IPO Working Group                                       D.Papadimitriou
Category: Internet Draft                                      O.Audouin
Document: draft-papadim-ipo-impairments-crosstalk-00.txt     J.-P.Faure
Expiration: May 2002                                           L.Noirie
                                                                Alcatel

                                                          November 2001



         Linear Crosstalk for Impairment-based Optical Routing

              draft-papadim-ipo-impairments-crosstalk-00.txt



Status of this Memo

   This document is an Internet-Draft and is in full conformance with
  all provisions of Section 10 of RFC2026 [1].

   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.

1. Abstract

   Optical in-band crosstalk between interfering optical channels has
   been identified (see [ITC-DIL]) as one of the major limitations to
   the diameter and the performance of photonic (or all-optical)
   networks. In this context in-band crosstalk remains a cause of
   optical signal degradation in switching elements included in a
   Photonic Cross-Connect (PXC).

   The aim of this draft is to extend the previous work dedicated to
   routing impairments [IPO-IMP]. It seeks to determine which are the
   additional linear crosstalk effects that need to be considered and
   which kind of engineering rules may be used to take these effects
   into account in constraint-based optical routing.

   Moreover, we propose to introduce IGP routing protocol extensions to
   transport information related to linear crosstalk relevant for
   wavelength routing decisions (i.e. the route of the wavelength
   throughout the optical network).

D.Papadimitriou et al. - Internet Draft  Expires May 2002           1

draft-papadim-ipo-impairments-crosstalk-00.txt              November 2001



2. 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 RFC-2119 [2].

3. Introduction

   [IPO-IMP] states that Optical crosstalk refers to the effect of
   other signals on the desired signal. It includes both coherent (i.e.
   intrachannel) crosstalk and incoherent (i.e. interchannel)
   crosstalk. Main contributors of crosstalk are the OADM and OXC sites
   that use a DWDM multiplexer/demultiplexer (MUX/DEMUX) pair. For a
   relatively sparse network where the number of OADM/OXC nodes on a
   path is low, crosstalk can be treated with a low margin in OSNR
   without being a binding constraint. But for some relatively dense
   networks where crosstalk might become a binding constraint, one
   needs to propagate the per-link crosstalk information to make sure
   that the end-to-end crosstalk which is the sum of the crosstalks on
   all the corresponding links to be within some limit, e.g. 25dB
   threshold with 1dB penalty ([Goldstein94]). Another way to treat it
   without having to propagate per-link crosstalk information is to
   have the system evaluate what the maximum number of OADM/OXC nodes
   that has a MUX/DEMUX pair for the worst route in the transparent
   domain for a low built-in margin. The latter one should work well
   where all the OXC/OADM nodes have similar level of crosstalk.ö

   While the above description proposes alternatives to overcome
   crosstalk impairments in all-optical environment, it doesnÆt propose
   a clear method to process this information in the context of
   constraint-based wavelength route computation, selection and
   allocation.

   The corresponding conclusion expressed in [IPO-IMP] is 'Crosstalk
   and effective passband narrowing due to filtering effects can be
   treated approximately as a constraint on the maximum allowable
   number of Optical Add-Drop Multiplexers (OADMs) and Photonic Cross-
   Connects (PXCs) in the transparent segment of the lightpath or
   optical channel.'

   Therefore, the aim of this memo is to provide a definition for the
   crosstalk constraint and to propose an efficient way to take these
   effects into account for dynamic constraint-based routing in
   Wavelength Switched (all-)optical Networks (WSoNs).

