Note: Descriptions are shown in the official language in which they were submitted.
CA 02357988 2001-09-27
TOPOLOGY DISCOVERY IN OPTICAL WDM NETWORKS
This invention relates to topology discovery in
optical WDM (wavelength division multiplex) communications
networks or systems.
Background
As communications networks become more complex, the
task of network management becomes increasingly difficult. An
important aspect of a network management system, or NMS,
relates to determining and maintaining an accurate record of
the topology or connectivity of the network. In optical WDM
communications networks this may involve knowledge of not only
connections of optical fibers among nodes of the network, but
also connections of optical fibers within the nodes, allocation
of wavelengths to respective optical fibers, and arrangements
and sequences of multiplexers and demultiplexers, or optical
band filters, within the nodes.
While it is conceivable to provide and maintain
manually a record of the topology of a network, this becomes
increasingly impractical as the network becomes more complex,
and has other disadvantages such as being subject to errors and
being slow and inconvenient to update and respond to changes.
In communications networks it has been proposed to
provide automatic discovery of the topo:logy of the network.
For example, such proposals are disclosed for ATM networks in
Suzuki United States Patent No. 5,796,736 issued August 18,
1998 entitled "ATM Network Topology Auto Discovery Method" and
in Chatwani et al. United States Patent No. 5,729,685 issued
March 17, 1998 and entitled "Apparatus For Determining The
Topology Of An ATM Network Or The Like Via Communication Of
Topology Information Between A Central Manager And Switches In
The Network Over A Virtual Service Path".
CA 02357988 2001-09-27
2
As disclosed in the latter patent, ATM cells include
so-called link advertisement messages (L'AMs) each of which
identifies an originating switch and port number and is
forwarded by a receiving or neighbour ATM switch to a network
manager. The network manager thereby develops information
profiling the topology of the network. Matching of LAM pairs
is carried out by the network manager to confirm bidirectional
NNI (Network-Network Interface) links.
While such a known arrangement may be effective for
discovering neighbours in an ATM network, it requires handling
of the LAMs specifically in each ATM switch and reduces
bandwidth of the network for data traffic. In addition, such a
known arrangement is not effective for determining the topology
of an optical communications network, in which for example an
optical fiber path between two nodes A and B may pass through
an intermediate node C. The ATM cell or packet level which
would be determined by such a known arrangement would indicate
that the nodes A and B are coupled together and would not show
the node C, whereas the actual topology in this case is that
the node A is coupled to the node C, and the node C is coupled
to the node B.
Also, Wood United States Patent No. 6,108,702 issued
August 22, 2000 and entitled "Method And Apparatus For
Determining Accurate Topology Features Of A Network" discloses
a system for monitoring packet traffic in a network to
determine topology features using logical groupings of ports
and/or devices. Orr et al. United States Patent No. 5,727,157
issued March 10, 1998 and entitled "Apparatus And Method For
Determining A Computer Network Topology" discloses determining
the topology of a computer network including data-relay devices
based on a comparison of source addresses heard by the various
data-relay devices.
CA 02357988 2001-09-27
3
In an optical network, Fee United States Patent No.
6,108,113 issued August 22, 2000 and entitled "Method And
System For Transporting Ancillary Network Data" discloses
superimposing a sub-carrier modulation signal at a relatively
low frequency (e.g. 1 MHz), containing ancillary network data,
on an optical carrier of a high bit rate (e.g. 1 to 10 GHz)
data signal. This patent discloses that the ancillary network
data can include any of numerous data types identifying any of
numerous network elements, and can be used for any of numerous
network management purposes one of which is listed as "Probing
Network Topology", but no further information in these respects
is disclosed.
Fatehi et al. United States Patent No. 5,892,606
issued April 6, 1999 and entitled "Maintenance Of Optical
Networks" discloses an apparatus for adding a dither signal to
an optical carrier modulated with an information signal, to
provide a method for monitoring and tracking end-to-end signal
routing in multi-wavelength optical networks. This patent
discloses that the monitoring can take place at any point in
the network. This patent is not concerned with topology
discovery.
A need exists for an improved method of topology
discovery which is particularly applicable to optical WDM
networks.
