U.S. patent application number 14/457201 was filed with the patent office on 2014-11-27 for directionless optical architecture and highly available network and photonic resilience methods.
This patent application is currently assigned to Ciena Corporation. The applicant listed for this patent is Loudon T. BLAIR, Michael Y. FRANKEL. Invention is credited to Loudon T. BLAIR, Michael Y. FRANKEL.
Application Number | 20140348504 14/457201 |
Document ID | / |
Family ID | 41063134 |
Filed Date | 2014-11-27 |
United States Patent
Application |
20140348504 |
Kind Code |
A1 |
BLAIR; Loudon T. ; et
al. |
November 27, 2014 |
DIRECTIONLESS OPTICAL ARCHITECTURE AND HIGHLY AVAILABLE NETWORK AND
PHOTONIC RESILIENCE METHODS
Abstract
A directionless optical architecture is described for
reconfigurable optical add/drop multiplexers (ROADMs) and
wavelength selective switches (WSSs). The directionless
architecture utilizes a directionless wavelength switch coupled
between client devices and ROADMs/WSSs to eliminate the need to
hard-wire client devices to a wavelength division multiplexed (WDM)
network. Accordingly, client device connections can be
automatically routed without manual intervention to provide a
highly resilient network design which can recover route diversity
during failure scenarios. Additionally, the present invention
minimizes deployments of costly optical transceivers while
providing superior resiliency. Further, the present invention
couples the directionless optical architecture and associated
optical protection mechanisms with existing mesh restoration
schemes to provide additional resiliency.
Inventors: |
BLAIR; Loudon T.; (Severna
Park, MD) ; FRANKEL; Michael Y.; (Baltimore,
MD) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
BLAIR; Loudon T.
FRANKEL; Michael Y. |
Severna Park
Baltimore |
MD
MD |
US
US |
|
|
Assignee: |
Ciena Corporation
Hanover
MD
|
Family ID: |
41063134 |
Appl. No.: |
14/457201 |
Filed: |
August 12, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12045933 |
Mar 11, 2008 |
8849115 |
|
|
14457201 |
|
|
|
|
Current U.S.
Class: |
398/49 |
Current CPC
Class: |
H04Q 2011/0081 20130101;
H04J 14/025 20130101; H04Q 11/0005 20130101; H04J 14/0284 20130101;
H04J 14/0209 20130101; H04J 14/0217 20130101; H04J 14/0291
20130101; H04J 14/0246 20130101; H04J 14/0206 20130101; H04J
14/0213 20130101; H04J 14/0212 20130101; H04J 14/0294 20130101;
H04Q 2011/0047 20130101; H04J 14/0227 20130101; H04J 14/029
20130101; H04J 14/0295 20130101; H04Q 11/0062 20130101 |
Class at
Publication: |
398/49 |
International
Class: |
H04J 14/02 20060101
H04J014/02; H04Q 11/00 20060101 H04Q011/00 |
Claims
1. A directionless reconfigurable optical add/drop multiplexer
(ROADM) node for high availability between switch or router ports,
the node communicatively coupled to an optical network via at least
two links, the node comprising: a directionless switch connected to
each of the at least two links and configured to selectively route
a lightpath from a client device to any of at least two links for
protection thereof; wherein the directionless switch is located
between the at least two degrees and the client device, and wherein
the client device is not hard-wired to any of the at least two
links.
2. The node of claim 1, wherein the directionless switch is
configured with a single input/output for the client device and the
protection is 1:1 optical protection.
3. The node of claim 2, wherein the 1:1 optical protection includes
a 1.times.2 optical splitter and a 1.times.2 tail end switch
between the directionless switch and the client device.
4. The node of claim 2, wherein the 1:1 optical protection includes
a 1+1 optical broadcast and a 1.times.2 tail end switch between the
directionless switch and the client device.
5. The node of claim 1, wherein the directionless switch is
configured with dual inputs/outputs for the client device and the
protection is 1+1 or 1:1.
6. The node of claim 1, wherein the directionless switch is
configured with dual inputs/outputs for the client device with each
of the dual inputs/outputs configured for a maximum fill rate of
around 50% of traffic, and wherein, responsive to a failure
affecting one of the dual inputs/outputs, the other of the dual
inputs/outputs is set to 100% of the traffic using fast reroute,
and the directionless switch is configured to switch the one of the
dual inputs/outputs for restoration, and responsive to the
restoration, the dual inputs/outputs are configured for the maximum
fill rate of around 50% of the traffic.
7. The node of claim 1, wherein the client device has one
protection transceiver for up to N working transceivers, and
wherein each of the N working transceivers and the one protection
transceiver are connected to the directionless switch through an
optical splitter which is configured to switch to the one
protection transceiver for any of the N working transceivers that
fail.
