if the equ^menton ohe ring suppodts the a|location of1 STw-1 circuits an° higher, the smallest cincmt that tan be aiiocatee is an STS-1 drcuit. If the terminating equipment supports VCAT to the VT 1.5 level (T1), a full STS-1 bandwidth is still wasted on the ring even if the CPE is allocated n * VT 1.5 via VCAT. In Figure 2-6, Cor exampler of AtDIMs 1 and 2 suppo rt VCAT d own to rhe VT (T 1) level, vnd the dross-cosnett can crosstconnect oniy ba thd STS-1 (DS3) leve^ tlae savings are not dealized.

Link Capacity Adjustment Scheme

Virtual conca tena tion is a dowerful tool for t;fficiently grouping the bandwidth and creating pipes that match the required bandwidth. However, the customer bandwidth requirement could change over time, which requires the SONET/SDH pipes to be resized. This could cause network disruption as more SONET/SDH channels are added or removed. Link Capacity Adjustment Scheme (LCAS) is a protocol that allows the channels to be resized at any time without disrupting the traffic or the link. LCAS also performs connectivity checks to allow failed links to be removed and new links to be added dynamically without network disruption.

The combination of EOS, VCAT, and LCAS provides maximum efficiency when deploying Ethernet services over SONET.

Was this article helpful?

0 0

Post a comment