   In WSoNs, unicast connections carried on lightpaths are optical
   point-to-point connections between two nodes. Such connection can
   span one or more network nodes and is used to transport packets or
   TDM circuits from source node (ingress) to destination node
   (egress). The network nodes can be Optical Cross-Connects (OXC) with
   non-transparent O-E-O interfaces or Photonic Cross-Connects (PXC)
   with transparent interfaces. PXCs do not perform any conversion of

D.Papadimitriou et al.  Internet Draft  Expires May 2002           2

draft-papadim-ipo-impairments-crosstalk-00.txt              November 2001


   the optical signal into the electrical domain or vice-versa (such
   devices are also referred to as All-Optical Cross-Connects). Optical
   networks comprised of OXCs only are referred to as 'non-transparent'
   WsoNs, while networks built of PXCs only are known as 'all-optical'
   or 'transparent' WSoNs.

   In WSoNs spanning a large geographical area, an optical signal (the
   wavelength) may traverse a number of intermediate nodes and long
   fiber segments. In order to enable the signal to flow over the
   desired wavelength in the optical domain, each intermediate PXC uses
   passive and lossy switching elements (i.e., power leaking due to
   isolation) using active electrical control mechanisms.

   The progressive losses experienced by the signal in all these nodes
   and long fiber segments necessitate the use of optical amplifiers
   (usually, erbium-doped fiber amplifiers (EDFAs) or Raman amplifiers)
   at strategic locations in the network, possibly at each node and
   within the fiber segments.

   Unfortunately, the PXCs and EDFAs, while offering transparent
   switching and loss compensation respectively for optical signals,
   may introduce significant transmission impairments, such as:
   - in-band crosstalk generation when two or more optical signals
     propagate through the same PXC
   - generation of Amplifier Spontaneous Emission (ASE) noise in EDFAs
     while providing signal amplification and wavelength dependence of
     EDFA gain

   These impairments, in the absolute sense, make the optical signal
   power gain a traffic-dependent and non-deterministic quantity. The
   in-band crosstalk and the ASE noise, generated at every intermediate
   node, propagate along with the optical signal over the assigned
   carrier wavelength; and all of them undergo variable gains at
   various wavelengths because of the traffic-dependent, non flat gain
   spectra of EDFAs.

   Thus, a signal degrades in quality as it traverses through switches
   and fiber segments while propagating along its assigned lightpath
   toward its destination, and the OSNR continues to decrease. When the
   signal finally arrives at the destination, the crosstalk and ASE
   noise that have accumulated along with the signal may result in
   significant degradation of the OSNR, which in turn might increase
   the receiver bit error rate (BER) beyond its acceptable threshold.

   In order to examine the reliability of the physical layer, one needs
   to capture all of these physical-layer limitations together and
   evaluate the achievable BER for a given lightpath.

   Note: considerations related to ASE noise are fully detailed in
         [IPO-IMP]




D.Papadimitriou et al.  Internet Draft    Expires May 2002           3

draft-papadim-ipo-impairments-crosstalk-00.txt              November 2001


4. Linear crosstalk

   In addition to non-linear crosstalk (described in [IPO-NLI]), some
   crosstalk occurs even in a perfectly linear optical channel because
   of the imperfect nature of various optical components such as
   filters, demultiplexers, and switches.

4.1 Out-of-band crosstalk

   Out-of-band crosstalk (also referred to as hetero-wavelength
   crosstalk) results from the leak of a fraction of the optical signal
   power from neighboring channels that interferes with the detection
   process in optical filters and demultiplexers.

   To maintain a given value of the Bit Error Rate (BER), the
   corresponding power penalty must be kept below a certain threshold.
   For instance, the power penalty can be limited around 0.3 dB when
   maintaining a BER of 10^(-12) and 0.2 dB when maintaining a BER of
   10^(-9). Notice that this threshold is dependent on the refinement
   of the filter.

   The out-of-band crosstalk induces at the receiver side a penalty on
   the required power to maintain a given value of the BER. The
   specification of a maximum acceptable penalty gives a maximum
   admissible crosstalk value. Moreover, the out-of-band crosstalk can
   always be further reduced at the receiver side using an enhanced
   filtering.

   However, this technique can not be applied to the in-band crosstalk.
   Consequently, in-band crosstalk is a critical parameter to be taken
   into account as impairment in constraint-based optical routing.