8. The node of claim 1, wherein the directionless switch includes a
receive side and a transmit side, and wherein the at least two
links each comprise a wavelength selective switch connected to the
receive side and a multiplexer connected to the transmit side.
9. The node of claim 1, further comprising: a second directionless
switch connected to each of the at least two links, wherein the
second directionless switch provides redundancy to the
directionless switch, wherein the client device has dual
inputs/outputs each provided to one of the directionless switch and
the second directionless switch, and wherein the directionless
switch and the second directionless switch are configured to
selectively route lightpaths from the dual inputs/outputs to any of
the at least two links for protection thereof
10. The node of claim 1, wherein the directionless switch is a
combination of small-scale switches added based on bandwidth at the
node.
11. A directionless switch method for high availability between
switch or router ports, the method comprising: receiving at least
two lightpaths from a client device; and selectively routing the at
least two lightpaths to separate degrees of a reconfigurable
optical add/drop multiplexer (ROADM) node, wherein each of the
separate degrees provides a link to an optical network; wherein the
client device is not hard-wired to any of the separate degrees.
12. The method of claim 10, wherein the receiving is from a single
input/output for the client device and the protection is 1:1
optical protection.
13. The method of claim 10, wherein the receiving is from dual
inputs/outputs for the client device and the protection is 1+1 or
1:1.
14. The method of claim 10, wherein the receiving is from dual
inputs/outputs for the client device with each of the dual
inputs/outputs configured for a maximum fill rate of around 50%,
and wherein responsive to a failure affecting one of the dual
inputs/outputs, the other of the dual inputs/outputs is set to 100%
using fast reroute, and the method further comprises switching the
one of the dual inputs/outputs for restoration, and responsive to
the restoration, the dual inputs/outputs are configured for the
maximum fill rate of around 50%.
15. The method of claim 10, wherein the client device has one
protection transceiver for up to N working transceivers, and
wherein each of the N working transceivers and the one protection
transceiver are connected to a directionless switch through an
optical splitter which is configured to switch to the one
protection transceiver for any of the N working transceivers that
fail.
16. A client device method for high availability between switch or
router ports using a directionless architecture, the method
comprising: communicating traffic, on a first port and a second
port, with an adjacent client device over an optical network,
wherein the first port and the second port are selectively routed
to separate degrees at a reconfigurable optical add/drop
multiplexer (ROADM) node over the optical network, wherein the
first port and the second port are each not hard-wired to the
separate degrees, and wherein each of the separate degrees provides
a link to the optical network; responsive to a failure affecting
the first port, rerouting the traffic, from the first port, to the
second port; and responsive to recovery of the first port,
rerouting some of the traffic from the second port to the first
port.
17. The method of claim 16, wherein, prior to the failure and
subsequent to the recovery, the first port and the second port are
each configured for a maximum fill rate of around 50%.
18. The method of claim 16, wherein the directionless architecture
comprises a directionless switch connected to each of the separate
degrees and configured to selectively route a lightpath from the
first port and the second port to any of the separate degrees, for
protection thereof
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present non-provisional patent application/patent is a
continuation of U.S. patent application Ser. No. 12/045,933, filed
on Mar. 11, 2008, and entitled "DIRECTIONLESS OPTICAL ARCHITECTURE
AND HIGHLY AVAILABLE NETWORK AND PHOTONIC RESILIENCE METHODS," the
contents of which are incorporated in full by reference herein.
FIELD OF THE INVENTION
[0002] The present invention relates generally to optical networks.
More particularly, the present invention provides a directionless
optical architecture for reconfigurable optical add/drop
multiplexers (ROADMs) and wavelength selective switches (WSSs)
which can be utilized to provide highly available network, photonic
resiliency, and wavelength optimization.
BACKGROUND OF THE INVENTION
[0003] As point-to-point Internet Protocol (IP) flows increase in
bandwidth, core router connections are being driven to higher
capacities. Today, core router interfaces are starting to move from
10 Gbps to 40 Gbps while 100 Gbps connections are already in the
planning stage. With this increase in capacity comes a heightened
responsibility to maintain high availability service by minimizing
the time that these very expensive, high bandwidth connections are
out of service due to failure events or scheduled maintenance
activities.
[0004] Referring to FIGS. 1 and 2, a network 10 illustrates core
routers 12a,12b,12c with mesh connections directly over a
statically provisioned wavelength division multiplexed (WDM)
transport layer. The core routers 12a,12b,12c can include IP
routers with direct optical interfaces, such as 10 Gbps, 40 Gbps,
etc. In this example, each core router 12a,12b,12c connects
optically to a Wavelength Selective Switch (WSS) 14. The network 10
includes multiple WSSs 14 at various geographically-diverse
locations 16 where regeneration of signals may take place or where
other signals may be added to or dropped from the WDM line. The
WSSs 14 are configured to receive a client signal, such as from the
core routers 12a,12b,12c, and to provide a WDM line signal formed
by multiplexing multiple client signals, such as with an optical
multiplexer with optical filters (not shown). Each node can also
include other components (not shown), such as optical amplifiers,
dispersion compensation modules (DCM), and the like.
