[Docs] [txt|pdf] [Tracker] [Email] [Diff1] [Diff2] [Nits] [IPR]

Versions: 00 01 02 03 04 05 06 07 08 09 RFC 5143

Network Working Group                                 Andrew G. Malis
INTERNET-DRAFT                                 Verizon Communications
Expiration Date: December 12, 2007
Category: Historic                                     Jeremy Brayley
                                                         John Shirron
                                                     ECI Telecom Inc.

                                                         Luca Martini
                                                        Cisco Systems

                                                      Steve Vogelsang
                                                       Alcatel-Lucent

                                                        June 12, 2007


   SONET/SDH Circuit Emulation Service over MPLS (CEM) Encapsulation
                     draft-malis-sonet-ces-mpls-08.txt


Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt.

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.

   This Internet-Draft will expire on December 6, 2007.

Copyright Notice

   Copyright (C) The IETF Trust (2007).







Malis, et al.         Expires December 12, 2007            [Page 1]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


Abstract

  This document describes a historical method for encapsulating
  SONET/SDH Path signals for transport across packet-switched networks
  (PSNs).  The PSNs explicitly supported by this document include MPLS
  and IP. Note that [RFC4842] describes the standards-track protocol
  for this functionality, and new implementations must use [RFC4842]
  rather than this document except when interoperability with older
  implementations is desired.

1. Conventions used in this document

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in [RFC2119].

2. Introduction

   This document describes a historical method for encapsulating
   SONET/SDH Path signals for transport across packet-switched networks
  (PSNs).

   The transmission system for circuit-oriented TDM signals is the
   Synchronous Optical Network (SONET) [T1.105], [GR-253]/Synchronous
   Digital Hierarchy (SDH) [G.707]. To support TDM traffic (which
   includes voice, data, and private leased line services) PSNs must
   emulate the circuit characteristics of SONET/SDH payloads.  MPLS
   labels and a new circuit emulation header are used to encapsulate
   TDM signals and provide the Circuit Emulation Service over MPLS (CEM)
   function. The MPLS encapsulation may be further encapsulated in IP
   for carriage across IP PSNs [RFC4023].

   This document also describes an optional extension to CEM called
   Dynamic Bandwidth Allocation (DBA).  This is a method for
   dynamically reducing the bandwidth utilized by emulated SONET/SDH
   circuits in the packet network.  This bandwidth reduction is
   accomplished by not sending the SONET/SDH payload through the packet
   network under certain conditions such as AIS-P or STS SPE
   Unequipped.

3. Scope

   This document describes how to provide CEM for the following digital
   signals:

   1. SONET STS-1 synchronous payload envelope (SPE)/SDH VC-3

   2. STS-Nc SPE (N = 3, 12, or 48)/SDH VC-4, VC-4-4c, VC-4-16c




Malis, et al.         Expires December 12, 2007            [Page 2]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   3. Unstructured SONET Emulation, where the entire SONET bit-stream
   (including the transport overhead) is packetized and transported
   across the PSN.

   For the remainder of this document, these constructs will be
   referred to as SONET/SDH channels.

   Other SONET/SDH signals, such as virtual tributary (VT) structured
   sub-rate mapping, are not explicitly discussed in this document;
   however, it can be extended in the future to support VT and lower
   speed non-SONET/SDH services. OC-192c SPE/VC-4-64c are also not
   included at this point, since most PSNs use OC-192c or slower
   trunks, and thus would not have sufficient capacity.  As trunk
   capacities increase in the future, the scope of this document can be
   accordingly extended.

4. CEM Encapsulation Format

   In order to transport SONET/SDH SPEs through a packet-oriented
   network, the SPE is broken into fragments.  A 32-bit CEM Header is
   pre-pended to each fragment.  The Basic CEM packet appears in Figure
   1.


             +-----------------------------------+
             |            CEM Header             |
             +-----------------------------------+
             |                                   |
             |                                   |
             |        SONET/SDH SPE Fragment     |
             |                                   |
             |                                   |
             +-----------------------------------+

             Figure 1. Basic CEM Packet


   The 32-bit CEM header has the following format:

      0                   1                   2                   3
      0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
      |D|R|Rvd|   Sequence Num    | Structure Pointer |N|P|   ECC-6   |
      +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                        Figure 2. CEM Header Format






Malis, et al.         Expires December 12, 2007            [Page 3]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


The above fields are defined as follows:

   D bit: Signals DBA Mode.  MUST be set to zero for Normal Operation.
   MUST be set to one if CEM is currently in DBA mode.  DBA is an
   optional mode during which trivial SPEs are not transmitted into the
   packet network.  See Table 1 and sections 7 and 8 for further
   details.  Note: for unstructured CEM, the D-bit MUST be set to zero.

   R bit: CEM-RDI.  This bit is set to one to signal to the remote CEM
   function that a loss of packet synchronization has occurred.

   Rvd: These bits are reserved for future use, and MUST be set to
   zero.

   Sequence Number:  This is a packet sequence number, which MUST
   continuously cycle from 0 to 1023.  It SHOULD begin at zero when a
   CEM LSP is created.