Summary of the Invention
According to one aspect of this invention there is
provided a method of determining topology of an optical WDM
(wavelength division multiplex) network in which optical
signals comprising a plurality of WDM optical channels are
communicated, comprising the steps of: modulating each optical
channel with a respective signal comprising a channel identity;
detecting the channel identities of all of the optical channels
CA 02357988 2001-09-27
4
in an optical signal at each of a plurality of points in the
network to produce a channel list for each of said points; and
identifying matched pairs of channel lists to determine optical
paths of the network between pairs of said points.
Another aspect of this invention provides a method of
determining topology of an optical WDM (wavelength division
multiplex) network in which optical signals comprising a
plurality of WDM optical channels are communicated among a
plurality of nodes of the network, comprising the steps of:
modulating each optical channel with a respective signal
comprising a channel identity; for each of a plurality of
optical paths entering or leaving each of a plurality of nodes,
determining a channel list of all the optical channels in an
optical signal on the optical path, this step comprising
detecting the channel identities of all of the optical channels
in an optical signal at each of a plurality of points; and
identifying matching channel lists to determine optical paths
of the network between the nodes.
The step of determining a channel list of all the
optical channels in an optical signal on an optical path
entering or leaving a node can comprise detecting the channel
identities of all of the optical channels in an optical signal
on the respective optical path.
Alternatively, this step can comprise, for each node:
detecting the channel identities of all of the optical channels
in an optical signal at a multiplex port of each of a plurality
of optical band filters of the node to produce a respective
channel list M; determining a respective channel list T for an
optical signal at a through port of the respective optical band
filter, the channel list T comprising channels of the
respective list M which are not within a pass band of the
optical band filter; identifying matching channel lists M and T
CA 02357988 2001-09-27
to determine optical paths within the node; and identifying
unmatched channel lists M or T as channel lists for optical
path entering or leaving the node.
In this case the step of identifying matching channel
5 lists M and T to determine optical paths within the node can
comprise identifying any optically transparent optical band
filters of the node for each of which the channel lists M and T
are the same; identifying matched pairs of the other channel
lists M and T for the node to determine optical paths between
respective ports of different optical band filters within the
node; and identifying any channel lists, from among said
matched pairs of channel lists for the node, matching said same
channel lists M and T to determine optical connections of said
optically transparent optical band filters within the node.
The invention also provides an optical WDM
(wavelength division multiplex) network comprising a plurality
of nodes and optical paths for communicating optical signals,
comprising a plurality of WDM optical channels, within and
among the nodes, the network comprising: a source for each
optical channel; a modulator for modulating each optical
channel with a respective signal comprising a channel identity;
a plurality of optical filters for combining optical channels
to produce optical signals and for separating optical signals
to derive optical channels from the optical signals; a
plurality of detectors for detecting the channel identities of
all of the optical channels in an optical signal at each of a
plurality of points in the network to produce a channel list
for each of said points; and a network management system for
identifying matched pairs of said channel lists to determine
optical paths of the network between pairs of said points.
In such a network the optical filters can comprise
optical band filters each having a multiplex port, an add or
CA 02357988 2001-09-27
6
drop port, and a through port, and said plurality of points in
the network can comprise multiplex ports of the optical band
filters. Preferably the network management system is arranged
to determine a channel list for a through port of an optical
band filter by omitting, from optical channels of a channel
list for the multiplex port of the respective optical band
filter, optical channels within a pass band of the optical band
filter.
Brief Description of the Drawings
The invention will be further understood from the
following description by way of example with reference to the
accompanying drawings, in which:
Fig. 1 illustrates a plurality of nodes, a network
management system, and communication paths therebetween forming
part of an optical WDM communications network incorporating an
embodiment of the invention;
Fig. 2 illustrates apparatus provided in one or more
of the nodes in accordance with an embodiment of the invention;
Fig. 3 illustrates one form of modulator which can be
used in the apparatus of Fig. 2;
Fig. 4 illustrates one form of detector which can be
used in the apparatus of Fig. 2;
Fig. 5 is a flow chart providing an outline of steps
carried out in performing a method in accordance with an
embodiment of the invention;
Fig. 6 is a flow chart illustrating in more detail
steps carried out for each node in performing the method
represented by Fig. 5;
CA 02357988 2001-09-27
7
Fig. 7 illustrates a plurality of optical band
filters with reference to which an aspect of an embodiment of
the invention is explained; and
Fig. 8 is a flow chart illustrating steps carried out
for the network in performing the method represented by Fig. 5.