[0005] The WSSs 14 connect each of the various locations 16 in a
mesh configuration through optical fibers. Conventionally, the core
routers 12a,12b,12c are connected through the WSSs 14 statically,
and bandwidth on each wavelength connection is typically traffic
engineered to a predetermined capacity below the maximum possible
capacity, such as 50%, 40%, 20%, etc., so as to accommodate a layer
three initiated roll over of traffic from a failed link to a
working link upon a network failure, such as a fiber cut, equipment
failure on the WSSs 14, failure on the core router 12, and the
like. For example, in FIG. 1, the core router 12a is provisioned
over the WSSs 14 to connect to the core routers 12b and 12c with a
maximum of 50% traffic over each link such that in the event of a
failure, sufficient capacity is available to accommodate both
working and protected traffic.
[0006] In FIG. 2, a failure 18, e.g. a fiber cut, optical
transceiver failure, network maintenance event (note, a maintenance
event has the same effect as a failure), etc., is illustrated
between the core routers 12a and 12b causing the traffic to be
interrupted on this link, i.e. 0% fill. Here, traffic from the core
routers 12a and 12b is rolled to the links between core routers 12a
to 12c and from core routers 12b to 12c at layer three by the core
routers 12a,12b, i.e. the routers located at either end of the
failed connection. Now, while the link between core routers 12a to
12b is out of service, the network 10 is vulnerable to additional
failures on links between core routers 12a to 12c (isolating router
12a) and/or between core routers 12b to 12c (isolating router
12b).
[0007] While the architecture of the network 10 is reasonably
efficient from a capital equipment ("CAPEX") perspective, it raises
some challenges that can impact the operating expenses ("OPEX")
required to operate and maintain this network 10. One challenge is
associated with how the links between the routers 12a,12b,12c are
protected. While the link between routers 12a to 12b is down, the
network 10 core is operating in a dangerous condition whereby any
second failure could potentially isolate a region of the network
10. The Median Time to Repair (MTTR) becomes a critical parameter
in the calculation of service availability and the corresponding
service level agreements (SLAs) that can be offered to end user
clients. Providing the ability to provide a new (third) path
through the network 10 in the event of such a condition could help
to minimize the MTTR for the connection and maintain high
connection availability.
[0008] Another challenge is associated with the coordination of
network maintenance activities between operations personnel who are
responsible for the IP network, i.e. routers 12, and those
responsible for the underlying transport connections, i.e. WSSs 14.
Because core router 12 interfaces are directly associated with a
statically defined WDM lightpath across the network 10, it is not
possible to separate the two events. Without careful cooperation
between operations personnel, it is possible that simultaneous
maintenance could occur on links between routers 12a to 12b (by
transport) and between routers 12a to 12c (by IP) causing
unnecessary network disruption. Clearly, providing a mechanism to
reconfigure the IP or optical layers independently is
advantageous.
[0009] The use of optical cross connects (OXCs) based on an
electrical switch fabric provides one possible solution that could
provide optical layer re-configurability in the face of network
failure or planned maintenance. However, there is concern that the
cost of 40 G or 100 G interfaces required to support core router
connections is not as cost effective as 10 G and therefore should
be minimized throughout the transmission path, i.e. OXCs would
require additional 40 G or 100 G interfaces. Furthermore,
dedicating 40 G or 100 G modules on an OXC to aggregate flows of
data is wasteful of precious backplane and switch capacity.
[0010] Referring to FIGS. 3 and 4, a network 20 illustrates core
routers 12d,12e using optical 1+1 broadcast with tail-end
protection on optical links between adjacent core routers 12d,12e.
Here, the routers 12d,12e are connected through a single optical
transceiver on each router 12d,12e. In this example, a
unidirectional path is shown from the router 12d to the router 12e.
At the router 12d, an optical splitter 22 is configured to split an
output from a transceiver on the router 12d into two identical
signals with each signal separately provided to a different WSS 14.
At the router 12e, a tail end switch 24 is configured to receive
outputs from two different WSSs 14. The switch 24 is configured to
switch between WSSs 14 responsive to a condition, such as loss of
signal.