   Structure Pointer: The Structure Pointer MUST contain the offset of
   the J1 byte within the CEM payload. The value is from 0 to 1,022,
   where 0 means the first byte after the CEM header. The Structure
   Pointer MUST be set to 0x3FF (1,023) if a packet does not carry the
   J1 byte.  See [T1.105], [G.707], and [GR-253] for more information
   On the J1 byte and the SONET/SDH payload pointer.  Note: for
   unstructured CEM, the Structure Pointer field MUST be set to 0x3FF.

   The N and P bits: Indicate negative and positive pointer adjustment
   events.  They are also used to relay SONET/SDH maintenance signals
   such as AIS-P.  See Table 1 and sections 7 and 8 for more details.
   Note: for unstructured CEM, the N and P bits MUST both be set to 0.

        +---+---+---+----------------------------------------------+
        | D | N | P |         Interpretation                       |
        +---+---+---+-------------+--------------------------------+
        | 0 | 0 | 0 | Normal Mode | No Ptr Adjustment              |
        | 0 | 0 | 1 | Normal Mode | Positive Ptr Adjustment        |
        | 0 | 1 | 0 | Normal Mode | Negative Ptr Adjustment        |
        | 0 | 1 | 1 | Normal Mode | AIS-P                          |
        |   |   |   |             |                                |
        | 1 | 0 | 0 | DBA Mode    | STS SPE Unequipped             |
        | 1 | 0 | 1 | DBA Mode    | STS SPE Unequipped Pos Ptr Adj |
        | 1 | 1 | 0 | DBA Mode    | STS SPE Unequipped Neg Ptr Adj |
        | 1 | 1 | 1 | DBA Mode    | AIS-P                          |
        +---+---+---+-------------+--------------------------------+

         Table 1. Interpretation of D, N, and P bits

   ECC-6: An Error Correction Code to protect the CEM header.  This
   offers the ability to correct single bit errors and detect up to two
   bit errors.  The ECC algorithm is described in Appendix B.  The ECC-


Malis, et al.         Expires December 12, 2007            [Page 4]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   6 can be optionally disabled at provisioning time.  If the ECC-6 is
   not utilized it MUST be set to zero.

   Note: Normal CEM packets are fixed in length for all of the packets
   of a particular emulated TDM stream.  This length is signaled using
   the CEM Payload Bytes parameter defined in [RFC4447], or is
   statically provisioned for each TDM stream.  Therefore, the length
   of each CEM packet does not need to be carried in the CEM header.

4.1 Transport Encapsulation

   In principle, CEM packets can be transported over any packet-
   oriented network.  The following sections describe specifically how
   CEM packets MUST be encapsulated for transport over MPLS or IP
   networks.

4.1.1 MPLS Transport

   To transport a CEM packet over an MPLS network, an MPLS label-stack
   MUST be pushed on top of the CEM packet.

   The last two labels prior to the CEM header are referred to as the
   Tunnel and Virtual Circuit (VC) labels.

   The VC label is required, and is the last label prior to the CEM
   Header.  The VC label MUST be used to identify the CEM connection
   within the MPLS tunnel.

   The optional tunnel label is immediately above the VC label on the
   label stack.  If present, the tunnel label MUST be used to identify
   the MPLS LSP used to tunnel the TDM packets through the MPLS network
   (the tunnel LSP).

   This is similar to the label stack usage defined in [RFC4447].


















Malis, et al.         Expires December 12, 2007            [Page 5]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


                 +-----------------------------------+
                 | Additional MPLS Labels (Optional) |
                 +-----------------------------------+
                 |       Tunnel Label (Optional)     |
                 +-----------------------------------+
                 |             VC Label              |
                 +-----------------------------------+
                 |            CEM Header             |
                 +-----------------------------------+
                 |                                   |
                 |                                   |
                 |       SONET/SDH SPE Fragment      |
                 |                                   |
                 |                                   |
                 +-----------------------------------+

              Figure 3. Typical MPLS Transport Encapsulation

4.1.2 IP Transport

   It is highly desirable to define a single encapsulation format that
   will work for both IP and MPLS.  Furthermore, it is desirable that
   the encapsulation mechanism be as efficient as possible.

   One way to achieve these goals is to map CEM directly onto IP by
   mapping the previously described MPLS packets onto IP.

   A mechanism for carrying MPLS over IP is described in [RFC4023].

   Using this encapsulation scheme would result in the packet format
   illustrated in figure 4.





















Malis, et al.         Expires December 12, 2007            [Page 6]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


                 +-----------------------------------+
                 |                                   |
                 |    IPv6/v4 Header [RFC4023]       |
                 |                                   |
                 +-----------------------------------+
                 |      Tunnel Label (Optional)      |
                 +-----------------------------------+
                 |             VC Label              |
                 +-----------------------------------+
                 |            CEM Header             |
                 +-----------------------------------+
                 |                                   |
                 |                                   |
                 |       SONET/SDH SPE Fragment      |
                 |                                   |
                 |                                   |
                 +-----------------------------------+

               Figure 4. MPLS Transport Encapsulation

5. CEM Operation

   The following sections describe CEM operation.

5.1 Introduction and Terminology

   There are two types of CEM: structured and unstructured.

   Unstructured CEM packetizes the entire SONET/SDH bit-stream
   (including transport overhead).