Detailed Description
Referring to Fig. 1, a simple optical communications
network is illustrated as comprising a plurality of, in this
example three, nodes 10, identified individually as Node A,
Node B, and Node C, which are coupled together via various
optical communications paths 12 represented by solid lines.
The paths 12 can comprise optical fibers and/or wavelengths on
optical fibers. The network also includes a network management
system (NMS) or station 14, to which all of the nodes 10 are
coupled via communications paths 16 represented by dashed
lines. The paths 16 can be optical or other (e.g. electrical)
communications paths.
As is known in the art, such a communications network
can include an arbitrary number of nodes 10 and NMSs 14, which
may be located together or separately from one another, with
various arrangements of communications paths among the nodes 10
and NMSs 14 having any desired configuration or topology.
Furthermore, such a communications network can be coupled to
other similar or different communications networks in various
manners. Accordingly, Fig. 1 serves merely to illustrate a
simple form of network for the purposes of describing an
embodiment of the invention.
As illustrated in Fig. 1, one of the paths 12, from
Node A to Node B, is divided into two parts 12-1 and 12-2 which
pass through Node C as illustrated by a dotted link in Node C.
Although this path may pass via one or more optical filters
CA 02357988 2001-09-27
8
which can drop or add optical channels in the Node C, it is
assumed that no optical channels are dropped from or added to
this path in Node C. Thus at a communications level, this path
only communicates signals from Node A to Node B. However, the
topology of the network is that this path comprises the path
12-1 from Node A to Node C, and the path 12-2 from Node C to
Node B. Discovery of the topology of the network by the NMS 14
is required, for example, to identify the paths 12-1 and 12-2
via Node C.
It can be appreciated that, without any interception
of signals on the dotted link path at Node C in the network of
Fig. 1, an analysis of signals at the communications level will
not identify the separate paths 12-1 and 12-2 via Node C, but
merely a signal path from Node A to Node B. It can also be
appreciated that this is a very simple example, and that in
practice an optical network may include far more than three
nodes, which may include multiple optical band filters for
multiplexing and demultiplexing numerous different wavelength
optical signals in a WDM network, and that there may be far
more optical communication paths, so that automatically and
accurately determining the topology of the network by the NMS
14 can be a very difficult task.
In embodiments of this invention, carrying out this
task is facilitated by providing each optical signal within the
network with a respective identifier, referred to as a channel
identity or CID. The CID is applied as a relatively low
frequency (e.g. about 1 MHz or less) amplitude modulation of
the optical signal, referred to as a dither tone, using any
desired coding scheme for the CID.
As one example, each CID may be constituted by a
repeated sequence of 14 bytes of data identifying the CID.
Numerous other ways of encoding unique CIDs for the respective
CA 02357988 2001-09-27
9
optical signals exist and any of these can be used, with
appropriate forms of detection and decoding of the CIDs.
Fig. 2 illustrates apparatus provided in one or more
of the nodes in accordance with an embodiment of the invention.
As illustrated in Fig. 2, each optical signal or channel, for
example supplied from a respective one of a plurality of
optical sources 20 modulated with respective data signals (not
shown), is supplied via a respective one of a plurality of
electronic variable optical attenuators (EVOAs) 22 supplied
with respective CIDs, whereby each optical signal or channel is
modulated with a respective CID as indicated above. As
illustrated in Fig. 2 for a WDM network, optical signals with
their respective CIDs are multiplexed in an optical multiplexer
24 to produce optical signals within an optical band on an
optical path 26.
The optical path 26 is coupled to an add port AD of
an optical band filter (OBF) 28 which serves to multiplex the
optical signals on the optical path 26 with other optical
channels supplied via an optical path 30 to a through port TH
of the OBF 28. These other optical channels on the optical
path 30 similarly are provided, for example in the same manner
as the signals from the optical sources 20, each with a
respective CID. The multiplexed optical signal produced by the
OBF 28 is supplied via a CID detector 32 to an optical path 34.
The CID detector 32, for example as further described
below, detects the CIDs of all of the optical channels present
in the optical signal at the multiplexed output port of the OBF
28 and hence on the optical path 34, and supplies these via one
of the paths 16 to the NMS 14.
