Ethernet over SDHEthernet Over SDH (EoS or EoSDH) or Ethernet over SONET refers to a set of protocols which allow Ethernet traffic to be carried over synchronous digital hierarchy networks in an efficient and flexible way. The same functions are available using SONET (a predominantly North American standard).

Ethernet frames which are to be sent on the SDH link are sent through an "encapsulation" block (typically Generic Framing Procedure or GFP) to create a synchronous stream of data from the asynchronous Ethernet packets. The synchronous stream of encapsulated data is then passed through a mapping block which typically uses virtual concatenation (VCAT) to route the stream of bits over one or more SDH paths. As this is byte interleaved, it provides a better level of security compared to other mechanisms for Ethernet transport.

After traversing SDH paths, the traffic is processed in the reverse fashion: virtual concatenation path processing to recreate the original synchronous byte stream, followed by decapsulation to converting the synchronous data stream to an asynchronous stream of Ethernet frames.

The SDH paths may be VC-4, VC-3, VC-12 or VC-11 paths. Up to 64 VC-11 or VC-12 paths can be concatenated together to form a single larger virtually concatenated group. Up to 256 VC-3 or VC-4 paths can be concatenated together to form a single larger virtually concatenated group. The paths within a group are referred to as "members". A virtually concatenated group is typically referred to by the notation <pathType>-<X>v, where <pathType> is VC-4, VC-3, VC-12 or VC-11 and X is the number of members in the group.

A 10-Mbit/s Ethernet link is often transported over a VC-12-5v which allows the full bandwidth to be carried for all packet sizes.

A 100-Mbit/s Ethernet link is often transported over a VC-3-2v which allows the full bandwidth to be carried when smaller packets are used (< 250 bytes) and Ethernet flow control restricts the rate of traffic for larger packets. But does only give ca.97Mbit/s, not full 100Mb.

A 1000-Mbit/s (or 1 GigE) Ethernet link is often transported over a VC-3-21v or a VC-4-7v which allows the full bandwidth to be carried for all packets.

The bandwidths possible are shown in the table below:

Container (SDH)Container (SONET) Type Payload Capacity (Mbit/s)

VC-11-Xv VT-1.5-Xv SPE Low Order X x 1.600 (X = 1 to 64)

VC-12-Xv VT-2-Xv SPE Low Order X x 2.176 (X = 1 to 64)

VC-3-Xv - Low Order X x 48.384 (X = 1 to 256)

VC-3-Xv STS-1-Xv SPE High Order X x 48.384 (X = 1 to 256)

VC-4-Xv STS-3c-Xv SPE High Order X x 149.76 (X = 1 to 256)

EoS also drops the "idle" packets of the Ethernet frame before encapsulating the Ethernet frame to GFP, which is recreated at the other end during decapsulation process. Hence this provide a better throughput compared to native Ethernet transport.

1

An additional protocol, called link capacity adjustment scheme (LCAS), allows the two endpoints of the SDH paths to negotiate which paths are working and can carry traffic versus which paths should not be used to carry traffic.

Link Capacity Adjustment Scheme or LCAS is a method to dynamically increase or decrease the bandwidth of virtual concatenated containers. The LCAS protocol is specified in ITU-T G.7042.

It allows on-demand increase or decrease of the bandwidth of the virtual concatenated group in a hitless manner. This brings bandwidth-on-demand capability for data clients like Ethernet when mapped into TDM containers.

LCAS is also able to temporarily remove failed members from the virtual concatenation group. A failed member will automatically cause a decrease of the bandwidth and after repair the bandwidth will increase again in a hitless fashion. Together with diverse routing this provides survivability of data traffic without requiring excess protection bandwidth allocation

Multiprotocol Label Switching (disingkat menjadi MPLS) adalah teknologi penyampaian paket pada jaringan backbone berkecepatan tinggi. Asas kerjanya menggabungkan beberapa kelebihan dari sistem komunikasi circuit-switched dan packet-switched yang melahirkan teknologi yang lebih baik dari keduanya. Sebelumnya, paket-paket diteruskan dengan protokol routing seperti OSPF, IS-IS, BGP, atau EGP. Protokol routing berada pada lapisan network (ketiga) dalam sistem OSI, sedangkan MPLS berada di antara lapisan kedua dan ketiga.