[0011] To date, this kind of protection has been implemented on the
short-reach link between core router interfaces and WDM
transceivers. This protection scheme is not designed to protect
against router 12d,12e or router port failure, however it does
provide resistance to transport layer failures associated with
optical layer components, i.e. the WSSs 14, and the optical fiber
itself
[0012] In FIG. 4, a failure 26 is illustrated on one link between
the routers 12e and 12d. Upon failure, protection decisions are
made locally, providing rapid (e.g., <50 ms) restoration times
and significantly reducing the amount of complex routing table
re-configuration at layer three. In this protection scheme, the
router 12d is broadcasting the same data signal to the router 12e
over two separate paths through WSSs 14. The splitter 22 is
unaffected by the failure 26, however the switch 24 is configured
to switch to the backup or protect path upon the failure 26. Also,
it is now possible to separate maintenance activities associated
with different IP and optical operations teams by moving the
physical path of the optical signal can be from the `primary` to
`backup` route.
[0013] However, because of the static nature of the optical
connectivity, it is not possible to reconfigure the optical layer
so as to restore diverse links between the end nodes. Instead,
after the link failure 26, the connection between the routers 12d
and 12e is now unprotected for as long as the damaged link is under
repair. For some carriers this is a significant issue. Depending on
the physical route of an optical fiber, the MTTR for the damaged
connection can be quite long (on the order of days to weeks). For
example, some fibers are routed through inhospitable terrain such
as over (or through) mountains or under lakes or seas where the
maintenance activity can involve lengthy procedures. In this case,
the carrier would like to re-establish a new `backup` route quickly
so as to restore diversity between end nodes and thus maintain the
promise of high availability to end user clients.
[0014] Referring to FIGS. 5 and 6, many carriers are now deploying
Wavelength Selective Switch (WSS) 14 technology into their networks
to form a Reconfigurable Optical Add-Drop Multiplexer (ROADM) 30
node. The WSS 14 (and ROADM 30)) provide all-optical wavelength
cross-connection functionality that supports lightpath
reconfigurability at the photonic level. The ability to redirect
lightpaths from one fiber direction to another direction makes the
WSS 14 a promising component of a solution to re-establish a backup
connections.
[0015] In FIG. 5, the ROADM 30 illustrates lightpath
reconfigurability for lightpaths 32 between different WSSs 14.
Note, in this configuration, the WSSs 14 may not require
connections to transceivers or regenerators, but rather may include
amplifiers, DCMs, multiplexers, etc. for all-optical pass-through
at the ROADM 30. Here, the lightpaths 32 can be redirected from one
output to any of the WSSs 14 (provided there is no wavelength
conflict).
[0016] In FIG. 6, an IP router 12 and Optical Transport Network
(OTN) platform 34 are connected to different WSSs 14. The OTN
platform 34 is illustrated as an example, and could also include a
SONET/SDH platform, an Ethernet platform, etc. Here, the add/drop
traffic from the lightpaths 32 is directly associated with a
specific direction (denoted by arrows 36). A shortcoming of ROADM
30 solutions as designed and implemented today is the fact that WDM
add/drop traffic is typically associated per direction (or fiber
degree) of the ROADM 30 node. The optical interface of the router
12, switch, platform 34, or the like wishing to communicate `north`
(i.e., in the direction of arrows 36) from a node must be
hard-wired to the WSSs 14 associated with that direction.
Therefore, while the pass-through traffic of today's ROADM 30 is
highly flexible, the static connectivity of the add/drop traffic
limits the flexibility of the end-to-end solution.
[0017] Referring to FIG. 7, ROADM 30 nodes are added at
intermediate junction nodes to the network 20 of FIGS. 3 and 4
illustrating core routers 12d, 12e using optical 1+1 broadcast with
tail-end protection on optical links between adjacent core routers
12d,12e. Adding the ROADM 30 to the core router 12d,12e
interconnection example provides optical pass through at
intermediate ROADM 30 nodes without the need for manual patching or
terminal regeneration. However, because the add/drop function is
dedicated on a per-direction basis, the solution with ROADM 30 is
no more capable of re-establishing a second backup connection as
the static network above.
[0018] Referring to FIG. 8, a conventional ROADM 40 illustrates the
limitations associated with current directional architectures. The
conventional ROADM 40 shows only drop-side connections for
illustration purposes, and those of ordinary skill in the art will
recognize that the ROADM 40 can also include add-side connections.
Connections 42 from receivers 44 are hard wired in the conventional
ROADM 40 and the intermediate ROADM 30 nodes in FIG. 7. A
receiver's 44 wavelength is fixed by a channel demultiplexer 46 to
a single value. Once plugged in, a transmitter cannot change
wavelength. As network grows, there may be lots of stranded ports,
and wavelengths cannot be dynamically re-optimized to reduce
blocking probability. Further, connections to the receivers 44
cannot be automatically altered due to the hard-wire connections 42
preventing rerouting to restore route diversity during
failures.