The optical path 34 should lead to, i.e. supply its
optical signal to, an optical input elsewhere in the optical
network. For example, it may lead to a through input port of
CA 02357988 2001-09-27
another OBF within the same node as the OBF 28, or to a
multiplex input port of another OBF, acting as a demultiplexer
or drop filter, within the same node as the OBF 28, or to an
input port of another OBF in a different node from that
5 containing the OBF 28.
Fig. 2 illustrates the optical path 34 as being
coupled via a path 36, represented by a dashed line, to an
optical path 38, which is coupled via another CID detector 40
to the multiplex input port of another OBF 42 having through
10 and drop output ports TH and DR respectively providing
demultiplexed optical output signals. Like the OBF 32, the CID
detector 40 detects the CIDs of all of the optical channels
present in the optical signal at the multiplexed input port of
the OBF 42, and hence on the optical path 38, and supplies
these via one of the paths 16 to the NMS 14.
It can be appreciated from the above description that
the optical path 36 may comprise an optical fiber between two
different nodes, one of which includes the OBF 28 and the other
of which includes the OBF 42, or it may comprise an optical
path within a node, the OBFs 28 and 42 both being within this
one node. As described above, the NMS 14 may be separate from
the nodes, incorporated within a node, or distributed among
nodes, with the paths 16 being provided accordingly.
Fig. 3 illustrates one form of modulator which can be
used in the apparatus of Fig. 2, for implementing each of the
EVOAs 22 as shown in Fig. 2.
Referring to Fig. 3, an optical signal to be provided
with a CID is supplied via an optical path 50, an EVOA 52, and
an optical tap 54 to an output optical path 56. A small
portion, e.g. 5%, of the optical signal which is tapped by the
optical tap 54 is detected by an optical detector 58, whose
electrical output is amplified and filtered by an AGC amplifier
CA 02357988 2001-09-27
11
and filter unit 60, an output of which is supplied to a digital
signal processor (DSP) unit 62. The DSP unit 62 provides a
controlled voltage bias to the EVOA 52 in accordance with a
respective CID for the optical signal, with which the DSP unit
62 is supplied for example from the NMS 14. The modulator of
Fig. 3 can modulate the optical signal with a desired amplitude
modulation depth, for example of about 1% to about 4%, at
frequencies of for example up to about 1 MHz, to provide the
desired form of CID modulation on the optical signal.
It is observed that the optical signal is also
modulated in known manner with a data signal typically at a
high bit rate. As described above with reference to Fig. 2,
the CID modulation is applied to each optical signal after its
data modulation. It can be appreciated that this need not be
the case, and instead the CID modulation can be applied to an
optical carrier which is subsequently modulated with data to be
carried.
Fig. 4 illustrates one form of detector which can be
used in the apparatus of Fig. 2, for implementing each of the
CID detectors 32 and 40 as shown in Fig. 2.
Referring to Fig. 4, an optical signal at the
multiplexed port of an optical band filter, for example at the
output of the OBF 28 acting as a multiplexer in Fig. 2 or at
the input of the OBF 42 acting as a demultiplexer in Fig. 2, on
an optical path 64 is supplied to an optical path 66 via an
optical tap 68. A small portion, e.g. 5%, of the optical
signal which is tapped by the optical tap 68 is detected by an
optical detector 70, whose electrical output is amplified and
filtered by an AGC amplifier and filter unit 72, an output of
which is supplied to a DSP unit 74. The DSP unit 74 derives
the CIDs of all of the optical channels which are present in
CA 02357988 2001-09-27
12
the optical signal on the path 64, and provides these as an
output to the NMS 14.
The forms and functions of the DSP units 62 and 74
depend on the particular form of CIDs used. For example, the
DSP unit 62 may incorporate a digital synthesizer and the DSP
unit 74 may incorporate a Fast Fourier Transform (FFT) or
Discrete Fourier Transform (DFT) function for detecting CID
tones.
As indicated above, the provision of a respective CID
for each optical channel, and the detection of the CIDs by the
CID detectors, is used in a method of determining the topology
of the network by the NMS 14 in accordance with an embodiment
of this invention. Fig. 5 is a flow chart providing an outline
of steps carried out in performing this method.
Referring to Fig. 5, a block 80 represents a first
step of providing each optical channel with a respective CID,
this being carried out for example as described above with
reference to Figs. 2 and 3. A subsequent block 82 represents a
step of detecting the CIDs and supplying these to the NMS 14,
this being carried out for example at the multiplexed ports of
the optical band filters as described above with reference to
Figs. 2 and 4. In this step the NMS 14 is also supplied with
related information as to the "colour" of, i.e. the wavelengths
added or dropped by, each respective optical band filter and
the direction of the optical signals, i.e. whether the
respective optical band filter is acting as a multiplexer like
the OBF 28 or as a demultiplexer like the OBF 42, as described
above with reference to Fig. 2.