Prinsip kerja MPLS ialah menggabungkan kecepatan switching pada layer 2 dengan kemampuan routing dan skalabilitas pada layer 3. Cara kerjanya adalah dengan menyelipkan label di antara header layer 2 dan layer 3 pada paket yang diteruskan. Label dihasilkan oleh Label-Switching Router dimana bertindak sebagai penghubung jaringan MPLS dengan jaringan luar. Label berisi informasi tujuan node selanjutnya kemana paket harus dikirim. Kemudian paket diteruskan ke node berikutnya, di node ini label paket akan dilepas dan diberi label yang baru yang berisi tujuan berikutnya. Paket-paket diteruskan dalam path yang disebut LSP (Label Switching Path).

Komponen MPLS :

← Label Switched Path (LSP): Merupakan jalur yang melalui satu atau serangkaian LSR dimana paket diteruskan oleh label swapping dari satu MPLS node ke MPLS node yang lain.

← Label Switching Router: MPLS node yang mampu meneruskan paket-paket layer-3

← MPLS Edge Node atau Label Edge Router (LER): MPLS node yang menghubungkan sebuah MPLS domain dengan node yang berada diluar MPLS domain

← MPLS Egress Node: MPLS node yang mengatur trafik saat meninggalkan MPLS domain

2

← MPLS ingress Node: MPLS node yang mengatur trafik saat akan memasuki MPLS domain

← MPLS label: merupakan label yang ditempatkan sebagai MPLS header ← MPLS node: node yang menjalankan MPLS. MPLS node ini sebagai control

protokol yang akan meneruskan paket berdasarkan label.

MSP(Multiplex Section Protection) is a per span protection.

A service line is protected using another line, called a protection line. If an error occurs, the protection mechanism should switch over to the protection line. There are two main protection schemes for the multiplex section:• 1+1 : Traffic is simultaneously transmitted over working and protecting lines (or cards if it is for hiT i.e. MSP bridge). The incoming traffic is select from the line that delivers signal in best condition (specifically switch fabric selector is responsible for making the selection in a HiT).o Switching type: unidirectional or bi-directionalo Operation type: revertive or non-revertive

• 1:N : A 1:N multiplex section protection system consists of N traffic-carrying multiplex sections that are to be protected by an additional multiplex section. In this scheme only one of the working sections can be protect at a time. The additional multiplex section can be used to carry low-priority traffic (unprotected) when it is not used as a protection section for the rest N working sections.o 1:N with N<=14 for STM-1/4/16 and N<=7 for STM-16/64o Switching type: bi-directionalo Operation type: revertive

• 1:1 : This is a special case of 1:N protection scheme. In case of a failure on the working path, traffic is switched to protecting path.o Switching type: bi-directionalo Operation type: revertive

SubNetwork Connection Protection. is a per path protection.

SNCP is a network protection mechanism for SDH networks providing path protection (end-to-end protection). The data signal is transmitted in a ring structure via two different paths and can be implemented in line or ring structures. The changeover criteria are specified individually when configuring a network element. A protection protocol is not required. The switchover to protection path occurs in the non-revertive mode, i.e. if traffic was switched to the protection path due to a transmission fault, there is no automatic switch-back to the original path once the fault is rectified, but only if there is a fault on the new path (the one labeled as “protecting” and currently services traffic).

SNCP is a 1+1 protection scheme (one working and one protection transport entity). Input traffic is broadcasted in two routes (one being the normal working route and the second one being the protection route).

Assume a failure free state for a path from a node B to a node A. Node B bridges the signal destined to A from other nodes on the ring, both on working and protecting routes. At node A, signals from these two routes are continuously monitored for path layer defects and the better quality signal is selected.Now consider a failure state where fiber between node A and node B is cut. The selector switches traffic on the standby route when the active route between node A and node B is failed.

In order to prevent any unnecessary or spurious protection switching in the presence of bit errors on both paths, a switch will typically occur when the quality of the alternate path

3

exceeds that of the current working path by some threshold (e.g., an order of magnitude better BER). Consecutively, any case of failure drops in SNCP’s decision mechanism.

4