   Structured CEM terminates the transport overhead and packetizes
   individual channels (STS-1/Nc) within the SONET/SDH frame.  Because
   structured CEM terminates the transport overhead, structured CEM
   implementations SHOULD meet the generic requirements for SONET/SDH
   Line Terminating Equipment as defined in [T1.105], [G.707], and
   [GR-253].

   Implementations MUST support structured CEM and MAY support
   unstructured CEM.

   Structured CEM MUST support a normal mode of operation and MAY
   support an optional extension called Dynamic Bandwidth Allocation
   (DBA).  During normal operation, SONET/SDH payloads are fragmented,
   pre-pended with the CEM Header, the VC label, and the PSN header,
   and then transmitted into the packet network.  During DBA mode, only
   the CEM header, the VC label, and PSN header are transmitted.  This
   is done to conserve bandwidth when meaningful user data is not
   present in the SPE, such as during AIS-P or STS SPE Unequipped.



Malis, et al.         Expires December 12, 2007            [Page 7]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


5.1.1 CEM Packetizer and De-Packetizer

   As with all adaptation functions, CEM has two distinct components:
   adapting TDM SONET/SDH into a CEM packet stream, and converting the
   CEM packet stream back into a TDM SONET/SDH.  The first function
   will be referred to as CEM Packetizer and the second as CEM De-
   Packetizer.  This terminology is illustrated in figure 5.


             +------------+              +---------------+
             |            |              |               |
   SONET --> |    CEM     | --> PSN  --> |      CEM      | --> SONET
    SDH      | Packetizer |              | De-Packetizer |      SDH
             |            |              |               |
             +------------+              +---------------+

   Figure 5. CEM Terminology

   Note: the CEM de-packetizer requires a buffering mechanism to
   account for delay variation in the CEM packet stream.  This
   buffering mechanism will be generically referred to as the CEM
   jitter buffer.

5.1.2 CEM DBA

   DBA is an optional mode of operation for structured CEM that only
   transmits the CEM Header, the VC label, and PSN Header into the
   packet network under certain circumstances such as AIS-P or STS
   Unequipped.

   If DBA is supported by a CEM implementation, the user SHOULD be able
   to configure if DBA will be triggered by AIS-P, STS Unequipped,
   both, or neither on a per channel basis.

   If DBA is supported, the determination of AIS-P and STS Unequipped
   MUST be based on the state of SONET/SDH Section, Line, and Path
   Overhead bytes.  DBA based on pattern detection within the SPE (i.e.
   all zeros, 7Es, or ATM idle cells) is for further study.

   During AIS-P, there is no valid payload pointer, so pointer
   adjustments cannot occur.  During STS Unequipped, the SONET/SDH
   payload pointer is valid, and therefore pointer adjustments MUST be
   supported even during DBA.  See Table 1 for details.

5.2 Description of Normal CEM Operation

   During normal operation, the CEM packetizer will receive a fixed
   rate byte stream from a SONET/SDH interface.  When a packet's worth
   of data has been received from a SONET/SDH channel, the CEM Header,
   the VC Label, and PSN Header are pre-pended to the SPE fragment and


Malis, et al.         Expires December 12, 2007            [Page 8]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   the resulting CEM packet is transmitted into the packet network.
   Because all normal CEM packets associated with a specific SONET/SDH
   channel will have the same length, the transmission of CEM packets
   for that channel SHOULD occur at regular intervals.

   At the far end of the packet network, the CEM de-packetizer will
   receive packets into a jitter buffer and then play out the received
   byte stream at a fixed rate onto the corresponding SONET/SDH
   channel.  The jitter buffer SHOULD be adjustable in length to
   account for varying network delay behavior.  The receive packet rate
   from the packet network should be exactly balanced by the
   transmission rate onto the SONET/SDH channel, on average.  The time
   over which this average is taken corresponds to the depth of the
   jitter buffer for a specific CEM channel.

   The CEM sequence numbers provide a mechanism to detect lost and/or
   mis-ordered packets.  The CEM de-packetizer MUST detect lost or mis-
   ordered packets.  The CEM de-packetizer MUST play out a programmable
   byte pattern in place of any dropped packets.  The CEM de-packetizer
   MAY re-order packets received out of order.  If the CEM de-
   packetizer does not support re-ordering, it MUST drop mis-ordered
   packets.

5.3 Description of CEM Operation during DBA

   (Note: DBA is only applicable to structured CEM.)

   There are several issues that should be addressed by a workable CEM
   DBA mechanism.  First, when DBA is invoked, there should be a
   substantial savings in bandwidth utilization in the packet network.
   The second issue is that the transition in and out of DBA should be
   tightly coordinated between the local CEM packetizer and CEM de-
   packetizer at the far side of the packet network.  A third is that
   the transition in and out of DBA should be accomplished with minimal
   disruption to the adapted data stream.

   Another goal is that the reduction of CEM traffic due to DBA should
   not be mistaken for a fault in the packet network or vice-versa.
   Finally, the implementation of DBA should require minimal
   modifications beyond what is necessary for the nominal CEM case.
   The mechanism described below is a reasonable balance of these
   goals.

   During DBA, packets MUST be emitted at exactly the same rate as they
   would be during normal operation.  This SHOULD be accomplished by
   transmitting each DBA packet after a complete packet of data has
   been received from the SONET/SDH channel.  The only change from
   normal operation is that the CEM packets during DBA MUST only carry
   the CEM header, the VC Label, and the PSN Header.