A block 84 in Fig. 5 represents a further step of
processing, individually for each node in the optical network,
the above information for the respective node to determine from
the detected CIDs inter-node optical paths which lead to or
CA 02357988 2001-09-27
13
from another node, i.e. for each node, each optical path like
the path 34 in Fig. 2 which supplies its optical signal to
another node, and each optical path like the path 38 in Fig. 2
via which an optical signal is supplied from another node.
This step can also determine the topology of optical devices
within the individual nodes. A further block 86 in Fig. 5
represents a final step of processing, for the optical network
as a whole, the inter-node optical paths to determine the
overall topology of the optical network. The steps represented
by the blocks 84 and 86 can also provide warnings or alarms to
a network operator in the event that errors or faults are
determined from the discovered topologyõ
Accordingly, in the step 82 the NMS 14 is supplied,
for each node 10, with the CIDs detected by the CID detectors,
such as the detectors 32 and 40 as described above, at the
multiplex port of each OBF of the respective node, together
with information as to the colour of the OBF and the optical
signal direction, i.e. whether this multiplex port is an
optical input or an optical output. For each node, the NMS
carries out the step 84 in Fig. 5, for example in accordance
with the more detailed steps of Fig. 6. Thus steps 88 to 100
of Fig. 6 serve to implement the step 84 of Fig. 5.
Referring to Fig. 6, in the step 88 the NMS 14
creates, for each OBF of the respective node 10, three ordered
lists of CIDs, referred to as lists M, B, and T. The list M is
a list of all of the CIDs detected by the respective CID
detector at the multiplex port of the OBF, supplied in the step
82. The list B is a subset of the list M, limited to the CIDs
in the list M for optical channels with wavelengths that are
within the colour of the OBF, i.e. these wavelengths are
selected by the band filter to be coupled between the add port
AD or drop port DR and the multiplex port of the OBF, so that
the list B contains the CIDs of optical channels that are added
CA 02357988 2001-09-27
14
or dropped by the OBF. The list T contains the other CIDs,
i.e. the CIDs of the list M which are not in the list B,
representing optical channels which are present at the through
port TH of the OBF. Each of these lists also has an associated
indication of whether it relates to optical channels at an
input to or an output from the respective OBF.
In determining the intra-node and inter-node topology
of the optical network, each of the created lists is used as an
identity for the optical path or fiber at the respective port
of the respective OBF and node. Any of these lists created in
the step 88 can be a null list, i.e. an empty list containing
no CIDs; for example a null list would be created for an OBF
port which is unconnected and carries no optical signals. At
the step 90 in Fig. 6, null lists are excluded from the further
topology-determining steps.
In the step 92, any lists for OBFs for which the
lists M and T are the same (except for the signal input/output
association) are temporarily set aside. Such lists correspond
to OBFs for which no optical channels are added or dropped, so
that the lists of CIDs at the multiplex and through ports of
the OBF are the same. An example of this is further described
below with reference to Fig. 7.
In the step 94, optical paths within the respective
node 10, i.e. intra-node paths, are determined by the NMS 14 by
matching pairs of the remaining lists M and T for the node, one
list of each pair being an output from one OBF and the other
list of the respective pair being an input to another OBF of
the node. Thus each such matched pair of lists also represents
an optical path or fiber between the ports associated with the
lists.
In the subsequent step 96, it is concluded that the
remaining lists M and T, i.e. those that are not null, set
CA 02357988 2001-09-27
aside, or paired within the node, represent ports of the node
that are coupled to other nodes, and hence the optical paths or
fibers coupled to these ports. The lists B identify their
respective ports as add or drop ports, depending upon whether
5 the ports are input or output ports. Except for OBFs whose
lists have been temporarily set aside, the topology of the
node, i.e. the optical paths between OBFs of the node, is
thereby determined.