4.2 In-band crosstalk

   In-band crosstalk (also referred to as homo-wavelength crosstalk)
   occurs during the switching of the optical signal in multiple
   devices such as PXCs, including (N x N) spatial switches. The in-
   band crosstalk for a given lightpath in the (N x N) switch is
   induced by the (N - 1) other lightpaths carried over the same
   wavelength due to the loss of the switch.

   Note that the out-of-band crosstalk may be transformed into in-band
   crosstalk in a PXC, when multiplexing optical signals coming from
   different demultiplexers through a switching stage.

   By definition (ITU-T G.692), the optical crosstalk is defined as the
   ratio of the combined total disturbing power due to signal power
   from all other channels, operating under all specified conditions,
   relative to the nominal signal power level in the desired channel,
   at the single-channel signal output reference points SD1 ... SDn
   according to Figure 1 (in G.692), within the resulting bandwidth of
   the optical demultiplexer and optical receiver, expressed in dB.


D.Papadimitriou et al.  Internet Draft   Expires May 2002           4

draft-papadim-ipo-impairments-crosstalk-00.txt              November 2001


   Therefore, the in-band crosstalk induced by an (N x N) switching
   element can be formulated as the fraction F (also referred to as the
   crosstalk level) of power leaking through the switch. When assuming
   equal power for all (N - 1) sources of coherent in-band crosstalk
   (due to the incomplete filtering caused by the partial overlap among
   the N channels), this leads to the r^2 factor: r^2 = F x (N - 1);
   where r^2 is defined as an intensive noise.

   Notice that this formulation of the induced in-band crosstalk
   reflects an ideal system where each of the N sources provides the
   same contribution to the in-band crosstalk.

   As described in [AGR-FOCS], the impact of the in-band crosstalk on
   the system performance can be evaluated by using the power penalty.
   When the in-band crosstalk is treated as an intensive noise, the
   power penalty D can be expressed as follows:

        D = - 10 log (1 - (r x Q)^2) where Q is the Q factor

   For instance, to keep the power penalty below 2dB when the Q factor
   = 8.6, the factor r must be such that r < 0.07, thus limiting the
   crosstalk level below û23dB when N = 2, below -36dB when N = 16 and
   -43dB when N = 100. Alternatively, giving a power penalty of 1dB
   (2dB) and a Q factor = 7 (BER = 10^(-12)), the crosstalk level must
   be maintained below -24dB (-21dB respectively).

   Note: this memo considers only first order crosstalk, i.e., the
   crosstalk flowing through propagating in the downstream direction
   (from ingress to egress) and producing in its turn additional
   crosstalk or higher order crosstalk) is not considered in this
   memo..

   However, the above formula is only valid when there is no optical
   noise within the system as detailed in [OFC-XTD]. It is demonstrated
   therein that the power penalty simulated and experimentally measured
   depends on the optical noise: a 1dB penalty for a low OSNR value
   induced -30dB crosstalk compared to the -24dB crosstalk without
   optical noise.

5. Impact on optical routing

   This section describes the impact of in-band crosstalk on optical
   routing.

5.1 Crosstalk computation for an optical channel

   Considering the establishment of an optical channel within a
   network, this channel will be a succession of ôjö different
   switching elements, each switching element (SE) having a XT(SE).

   Then, the total cumulated crosstalk over the whole optical channel
   or path, XT(path), is given by:


D.Papadimitriou et al.  Internet Draft   Expires May 2002           5

draft-papadim-ipo-impairments-crosstalk-00.txt             November 2001


      XT(path) = Sum(XT(SE)[j])

      where XT(SE)[j] is the XT induced by the switching element 'j'

   With this simple formula, it is possible to compute the total
   cumulated XT for any optical channel switched through a sequence of
   S switching elements.

   If the (N x N) switch element is composed of n x (M x M) switching
   sub-elements (M < N) then simply XT(SE) = Sum(XT(SSE)[i]), where
   XT(SSE)[i] is the XT induced by the switching sub-element 'i'.