Malis, et al.         Expires December 12, 2007            [Page 9]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   Because some links have a minimum supported packet size, the CEM
   packetizer MAY append a configurable number of bytes immediately
   after the CEM-header to pad out the CEM packet to reach the minimum
   supported packet size.  The value of the padding bytes is
   implementation specific.  The D-bit MUST be set to one, to indicate
   that DBA is active.

   The CEM de-packetizer MUST assume that each packet received with the
   D-bit set represents a normal-sized packet containing an AIS-P or
   SPE Unequipped payload as noted by N and P.  See Table 1.  The CEM
   de-packetizer MUST accept DBA packets with or without padding.

   This allows the CEM packetization and de-packetization logic during
   DBA to be similar to the nominal case.  It insures that the correct
   SONET/SDH indication is reliably transmitted between CEM adaptation
   points.  It minimizes the risk of under or over running the jitter
   buffer during the transition in and out of DBA.  And, it guarantees
   that faults in the packet network are recognized as distinctly
   different from line conditioning on the SONET/SDH interfaces.

5.4 Packet Synchronization

   A key component in declaring the state of a CEM service is whether
   or not the CEM de-packetizer is in or out of packet synchronization.
   The following paragraphs describe how that determination is made.

5.4.1 Acquisition of Packet Synchronization

   At startup, a CEM de-packetizer will be out of packet
   synchronization by default.  To declare packet synchronization at
   startup or after a loss of packet synchronization, the CEM de-
   packetizer must receive a configurable number of CEM packets with
   sequential sequence numbers.

5.4.2 Loss of Packet Synchronization

   Once a CEM de-packetizer is in packet sync, it may encounter a set
   of events that will cause it to lose packet synchronization.

   As discussed in section 6.2, a CEM de-packetizer MAY or MAY NOT
   support re-ordering of mis-ordered packets.

   If a CEM de-packetizer supports re-ordering, then the determination
   that packet synchronization has been lost cannot be made at the time
   the packets are received from the PSN.  Instead, the determination
   MUST be made as the packets are being played out onto the SONET/SDH
   interface.  This is because it is only at play-out time that the
   determination can be made as to whether the entire emulated
   SONET/SDH stream was received from the PSN.



Malis, et al.         Expires December 12, 2007           [Page 10]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   If a CEM de-packetizer does not support re-ordering, a number of
   approaches may be used to minimize the impact of mis-ordered or lost
   packets on the final re-assembled SONET/SDH stream.  For example,
   AAL1 [I.363.1] uses a simple state-machine to re-order packets in a
   subset of possible cases.  The algorithm for these state-machines is
   outside of the scope of CEM.  However, the final determination as to
   whether or not to declare loss of packet synchronization MUST be
   based on the same criteria as for implementations that do support
   re-ordering.

   Whether or not a CEM implementation supports re-ordering, the
   declaration of loss of packet synchronization MUST be based on the
   following criteria.

   As packets are played out towards the SONET/SDH interface, the CEM
   de-packetizer will encounter empty packets in the place of packets
   that were dropped by the PSN, or effectively dropped due to
   limitations of the CEM implementation.  If the CEM de-packetizer
   encounters more than a configurable number of sequential dropped
   packets, the CEM de-packetizer MUST declare loss of packet
   synchronization.

6. SONET/SDH Maintenance Signals

   There are several issues that must be considered in the mapping of
   maintenance signals between SONET/SDH and a PSN.  A description of
   how these signals and conditions are mapped between the two domains
   is described below.

   For clarity, the mappings are split into two groups: SONET/SDH to
   PSN, and PSN to SONET/SDH.

6.1 SONET/SDH to PSN

   The following sections describe how SONET/SDH Maintenance Signals
   and Alarm conditions are mapped into a Packet Switched Network.

6.1.1 AIS-P Indication

   In a SONET/SDH network, SONET/SDH Path outages are signaled using
   maintenance alarms such as Path AIS (AIS-P).  In particular, AIS-P
   indicates that the SONET/SDH Path is not currently transmitting
   valid end-user data, and the SPE contains all ones.

   It should be noted that for structured CEM nearly every type of
   service-effecting section or line defect will result in an AIS-P
   condition.

  The SONET/SDH hierarchy is illustrated below.



Malis, et al.         Expires December 12, 2007           [Page 11]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


                               +----------+
                               |   PATH   |
                               +----------+
                                    ^
                                    |
                                  AIS-P
                                    |
                                    |
                               +----------+
                               |   LINE   |
                               + ---------+
                                  ^     ^
                                  |     |
                                AIS-L   +------ LOP
                                  |
                                  |
                               +----------+
                               | SECTION  |
                               +----------+
                                  ^    ^
                                  |    |
                                  |    |
                                 LOS  LOF


                       Figure 6.  SONET/SDH Fault Hierarchy.

   Should the Section Layer detect a Loss of Signal (LOS) or Loss of
   Frame (LOF) condition, it sends AIS-L up to the Line Layer.  If the
   Line Layer detects AIS-L or Loss of Path (LOP), it sends AIS-P to
   the Path Layer.