In the step 98, the list M (or the list T which is
10 the same) of each OBF whose lists have been temporarily set
aside are matched with the lists of the optical paths of the
node determined in the steps 94 and 96, and the determined
intra-node topology is expanded to include these OBFs in the
respective optical paths. As further described below, where an
15 optical path includes only one such OBF, its position and
optical connections are thereby fully determined, but for an
optical path including two or more such OBFs the order of such
OBFs within this optical path is undetermined (but this is
relatively unimportant because such OBFs do not add or drop any
optical channels).
In the final step 100 of Fig. 6, it is concluded that
any remaining matching lists M and T which were set aside in
the step 92 relate to OBFs which are optically transparent
within the node, i.e. these lists relate to at least one OBF
which is coupled between ports of the node that are coupled to
other nodes, but which OBF does not drop or add any optical
channels so that the node is optically transparent to the
associated optical signal.
For example, if for a particular OBF in the node the
list M relates to an input port of the OBF and the list T
relates to an output port of the OBF, and these lists are the
same and not null, then they are set aside at the step 92. if
CA 02357988 2001-09-27
16
at the step 98 these lists are not matched with another list in
the node, then it is concluded at the step 100 that the list M
for this OBF constitutes an input optical path of the node, the
list T represents an output optical path of the node, and that
optical signals are coupled transparently, i.e. without any
dropping or adding of optical channels, from this node input to
this node output via this OBF. As in the case of the step 98,
if at the step 100 it is determined that the matching lists M
and T of a plurality of OBFs in the node are the same, then it
is concluded that these OBFs are coupled in sequence with one
another and that the particular order of the OBFs in this
sequence can not be determined.
The steps of Fig. 6 are further explained by way of
example with reference to Fig. 7, which illustrates one of
numerous possible arrangements of OBFs within a node. As shown
in Fig. 7, the node is assumed to contain seven OBFs which are
identified as band filters Fl to F7. As illustrated, the OBFs
Fl to F4 are arranged as drop or demultiplexing filters, and
the OBFs F5 to F7 are arranged as add or multiplexing filters.
For each of the filters Fl to F7, a rectangle at the multiplex
port of the filter represents the presence of a respective CID
detector as described above. For the drop filters Fl to F4 the
multiplex port is an optical input, and for the add filters F5
to F7 the multiplex port is an optical output, of the
respective OBF.
In the step 88 of Fig. 6, the three ordered lists M,
B, and T are created by the NMS 14 for each of the OBFs Fl to
F7. For clarity, these lists are represented as M1, Bi, and Ti
for the OBF F1, M2, B2, and T2 for the OBF F2, and so on, and
these list designations are illustrated in Fig. 7 adjacent to
the OBF ports to which they relate. It can be appreciated that
the lists Ml to M4, B5 to B7, and T5 to T7 are identified as
CA 02357988 2001-09-27
17
optical inputs for their respective OBFs, and the other lists
are identified as optical outputs for their respective OBFs.
As illustrated in Fig. 7, it is assumed that no
optical channels are dropped by the OBFs F2 and F3, and that no
optical channels are added by the OBFs F5 and F7, so that their
drop/add ports are unconnected and the lists B2, B3, B5, and B7
are null lists and are excluded at step 90 in Fig. 6. Further,
for each of these OBFs F2, F3, F5, and F7 the respective list T
is the same as the respective list M, sc> that these lists for
these OBFs are set aside at step 92 in Fig. 6.
Of the remaining lists M and T, in the step 94 of
Fig. 6 the NMS 14 pairs the list T1 (output) with the list M4
(input), and pairs the list T4 (output) with the list T6
(input), in each case concluding that there is an optical path
between the ports corresponding to the paired lists. In the
step 96 of Fig. 6 the NMS 14 concludes correctly that the
remaining lists M and T, i.e. the lists M1 and M6, correspond
to inter-node optical paths, i.e. respectively an input port
and an output port of the node, to which inter-node optical
paths or fibers are coupled. The NMS 14 also recognizes from
their input and output associations that the lists B1 and B4
relate to optical channels dropped in the node and that the
list B6 relates to optical channels added in the node.
In the step 98 of Fig. 6, the NMS 14 matches the set-
aside lists M2 and M3 each to the list T1 or M4 to determine
that the OBFs F2 and F3 are present in the optical path T1-M4
as determined above, but it can not determine the order of the
OBFs F2 and F3 within this optical path (i.e. whether the OBF
F2 precedes the OBF F3 as illustrated, or whether the OBF F3
precedes the OBF F2). In addition, the NMS 14 matches the set-
aside list M5 to the list T4 or T6 to determine that the OBF F5
is present in the optical path T4-T6 as determined above; in
CA 02357988 2001-09-27
18
this case it concludes by matching inputs and outputs that
there are optical paths T4-T5 and M5-T6.