[0019] Accordingly, the downtime associated with any network
service (e.g., Ethernet, SONET/SDH, Fibre Channel, etc.) is
directly associated with the quality of a network service and the
associated Service Level Agreement (SLA) between a service provider
(i.e., carrier) and a client. In addition to the conventional
optical protection schemes described above, network elements (NEs)
with protection schemes such as SONET Bi-directional Line Switched
Ring/Uni-directional Path Switched Ring (BLSR/UPSR), SDH Multiplex
Section-Shared Protection Ring/Sub-network Connection Protection
(MSSPRing/SNCP), etc. have been developed as network `self-healing`
mechanisms.
[0020] More recently, mesh restoration has been implemented in
networks to improve service availability by providing access to
multiple backup paths. The ability to access more than one backup
path through the network increases the probability that service
will stay available to the end user, thus decreasing the average
downtime a network experiences over the course of a year.
[0021] One of the major challenges facing a number of network
operators today is a high incidence of fiber cuts occurring
randomly in the network. Such failures are a common occurrence in
developing nations such as India where significant new
infrastructure building is taking place (resulting in lots of
digging up of fiber cables). Because most of these carriers are
currently using ring protection methods to protect service, their
networks are only able to accommodate one fiber failure on a single
ring at any one time. High fiber failure probability therefore
leads to the isolation of network elements when two cuts occur
simultaneously on the same ring and therefore results in a
degradation of end-to end service.
[0022] To overcome this challenge, some carriers can geographically
partition their existing SONET/SDH ring networks into small
cascaded rings such that the probability of two fiber failures
occurring on the same ring is reduced. This helps increase service
availability but, in many cases, does not allow the carriers to
meet their target availability objective (particularly for high
value e.g. banking clients who demand `always on` service). They
are also now investigating the use of mesh restoration to increase
their service availability.
[0023] A challenge with both ring protection and mesh restoration
is the fact that fiber failures are statistically dependent upon
the distance between the switching nodes in the network. So, even
if mesh restoration is used, if the distance between mesh
restoration switch sites is too long, then there still exists an
increased probability that a service node can be isolated due to
simultaneous failures on each of the (multiple) links connected to
that node . . . thus losing service.
[0024] One approach used to increase resiliency performance and
availability is to combine SDH or SONET ring protection with SONET
or SDH mesh restoration. This capability can be combined through a
Virtual Line Switched Ring (VLSR) or SNCP protection plus backup
mesh restoration on an Optical switch platform. This clearly
provides the benefit of a deterministic 50 ms protection time plus
mesh restoration availability.
[0025] However, this typically needs to be implemented by a single
SDH or SONET vendor. Unfortunately, while perhaps possible, the
interaction of SDH/SONET between different vendor's equipment for
ring protection is highly complicated and not advised. For a number
of reasons including Data Communication Channel (DCC) transparency,
different use of (and response to) SONET/SDH overhead bytes, etc.
very little success has been achieved in the industry in the area
of SONET/SDH inter-working between different vendors. Consequently,
it does not make engineering or operational sense to operate
SDH/SONET rings between mesh restoration nodes belonging to one
vendor and SDH/SONET rings belonging to a second vendor. This has
been shown to be operationally challenging to engineer.
[0026] Also, in some network designs, it may not be cost effective
(or prudent from a traffic management perspective) to put a large
cross-connect with mesh restoration capabilities at every node in
the network. The use of a limited number of cross-connects plus
lower cost equipment in between may provide a more economic
solution.
BRIEF SUMMARY OF THE INVENTION
[0027] In various exemplary embodiments, the present invention
provides a directionless (i.e., a direction-independent) optical
architecture for reconfigurable optical add/drop multiplexers
(ROADMs) and wavelength selective switches (WSSs). The
directionless architecture utilizes a wavelength switch coupled
between client devices and ROADMs/WSSs to eliminate the need to
hard-wire client devices to a wavelength division multiplexed (WDM)
network. Accordingly, client device connections can be
automatically routed without manual intervention to provide a
highly resilient network design which can recover route diversity
during failure scenarios. Additionally, the present invention
minimizes deployments of costly optical transceivers while
providing superior resiliency. Further, the present invention
couples the directionless optical architecture and associated
optical protection mechanisms with existing mesh restoration
schemes to provide additional resiliency. Here, the present
invention provides mesh restoration, such as through SONET/SDH,
across a G.709/OTN enabled sequence of rings. The G.709 rings
provide high availability connectivity for the SONET/SDH
connections.