   In normal mode during AIS-P, structured CEM packets are generated as
   usual.  The N and P bits MUST be set to 11 binary to signal AIS-P
   explicitly through the packet network.  The D-bit MUST be set to
   zero to indicate that the SPE is being carried through the packet
   network.  Normal CEM packets with the SPE fragment, CEM Header, the
   VC Label, and PSN Header MUST be transmitted into the packet
   network.

   However, to conserve network bandwidth during AIS-P, DBA MAY be
   employed.  If DBA has been enabled for AIS-P and AIS-P is currently
   occurring, the N and P bits MUST be set to 11 binary to signal AIS,
   and the D-bit MUST be set to one to indicate that the SPE is not
   being carried through the packet network.  Only the CEM header, the
   VC Label, and the PSN Header MUST be transmitted into the packet
   network.

   Also note that this differs from the outage mechanism in [RFC4447],



Malis, et al.         Expires December 12, 2007           [Page 12]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   which withdraws the VC label as a result of an endpoint outage.  TDM
   circuit emulation requires the ability to distinguish between the
   de-provisioning of a circuit (which causes the VC label to be
   withdrawn), and temporary outages (which are signaled using AIS-P).

6.1.2 STS SPE Unequipped Indication

   The STS SPE Unequipped Indication is a slightly different case than
   AIS-P.  When byte C2 of the Path Overhead (STS path signal label) is
   00h and Byte B3 (STS Path BIP-8) is valid, it indicates that the SPE
   is unequipped.  Note: this is typically signaled by setting the
   entire SPE to zeros.

   For normal structured CEM operation during SPE Unequipped, the N and
   P bits MUST be interpreted as usual.  The SPE MUST be transmitted
   into the packet network along with the CEM Header, the VC Label, and
   PSN Header, and the D-Bit MUST be set to zero.

   If DBA has been enabled for STS SPE Unequipped and the Unequipped
   condition is occurring on the SONET/SDH channel, the D-bit MUST be
   set to one to indicate DBA is active.  Only the CEM Header, the VC
   Label, and PSN Header MUST be transmitted into the packet network.
   The N and P bits MUST be used to signal pointer adjustments as
   normal.  See Table 1 and section 8 for details.

6.1.3 CEM-RDI

   The CEM function MUST send CEM-RDI towards the packet network during
   loss of packet synchronization.  This MUST be accomplished by
   setting the R bit to one in the CEM header.  This applies for both
   structured and unstructured CEM.

6.2 PSN to SONET/SDH

   The following sections discuss how the various conditions on the
   packet network are converted into SONET/SDH indications.

6.2.1 AIS-P Indication

   There are several conditions in the packet network that will cause
   the structured CEM de-packetization function to send an AIS-P
   indication onto a SONET/SDH channel.

   The first of these is the receipt of structured CEM packets with the
   N and P bits set to one, and the D-bit set to zero.  This is an
   explicit indication of AIS-P being received at the far-end of the
   packet network, with DBA disabled for AIS-P.  The CEM de-packetizer
   MUST play out the received SPE fragment (which will incidentally be
   carrying all ones), and MUST configure the SONET/SDH Overhead to
   signal AIS-P as defined in [T1.105], [G.707], and [GR-253].


Malis, et al.         Expires December 12, 2007           [Page 13]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   The second case is the receipt of structured CEM packets with the N
   and P bits set to one, and the D-bit set to one.  This is an
   explicit indication of AIS-P being received at the far-end of the
   packet network, with DBA enabled for AIS-P.  The CEM de-packetizer
   MUST play out one packet's worth of all ones for each packet
   received, and MUST configure the SONET/SDH Overhead to signal AIS-P
   as defined in [T1.105], [G.707], and [GR-253].

   A third case that will cause a structured CEM de-packetization
   function to send an AIS-P indication onto a SONET/SDH channel is
   loss of packet synchronization.

6.2.2 STS SPE Unequipped Indication

   There are three conditions in the packet network that will cause the
   CEM function to transmit STS SPE Unequipped indications onto the
   SONET/SDH channel.

   The first, which is transparent to CEM, is the receipt of regular
   CEM packets that happen to be carrying an SPE that contains the
   appropriate Path overhead to signal STS SPE unequipped.  This case
   does not require any special processing on the part of the CEM de-
   packetizer.

   The second case is the receipt of structured CEM packets that have
   the D-bit set to one to indicate DBA active and the N and P bits set
   to 00 binary, 01 binary, or 10 binary to indicate SPE Unequipped
   with or without pointer adjustments.  The CEM de-packetizer MUST use
   this information to transmit a packet of all zeros onto the
   SONET/SDH interface, and adjust the payload pointer as necessary.

   The third case when a structured CEM de-packetizer MUST send an STS
   SPE Unequipped Indication towards the SONET/SDH interface is when
   the VC-label has been withdrawn due to de-provisioning of the
   circuit.

7. Clocking Modes

   It is necessary to be able to regenerate the input service clock at
   the output interface.  Two clocking modes are supported: synchronous
   and asynchronous.  Selection of the clocking mode is made as part of
   service provisioning.  Both ends of the emulated circuit must be
   configured with the same clocking mode.

7.1 Synchronous

   When synchronous SONET/SDH timing is available at both ends of the
   circuit, the issue of clock recovery becomes much simpler.




Malis, et al.         Expires December 12, 2007           [Page 14]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


7.1.1 Synchronous Unstructured CEM

   For unstructured CEM, the external clock is used to clock each bit
   onto the optical carrier.