   However, in real systems, the individual XT(SSE) values will be
   difficult to measure; this aspect is not considered in the remaining
   sections of this document.

5.3 Crosstalk constraint

   In Section 4, we have demonstrated that the crosstalk is an additive
   variable along an optical path (including several nodes) which
   depends on the Q factor and the channel spacing (more precisely the
   number of channels). Therefore, the cumulative XT value over the
   whole optical path (XT(path)) can be used to determine the maximum
   number of hops (i.e., PXCs) that this channel can traverse.

   Consequently, the crosstalk routing constraint can be expressed as
   follows: after switching an optical channel through a sequence of
   PXCs the total cumulative crosstalk XT(path) must be such that:

        - 10 log (1 - Q^2 x XT(path)) < D

   When XT(path) fulfills this constraint, the corresponding optical
   channel is not limited by the in-band crosstalk induced by the
   switching elements of the PXCs along the optical path (though
   depending on the OSNR). This suggests a limiting D factor compatible
   with the OSNR margin. Using a conservative value for the power
   penalty limit guarantees the feasibility of the optical channel in
   the worst OSNR case; however, this will lead to reject some feasible
   paths.

   Therefore, by applying the above formula which does not take into
   account OSNR effects, when the Q factor is known and using D as a
   constraint (usually 1 or 2dB power penalty), one can determine with
   the above formula the maximum number of switching elements (and
   subsequently the maximum number of PXCs) that a given optical
   channel can traverse.

   Another approach would be to have a table, based on measurements or
   simulations, giving the different XTmax values as a function of the
   targeted Q (since a single Q value could not be sufficient for all
   optical channels when providing QoS differentiation for instance)
   and cumulative OSNR per channel (OSNR(path)). XT(path) would then be


D.Papadimitriou et al.  Internet Draft   Expires May 2002           6

draft-papadim-ipo-impairments-crosstalk-00.txt             November 2001


   compared to the proper XTmax retrieved from this table. As such,
   this approach works without any complex formula.

   Remember also that any PXC will be designed to limit as much as
   possible the in-band crosstalk.

6. Traffic-engineering routing protocol extension

   Using the approach initiated in [IPO-ORI], the in-band crosstalk
   constraint can be integrated into the OSPF-TE Link-State
   Advertisement (LSA) or ISIS-TE Link State PDU (LSP). As mentioned
   here above, the XT(SE) parameter may be flooded using a dedicated
   extension to the Router TLV defined for IGP TE-Routing protocol.

   In OSPF, this XT(SE) parameter is included in a common sub-TLV of
   the Router TLV in the Traffic Engineering LSA since this parameter
   defines a node (and not a link) TE attribute. The Type value of this
   sub-TLV is to be attributed (TBA). The length of this sub-TLV is 4
   octets and the corresponding value specifies the XT(SE) value (in
   IEEE floating point format) per switching element. The format of the
   XT(SE) sub-TLV is shown below:

     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
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |  Type = TBA                   |         Length = 4            |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
    |                             XT(SE)                            |
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   In IS-IS, the XT(SE) parameter is included in a sub-TLV (whose type
   is TBA) of the Traffic Engineering router ID TLV (type 134). The
   length of this sub-TLV is 4 octets and the corresponding value
   specifies the XT(SE) value (in IEEE floating point format) per
   switching element. More precisely, the following sub-TLV is added:
      - Sub-TLV type: TBA
      - Length (in bytes): 4
      - Name: XT(SE)

7. Security Considerations

   This memo does not imply additional security issues than the one
   considered in [ISIS] and [OSPF].

8. References

   1  Bradner, S., "The Internet Standards Process -- Revision 3", BCP
      9, RFC 2026, October 1996.

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

   3. [AGR-FOCS] Govind P. Agrawal, ôFiber-Optic Communication

D.Papadimitriou et al. Internet Draft  Expires May 2002           7

draft-papadim-ipo-impairments-crosstalk-00.txt             November 2001


      Systemsö, Second Edition, Wiley Series in Microwave and Optical
      Engineering, March 1997.