[0028] In an exemplary embodiment of the present invention, a
directionless optical system includes a client device; a
reconfigurable optical node including one or more degrees; and a
switch connected to each of the one or more degrees; wherein the
client device connects to the switch, and wherein the switch is
configured to route signals between the client device and the one
or more degrees. Optionally, the directionless optical system
further includes a splitter including an input port connected to a
transmitter on the client device and dual output ports connected to
the switch, wherein the splitter is configured to receive an input
from the transmitter on the input port and provide duplicate
signals of the input from the transmitter on the dual output ports;
and a tail-end switch including dual input ports connected to the
switch and an output port connected to a receiver on the client
device, wherein the tail-end switch is configured to receive
duplicate input signals from the switch on the dual input ports and
provide one of the duplicate input signals to the receiver. The
switch includes a first switch module and a second switch module
for redundancy, and wherein each of the dual output ports and dual
input ports connect separately to the first switch module and the
second switch module. The one or more degrees can includes two or
more degrees; wherein the first switch module and the second switch
module are configured to connect the duplicate signals on the dual
output ports to separate degrees of the two or more degrees; and
wherein the first switch module and the second switch module are
configured to connect the duplicate input signals on the dual input
ports to separate degrees of the two or more degrees. The two or
more degrees can also include an additional degree not utilized by
the dual output ports and the dual input ports; and wherein,
responsive to a failure on one of the separate degrees, the first
switch module and the second switch module are configured to
redirect one of the dual output ports and one of the dual input
from the one of the separate degrees to the additional degree to
provide route diversity.
[0029] Alternatively, the directionless optical system further
includes a first transceiver and a second transceiver on the client
device, wherein the first transceiver and the second transceiver
are connected to the switch; wherein the one or more degrees
includes two or more degrees; and wherein the switch is configured
to connect the first transceiver and the second transceiver to
separate degrees of the two or more degrees. The switch can include
a first switch module and a second switch module for redundancy,
and wherein the first transceiver is connect to the first switch
module and the second switch module is connected to the second
switch module. The two or more degrees can also include an
additional degree not utilized by the first transceiver and the
second transceiver; and wherein, responsive to a failure on one of
the separate degrees, the first switch module and the second switch
module are configured to redirect one of the first transceiver and
the second transceiver from the one of the separate degrees to the
additional degree to provide route diversity. Optionally, the
client device includes a router, and wherein the first transceiver
and the second transceiver utilize layer three protection
mechanisms. The reconfigurable optical node can include a
wavelength selective switch for each of the one or more degrees;
and a multi-channel fixed filter including one of a band-wide and
cyclic Arrayed Waveguide connected to the wavelength selective
switch and the switch.
[0030] Optionally, the directionless optical system further
includes a regenerator/wavelength converter connected to the
switch; wherein the switch is configured to connect a signal from
the one or more degrees to the regenerator/wavelength converter.
The one or more degrees can include two or more degrees, and
wherein the reconfigurable optical node includes a first wavelength
selective switch connected to a first set of one or more
demultiplexers and to a first set of one or more multiplexers; and
a second wavelength selective switch connected to a second set of
one or more demultiplexers and to a second set of one or more
multiplexers; wherein each of the first set of one or more
demultiplexers, the first set of one or more multiplexers, the
second set of one or more demultiplexers, and the second set of one
or more multiplexers include a connection to the switch. The switch
can include two or more optical switches, and wherein the first set
of one or more demultiplexers, the first set of one or more
multiplexers, the second set of one or more demultiplexers, and the
second set of one or more multiplexers are each wavelength
independent. Alternatively, the directionless optical system
further includes a wavelength division multiplex platform including
a transponder, the splitter, and the tail-end switch, wherein the
transponder connects to the client device; wherein the client
device includes mesh restoration with a hold-off timer configured
to allow the splitter and the tail-end switch to restore service
prior to initiated mesh restoration.
[0031] In another exemplary embodiment of the present invention, a
highly available directionless optical method includes providing a
first path and a second path through a network between a first
device and a second device, wherein the first device and the second
device are each dynamically connected to the first path and the
second path; upon a failure on the first path, restoring service
between the first device and the second device on the second path;
and reconfiguring the first path to a third path by dynamically
changing a connection from the first path on each of the first
device and the second device to the third path. Each of the first
path and the second path includes an optical network; wherein
dynamically connected includes a directionless optical switch
between each of the first path and the second path and each of the
first device and the second device. The highly available
directionless optical method further includes performing mesh
restoration at each of the first device and the second device
following a hold-off time period, wherein the hold-off time period
is configured to enable service restoration optically between the
first device and the second device.