7.1.2 Synchronous Structured CEM

   For structured CEM, the external clock is used to clock the
   SONET/SDH carrier.  The N and P bits are used to signal negative or
   positive pointer justification events between structured CEM end-
   points.

   If there is a frequency offset between the frame rate of the
   transport overhead and that of the SONET/SDH SPE, then the alignment
   of the SPE shall periodically slip back or advance in time through
   positive or negative stuffing. The N and P bits are used to replay
   the pointer adjustment events and eliminate transport jitter.

   During a negative pointer adjustment event, the H3 byte from the
   SONET/SDH stream is incorporated into the CEM packet payload in
   order with the rest of the SPE.  During a positive pointer
   adjustment event, the stuff byte is not included in the CEM packet
   payload.

   The pointer adjustment event MUST be transmitted in three
   consecutive packets by the packetizer. The de-packetizer MUST play
   out the pointer adjustment event when the first packet of the three
   with N/P bit set is received.

   The CEM de-packetizer MUST utilize the CEM sequence numbers to
   insure that SONET/SDH pointer adjustment events are not played any
   more frequently than once per every three CEM packets transmitted by
   the remote CEM packetizer.

   References [T1.105], [G.707],and [GR-253] specify that pointer
   adjustment events MUST be separated by three SONET/SDH frames
   without a pointer adjustment event.  In order to relay all legal
   pointer adjustment events, the packet size for a specific circuit
   MUST be no larger than (783 * 4 * N)/3, where N is the STS-Nc
   multiplier.

   However, some SONET/SDH equipment allows pointer adjustments to
   occur in back to back SONET/SDH frames.  In order to support this
   possibility, the packet size for a particular circuit SHOULD be no
   larger than (783*N)/3.  Where N is the STS-Nc multiplier.

   Since the minimum value of N is one, CEM implementations SHOULD
   support a minimum payload length of 783/3 or 261 bytes.  Smaller
   payload lengths MAY be supported as an option.



Malis, et al.         Expires December 12, 2007           [Page 15]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


7.2 Asynchronous

   If synchronous timing is not available, other methods MAY be
   employed to regenerate the circuit timing.

   For structured CEM, the CEM packetizer SHOULD generate the N and P
   bits as usual.  However, without external synchronization, this
   information is not sufficient to reliably justify the SPE within the
   SONET/SDH transport framing at the CEM de-packetizer.  The de-
   packetizer MAY employ an adaptive algorithm to introduce pointer
   adjustment events to map the CEM SPE to the SONET/SDH transport
   framing.  Regardless of whether the N and P bits are used by the de-
   packetizer as part of the adaptive clock recovery algorithm, they
   MUST still be used in conjunction with the D-bit to signal AIS-P,
   SPE Unequipped, and DBA.

   For unstructured CEM, the CEM de-packetizer MAY use an adaptive
   clock recovery technique to regenerate the SONET/SDH transport
   clock.

   An example adaptive clock recovery method can be found in section
   3.4.2 of [VTOA].

8. CEM LSP Signaling

   For maximum network scaling in MPLS applications, CEM LSP signaling
   may be performed using the LDP Extended Discovery mechanism as
   augmented by the PWid FEC Element defined in [RFC4447].  MPLS traffic
   tunnels may be dedicated to CEM, or shared with other MPLS-based
   services.  The value 0x8008 is used for the PWE3 PW Type in the PWid
   FEC Element in order to signify that the LSP being signaled is to
   carry CEM.  Note that the generic control word defined in [GR-253]
   is not used, as its functionality is included in the CEM
   encapsulation header.

   Alternatively, static label assignment may be used, or a dedicated
   traffic engineered LSP may be used for each CEM service.

   Normal CEM packets are fixed in length for all of the packets of a
   particular emulated TDM stream.  This length is signaled using the
   CEM Payload Bytes parameter defined in [RFC4447], or is statically
   provisioned for each CEM service.

   At this time, other aspects of the CEM service MUST be statically
   provisioned.

9. Security Considerations

   The CEM encapsulation is subject to all of the general security
   considerations discussed in [RFC3985] and [RFC4447].  In addition,


Malis, et al.         Expires December 12, 2007           [Page 16]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   this document specifies only encapsulations, and not the protocols
   used to carry the encapsulated packets across the PSN.  Each such
   protocol may have its own set of security issues, but those issues
   are not affected by the encapsulations specified herein. Note that
   the security of the transported CEM service will only be as good as
   the security of the PSN. This level of security may be less rigorous
   then that available from a native TDM service due to the inherent
   differences between circuit-switched and packet-switched public
   networks.

10. IANA Considerations

   IANA has already allocated the PWE3 PW Type value 0x0008 for this
   specification; however, the current entry in the IANA database
   requires a correction. The entry should properly say:

0x0008 SONET/SDH Circuit Emulation Service Over MPLS Encapsulation [CEM]

   And the CEM reference should say:

[CEM]  "SONET/SDH Circuit Emulation Service over MPLS (CEM)
       Encapsulation", draft-malis-sonet-ces-mpls-07.txt
       (work in progress)

11. References

11.1 Normative References

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
        Requirement Levels", BCP 14, RFC 2119, March 1997.