There is no match in the step 98 for the list M7,
Accordingly, in the step 100 of Fig. 6 the NMS 14 concludes
that the list T7 (input) and the list M7 (output) relate
respectively to an input port and an output port of the node,
corresponding to inter-node optical paths, and that the OBF F7
provides an optically transparent coupling from this input port
to this output port of the node.
It should be appreciated that the simple example of
Fig. 7 is provided only by way of additional explanation of the
steps of Fig. 6, and that these steps can be carried out to
determine arbitrary topologies of nodes.
In the step 86 of Fig. 6, the NMS performs a sequence
of steps to determine the inter-node, or network, topology in a
manner which is generally similar to that described above for
determining the intra-node topology for each node. Steps 102,
104, and 106, illustrated in Fig. 8 and further described
below, constitute the step 86 of Fig. 6 and generally
correspond for the network topology to the steps 92, 94, and 98
respectively of Fig. 6 as described above for the node
topology. To this end, the NMS 14 makes use of the lists M and
T, e.g. M1, M6, M7, and T7 as described above with reference to
Fig. 7, which it has concluded in the steps 96 and 100
represent inter-node optical connections for each node. A
further step 108 in Fig. 8 provides an optional additional
sanity check for the network.
Referring to Fig. 8, the NMS 14 initially determines
at the step 92 whether any node has two matching lists for
input and output; if so these are set aside with the conclusion
that the node is optically transparent to the optical channels
to which these lists relate. Using the example of Fig. 7, the
CA 02357988 2001-09-27
19
NMS 14 determines that the respective node has matching lists
T7 and M7 for two of its inter-node ports, and accordingly
these are set aside in the step 92, the node being optically
transparent for optical traffic between these ports.
In the step 104, the NMS 14 determines inter-node
paths by matching pairs of the remaining inter-node lists for
the network, one list of each pair being an output from one
node and the other list of the respective pair being an input
to another node of the network. Any unpaired lists are
concluded to represent errors (e.g. an unconnected optical
fiber) for which a warning or alarm is provided by the NMS 14
to an operator.
In the step 106, the matching lists, for input and
output inter-node optical channels, of any node concluded at
the step 102 as being optically transparent for this traffic,
are matched with lists of the determined inter-node optical
paths. The NMS 14 accordingly expands the determined inter-
node topology to include each optically transparent node in the
respective inter-node optical path, in a similar manner to that
described above for intra-node OBFs which do not add or drop
any optical channels. As in that case, in the event that the
same inter-node optical path includes a plurality of nodes
which are optically transparent to the traffic on the optical
path, then the order or sequence of these nodes in the optical
path is not determined by the NMS 14.
As indicated by the step 108, the NMS 14 can
optionally perform a further sanity check on the determined
topology of the network, that incoming and outgoing optical
paths are symmetrical with respect to the OBFs to which these
opposite-direction optical paths are connected, as is normally
the case in an optical network. Such a check facilitates
detecting an error of one node receiving optical signals from
CA 02357988 2001-09-27
another node but transmitting optical signals to a third node.
Other further and/or additional checks may also be performed by
the NMS 14; for example, the determined topology may be
compared with a previous topology of the same network to
5 identify any differences.
As described above, the CID detectors are provided at
the multiplexed ports of the optical filters, but this need not
be the case and the CID detectors can be provided additionally
or instead at other points in the optical paths. In addition,
10 as described above there is an initial process (the steps of
Fig. 6) of determining intra-node topology and hence ports of
each node, and a subsequent process (the steps of Fig. 8) of
determining the inter-node network topology, but again this
need not be the case.
15 For example, instead a CID detector may be provided
for each optical path entering or leaving each node, i.e. for
each node port, to provide a respective list of channels at
each node port. The steps 102, 104, and. 106 of Fig. 8 as
described above can be carried out in respect of these lists,
20 to determine the inter-node network topology without any
determination of the topology within each node.
Thus although particular embodiments of the invention
are described above, it can be appreciated that numerous
modifications, variations, and adaptations may be made without
departing from the scope of the invention as defined in the
claims.