[0032] In yet another exemplary embodiment of the present
invention, a highly available directionless optical network
includes a first device including an optical connection to a first
optical switch; a first optical platform connected to the first
optical switch, wherein the first optical platform includes
multiple degrees; a second device including an optical connection
to a second optical switch; a second optical platform connected to
the second optical switch, wherein the second optical platform
includes multiple degrees; and a plurality of interconnected
reconfigurable optical nodes between the first optical platform and
the second optical platform; wherein the first optical switch and
the second optical switch connect the first optical platform to the
second optical platform through a first path and a second path;
wherein the first path utilizes a first degree of the multiple
degrees on the first optical platform and a first degree of the
multiple degrees on the second optical platform; wherein the second
path utilizes a second degree of the multiple degrees on the first
optical platform and a second degree of the multiple degrees on the
second optical platform; and wherein the first optical switch is
configured to switch connections from one of the first path and the
second path to a third degree of the multiple degrees on the first
optical platform and a third degree of the multiple degrees on the
second optical platform. Optionally, the plurality of
interconnected reconfigurable optical nodes are configured in a
cascaded protection ring configuration between the first optical
platform and the second optical platform; and wherein the first
optical platform and the second optical platform are configured to
provide optical 1+1 protection. The first device and the second
device can include mesh restoration with a hold-off timer operable
to delay mesh restoration for a time period allowing the optical
1+1 protection to reestablish service. The highly available
directionless optical network can further include one or more
regenerators at one or more of the plurality of interconnected
reconfigurable optical nodes; wherein each of the one or more of
the plurality of interconnected reconfigurable optical nodes
include an optical switch connected between the one or more
regenerators and the one or more of the plurality of interconnected
reconfigurable optical nodes.
[0033] In yet another exemplary embodiment of the present
invention, a directionless reconfigurable optical add/drop node
includes a wavelength selective switch connected to a multi-channel
fixed filter, wherein the multi-channel fixed filter includes one
of a band-wide and cyclic Arrayed Waveguide; an optical switch
connected to the multi-channel fixed filter; and a client device
connected to the optical switch; wherein the wavelength selective
switch is configured to provide single channel selectivity across a
plurality of wavelengths; wherein the multi-channel fixed filter
receives one of the plurality of wavelengths within a multi-channel
range of the multi-channel fixed filter; and wherein the optical
switch routes one of the plurality of wavelengths to the client
device.
BRIEF DESCRIPTION OF THE DRAWINGS
[0034] The present invention is illustrated and described herein
with reference to the various drawings, in which like reference
numbers denote like method steps and/or system components,
respectively, and in which:
[0035] FIGS. 1 and 2 are diagrams of a network of core routers with
mesh connections directly over a statically provisioned wavelength
division multiplexed (WDM) transport layer;
[0036] FIGS. 3 and 4 are diagrams of a network of core routers
using optical 1+1 broadcast with tail-end protection on optical
links between adjacent core routers;
[0037] FIGS. 5 and 6 are diagrams of conventional Reconfigurable
Optical Add-Drop Multiplexer (ROADM) with Wavelength Selective
Switch (WSS) technology;
[0038] FIG. 7 is a diagram of the network in FIGS. 3 and 4 with
conventional ROADM nodes are added at intermediate junction
nodes;
[0039] FIG. 8 is a diagram of a conventional ROADM illustrating the
limitations associated with current directional architectures;
[0040] FIG. 9 is a diagram of a directionless architecture with a
directionless wavelength switch located between client devices and
a ROADM according to an exemplary embodiment of the present
invention;
[0041] FIGS. 10 and 11 are diagrams of a directionless architecture
with a directionless wavelength switch located between a router and
a ROADM to provide optical 1+1 protection from the router according
to an exemplary embodiment of the present invention;
[0042] FIG. 12 is a diagram of directionless architectures for
ROADMs with redundant directionless switches for multiple degrees
according to an exemplary embodiment of the present invention;
[0043] FIGS. 13-16 are diagrams of a network connecting two IP
routers with single ports and optical layer protection through
ROADMs configured with a directionless architecture according to an
exemplary embodiment of the present invention;
[0044] FIGS. 17-20 are diagrams of a network connecting two IP
routers with dual ports through ROADMs configured with a
directionless architecture according to an exemplary embodiment of
the present invention;
[0045] FIGS. 21-24 are diagrams of a network connecting three IP
routers with dual ports through ROADMs configured with a
directionless architecture according to an exemplary embodiment of
the present invention;
[0046] FIGS. 25-29 are diagrams of a high availability router
interconnect with two routers interconnected over an optical
network utilizing a directionless architecture according to an
exemplary embodiment of the present invention;
[0047] FIGS. 30-34 are diagrams of a high availability router
interconnect with two routers interconnected with dual ports over
an optical network utilizing a directionless architecture according
to an exemplary embodiment of the present invention;
[0048] FIGS. 35-37 are diagrams of a degree four ROADM node with
redundant directionless switches for a directionless architecture
according to an exemplary embodiment of the present invention;
[0049] FIG. 38 is a diagram of a directionless ROADM node with a
single large switch according to an exemplary embodiment of the
present invention;
[0050] FIG. 39 is a diagram of a ROADM configuration with multi-X
fixed filters according to an exemplary embodiment of the present
invention;
[0051] FIG. 40 is a diagram of a directionless ROADM node with one
or more small scale switches and multi-X fixed filters according to
an exemplary embodiment of the present invention;
[0052] FIG. 41 is a diagram of a network illustrating connections
between two offices utilizing a combination of mesh restoration and
optical ring restoration according to exemplary embodiment of the
present invention;
[0053] FIG. 42 is a diagram of the network from FIG. 41 with two
simultaneous fiber failures according to an exemplary embodiment of
the present invention; and
[0054] FIGS. 43-45 are diagrams of a network illustrating a
large-scale deployment of cross-connects with cascaded protection
rings according to an exemplary embodiment of the present
invention.