   [T1.105]  American National Standards Institute, "Synchronous Optical
        Network (SONET) - Basic Description including Multiplex
        Structure, Rates and Formats," ANSI T1.105-1995.

   [G.707]  ITU Recommendation G.707, "Network Node Interface For The
        Synchronous Digital Hierarchy", 1996.

   [RFC4447]  Martini, L. et al, "Pseudowire Setup and Maintenance using
        the Label Distribution Protocol (LDP)", RFC 4447, April 2006.

   [GR-253]  Telcordia Technologies, "Synchronous Optical Network
        (SONET) Transport Systems: Common Generic Criteria",
        GR-253-CORE, Issue 3, September 2000.

   [RFC4023]  Worster, T. et al, "Encapsulating MPLS in IP or Generic
        Routing Encapsulation (GRE)", RFC 4023, March 2005.

   [I.363.1]  ITU-T, "Recommendation I.363.1, B-ISDN Adaptation Layer
        Specification: Type AAL1", Appendix III, August 1996.


Malis, et al.         Expires December 12, 2007           [Page 17]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   [VTOA]  ATM Forum, "Circuit Emulation Service Interoperability
        Specification Version 2.0", af-vtoa-0078.000, January 1997.

   [RFC4842] Malis, A. et al, "Synchronous Optical Network/Synchronous
        Digital Hierarchy (SONET/SDH) Circuit Emulation over Packet
        (CEP)", RFC 4842, April 2007

11.2 Informative References

   [RFC3985] Bryant, S. and P. Pate, "PWE3 Architecture", RFC 3985,
        March 2005.

12. Acknowledgments

   The authors would like to thank Mitri Halabi, Bob Colvin, and Ken
   Hsu, all formerly of Vivace Networks and Tellabs, and Tom Johnson,
   Marlene Drost, Ed Hallman, and Dave Danenberg, all formerly of
   Litchfield Communications, for their contributions to the document.

13. Authors' Addresses

   Andrew G. Malis
   Verizon Communications
   40 Sylvan Road
   Waltham, MA 02451
   Email: andrew.g.malis@verizon.com

   Jeremy Brayley
   ECI Telecom Inc.
   Omega Corporate Center
   1300 Omega Drive
   Pittsburgh, PA 15205
   Email: jeremy.brayley@ecitele.com

   John Shirron
   ECI Telecom Inc.
   Omega Corporate Center
   1300 Omega Drive
   Pittsburgh, PA 15205
   Email: john.shirron@ecitele.com


   Luca Martini
   Cisco Systems, Inc.
   9155 East Nichols Avenue, Suite 400
   Englewood, CO, 80112
   Email: lmartini@cisco.com

   Steve Vogelsang
   Alcatel-Lucent


Malis, et al.         Expires December 12, 2007           [Page 18]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   600 March Road
   Kanata, ON K2K 2T6
   Canada
   Email: steve.vogelsang@alcatel-lucent.com
















































Malis, et al.         Expires December 12, 2007           [Page 19]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


Appendix A. SONET/SDH Rates and Formats

   For simplicity, the discussion in this section uses SONET
   terminology, but it applies equally to SDH as well.  SDH-equivalent
   terminology is shown in the tables.

   The basic SONET modular signal is the synchronous transport signal-
   level 1 (STS-1). A number of STS-1s may be multiplexed into higher-
   level signals denoted as STS-N, with N synchronous payload envelopes
   (SPEs). The optical counterpart of the STS-N is the Optical Carrier-
   level N, or OC-N. Table 2 lists standard SONET line rates discussed
   in this document.


     OC Level          OC-1    OC-3    OC-12      OC-48     OC-192
     SDH Term             -   STM-1    STM-4     STM-16     STM-64
     Line Rate(Mb/s) 51.840 155.520  622.080  2,488.320  9,953.280

                    Table 2. Standard SONET Line Rates


   Each SONET frame is 125 us and consists of nine rows. An STS-N frame
   has nine rows and N*90 columns. Of the N*90 columns, the first N*3
   columns are transport overhead and the other N*87 columns are SPEs.
   A number of STS-1s may also be linked together to form a super-rate
   signal with only one SPE. The optical super-rate signal is denoted
   as OC-Nc, which has a higher payload capacity than OC-N.

   The first 9-byte column of each SPE is the path overhead (POH) and
   the remaining columns form the payload capacity with fixed stuff
   (STS-Nc only).  The fixed stuff, which is purely overhead, is N/3-1
   columns for STS-Nc.  Thus, STS-1 and STS-3c do not have any fixed
   stuff, STS-12c has three columns of fixed stuff, and so on.

   The POH of an STS-1 or STS-Nc is always nine bytes in nine rows. The
   payload capacity of an STS-1 is 86 columns (774 bytes) per frame.
   The payload capacity of an STS-Nc is (N*87)-(N/3) columns per frame.
   Thus, the payload capacity of an STS-3c is (3*87 - 1)*9 = 2,340
   bytes per frame. As another example, the payload capacity of an STS-
   192c is 149,760 bytes, which is exactly 64 times larger than the
   STS-3c.