   4. [GYS-XT] T. Gyselings, ôInvestigation and Reduction of CrossTalk
      in Wavelength Division Multiplexed All-Optical Cross-Connectsö,
      PhD Thesis, INTEC, Universiteit Gent.

   5. [ITC-DIL] K. Padmanabhan and A.N. Nevrali, ôDilated networks for
      photonic switchingö, IEEE Trans. Commun., Vol.35, pp 1357-1356,
      Dec. 1987.

   6. [IPO-IMP] A. Chiu et al., 'Impairments and Other Constraints On
      Optical Layer Routing', Internet Draft, Work in progress, draft-
      ietf-ipo-impairments-00.txt, May 2001.

   7. [IPO-NLI] D. Papadimitriou et al. 'Non-linear Routing Impairments
      in Wavelength Switched Optical Networks', Internet Draft, Work in
      Progress, draft-papadimitriou-ipo-non-linear-routing-impairm-
      01.txt, November 2001.

   8. [IPO-ORI] A. Banerjee et al., 'Impairment Constraints for Routing
      in All-Optical Networks', Internet Draft, Work in progress,
      draft-banerjee-routing-impairments-00.txt, May 2001.

   9. [OFC-XTD] L. Noirie et al., 'Crosstalk-induced degradation in an
      optical-noise-limited detection system', Paper presented during
      OFCÆ99, TuR4, San Diego, USA, 21-26 February 1999.

10. Acknowledgments

   The authors would like to thank B.Sales and E.Desmet for their
   constructive comments and inputs.

11. Author's Addresses

   Dimitri Papadimitriou
   Alcatel
   Francis Wellesplein 1,
   B-2018 Antwerpen, Belgium
   Phone: +32 3 240-8491
   Email: dimitri.papadimitriou@alcatel.be

   Jean-Paul Faure
   Alcatel
   Route de Nozay
   91461 Marcoussis Cedex, France
   Phone: +33 1 6963-1307
   Email: jean-paul.faure@ms.alcatel.fr

   Olivier Audouin
   Alcatel
   Route de Nozay
   91461 Marcoussis Cedex, France

D.Papadimitriou et al.  Internet Draft  Expires May 2002           8

draft-papadim-ipo-impairments-crosstalk-00.txt             November 2001


   Phone: +33 1 6963-2365
   Email: olivier.audouin@ms.alcatel.fr

   Ludovic Noirie
   Alcatel
   Route de Nozay
   91461 Marcoussis Cedex, France
   Phone: +33 1 6963-3476
   Email: ludovic.noirie@ms.alcatel.fr


























D.Papadimitriou et al.  Internet Draft  Expires May 2002           9

draft-papadim-ipo-impairments-crosstalk-00.txt            November 2001


Full Copyright Statement

   "Copyright (C) The Internet Society (date). All Rights Reserved.
   This document and translations of it may be copied and furnished to
   others, and derivative works that comment on or otherwise explain it
   or assist in its implementation may be prepared, copied, published
   and distributed, in whole or in part, without restriction of any
   kind, provided that the above copyright notice and this paragraph
   are included on all such copies and derivative works. However, this
   document itself may not be modified in any way, such as by removing
   the copyright notice or references to the Internet Society or other
   Internet organizations, except as needed for the purpose of
   developing Internet standards in which case the procedures for
   copyrights defined in the Internet Standards process must be
   followed, or as required to translate it into languages other than
   English. The limited permissions granted above are perpetual and
   will not be revoked by the Internet Society or its successors or
   assigns.

   This document and the information contained herein is provided on an
   "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
   TASK FORCE DISCLAIMS 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."





























D.Papadimitriou et al.  Internet Draft   Expires May 2002          10