   There are 8,000 SONET frames per second. Therefore, the SPE size,
   (POH plus payload capacity) of an STS-1 is 783*8*8,000 = 50.112
   Mb/s. The SPE size of a concatenated STS-3c is 2,349 bytes per frame
   or 150.336 Mb/s. The payload capacity of an STS-192c is 149,760
   bytes per frame, which is equivalent to 9,584.640 Mb/s. Table 2
   lists the SPE and payload rates supported.




Malis, et al.         Expires December 12, 2007           [Page 20]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   SONET STS Level     STS-1   STS-3c  STS-12c    STS-48c   STS-192c
   SDH VC Level            -     VC-4  VC-4-4c   VC-4-16c   VC-4-64c
   Payload Size(Bytes)   774    2,340    9,360     37,440    149,760
   Payload Rate(Mb/s) 49.536  149.760  599.040  2,396.160  9,584.640
   SPE Size(Bytes)       783    2,349    9,396     37,584    150,336
   SPE Rate(Mb/s)     50.112  150.336  601.344  2,405.376  9,621.504

                      Table 2. Payload Size and Rate


   To support circuit emulation, the entire SPE of a SONET STS or SDH
   VC level is encapsulated into packets, using the encapsulation
   defined in section 5, for carriage across packet-switched networks.

Appendix B. ECC-6 Definition

   ECC-6 is an Error Correction Code to protect the CEM header.  This
   provides single bit correction and the ability to detect up to two
   bit errors.


   Error Correction Code:


   |---------------Header bits 0-25 -------------------| ECC-6 code|
   0                   1                   2                   3
   0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |1 1 1 1 1 0 0 0 1 0 0 0 1 1 1 1 1 0 1 0 0 0 1 0 1 1|1 0 0 0 0 0|
   |1 1 1 1 0 1 0 0 0 1 0 0 1 0 0 0 0 1 0 1 1 1 1 1 1 1|0 1 0 0 0 0|
   |1 0 0 0 1 1 1 1 0 0 1 0 1 1 1 0 0 0 1 1 1 1 0 0 1 1|0 0 1 0 0 0|
   |0 1 0 0 1 1 1 1 0 0 0 1 1 0 0 1 1 1 1 1 0 0 1 1 0 1|0 0 0 1 0 0|
   |0 0 1 0 0 0 1 0 1 1 1 1 1 1 0 0 1 1 1 1 1 0 1 0 1 0|0 0 0 0 1 0|
   |0 0 0 1 0 0 0 1 1 1 1 1 0 0 1 1 0 0 1 1 0 1 1 1 1 1|0 0 0 0 0 1|
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                      Figure 7. ECC-6 Check Matrix X


   The ECC-6 code protects the 32 bit CEM header as follows:

   The encoder generates the 6 bit ECC using the matrix shown in Figure
   7.  In brief, the encoder builds another 26 column by 6 row matrix
   and calculates even parity over the rows.  The matrix columns
   represent CEM header bits 0 through 25.

   Denote each column of the ECC-6 check matrix by X[], and each column
   of the intermediate encoder matrix as Y[].  CEM[] denotes the CEM
   header and ECC[] is the error correction code that is inserted into
   CEM header bits 26 through 31.


Malis, et al.         Expires December 12, 2007           [Page 21]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


   for i = 0 to 25 {
        if CEM[i] = 0 {
                Y[i] = 0;
        } else {
                Y[i] = X[i];
        }
   }

   In other words, for each CEM header bit (i) set to 1, set the
   resulting matrix column Y[i] according to Figure 7.

   The final ECC-6 code is calculated as even parity of each row in Y
   (i.e. ECC[k]=CEM[25+k]=even parity of row k).

   The receiver also uses matrix X to calculate an intermediate matrix
   Y' based on all 32 bits of the CEM header.  Therefore Y' is 32
   columns wide and includes the ECC-6 code.

   for i = 0 to 31 {
        if CEM[i] = 0 {
                Y'[i] = 0;
        } else {
                Y'[i] = X[i];
        }
   }

   The receiver then appends the incoming ECC-6 code to Y as column 32
   (ECC[0] should align with row 0) and calculates even parity for each
   row.  The result is a single 6 bit column Z.  If all 6 bits are 0,
   there are no bit errors (or at least no detectable errors).
   Otherwise, it uses Z to perform a reverse lookup on X[] from Figure
   7.  If Z matches column X[i], then there is a single bit error.  The
   receiver should invert bit CEM[i] to correct the header.  If Z fails
   to match any column of X, then the CEM header contains more than one
   bit error and the CEM packet MUST be discarded.

   Note that the ECC-6 code provides single bit correction and 2-bit
   detection of errors within the received ECC-6 code itself.














Malis, et al.         Expires December 12, 2007           [Page 22]

              SONET/SDH Circuit Emulation over MPLS   June 12, 2007


Full Copyright Statement

   Copyright (C) The IETF Trust (2007).

   This document is subject to the rights, licenses and restrictions
   contained in BCP 78, and except as set forth therein, the authors
   retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY, THE IETF TRUST AND
   THE INTERNET ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS
   OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF
   THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.


Intellectual Property

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at
   ietf-ipr@ietf.org.


Acknowledgment

   Funding for the RFC Editor function is provided by the IETF
   Administrative Support Activity (IASA).






Malis, et al.         Expires December 12, 2007           [Page 23]


Html markup produced by rfcmarkup 1.107, available from http://tools.ietf.org/tools/rfcmarkup/