DETAILED DESCRIPTION OF THE INVENTION
[0055] In various exemplary embodiments, the present invention
provides a directionless optical architecture for reconfigurable
optical add/drop multiplexers (ROADMs) and wavelength selective
switches (WSSs). The directionless architecture utilizes a
directionless wavelength switch coupled between client devices and
ROADMs/WSSs to eliminate the need to hard-wire client devices to a
wavelength division multiplexed (WDM) network. Accordingly, client
device connections can be automatically routed without manual
intervention to provide a highly resilient network design which can
recover route diversity during failure scenarios. Additionally, the
present invention minimizes deployments of costly optical
transceivers while providing superior resiliency. Further, the
present invention couples the directionless optical architecture
and associated optical protection mechanisms with existing mesh
restoration schemes to provide additional resiliency. Here, the
present invention provides mesh restoration, such as through
SONET/SDH, across a G.709/OTN enabled sequence of rings. The G.709
rings provide high availability connectivity for the SONET/SDH
connections.
[0056] Advantageously, the present invention can provide
connections with high availability between switch or router ports.
The present invention can also be utilized with SONET/SDH/Optical
Transport Network (OTN) terminals and the like. As data rates
increase to 40 Gbps, 100 Gbps, and the like, the present invention
provides the potential for lower cost protection than electronic
switching equipment. However, the present invention maintains
carrier-grade protection requirements to offer superior
resiliency.
[0057] Referring to FIG. 9, a directionless architecture 50 is
illustrated with a directionless wavelength switch 52 located
between client devices, such as an IP router 12 and an OTN platform
34, and a ROADM 30 according to an exemplary embodiment of the
present invention. As described herein, directionless refers to a
direction independent architecture where connected ports can be
routed to any device or degree through the directionless wavelength
switch 52. Advantageously, the directionless wavelength switch 52
effectively removes the hard-wired connection from the devices
12,34 to the ROADM 30 enabling network flexibility in the ROADM 30.
In FIG. 9, each of the IP router 12 and OTN platform 34 can select
any direction in the ROADM 30 using the wavelength switch 52 to
reroute an input/output to/from a different WSS 14.
[0058] Connections from the IP router 12 and OTN platform 34 are to
the switch 52, i.e. not hard-wired to WSSs 14. The switch 52
provides connections to each WSS 14 in the ROADM 30. Based on the
switch 52 configuration, the IP router 12 and OTN platform 34 can
have their signals routed in different directions through the ROADM
30 without manual patching of connections.
[0059] For example, both the IP router 12 and OTN platform 34
include dual inputs/outputs to the directionless wavelength switch
52 for 1+1/1:1 or the like protection. A failure in the path or on
the equipment causes the dual inputs/outputs to switch. The switch
52 can be utilized to switch the failed port to another WSS 14 to
provide another route in the network ensuring resiliency during a
failure.
[0060] The directionless wavelength switch 52 can include a Micro
Electro-Mechanical Systems (MEMS), a liquid crystal, an inkjet, a
thermal mechanism, a non-linear mechanism, an acousto-optic
mechanism, and the like for the physical embodiment. As described
herein, the wavelength switch 52 can be scaled to variable port
sizes (i.e., depending on application size), and can also include
redundancy through multiple physical devices to ensure resiliency.
Generally, the directionless wavelength switch 52 is configured to
receive an input including one or more wavelengths, and to switch
the input to an output port based on provisioning. The
directionless wavelength switch 52 can switch at a wavelength level
or at a multiplexed wavelength level (i.e., groups of wavelengths).
Advantageously, the scaling requirements of the directionless
wavelength switch 52 are less than those required for prior art
designs that require larger port counts.
[0061] Referring to FIG. 10, a directionless architecture 50 is
illustrated with a directionless wavelength switch 52 located
between an IP router 12 and a ROADM 30 to provide optical 1+1
protection from the IP router 12 according to an exemplary
embodiment of the present invention. Here, the router 12 utilizes a
single input/output to connect to the switch 52, and the switch 52
can be configured to provide 1:1 optical protection by switching
the single input between two diverse paths through the ROADM 30
providing a working 54 and protection 56 path.
[0062] Referring to FIG. 11, a directionless architecture 50 is
illustrated with a directionless wavelength switch 52 located
between one or more optical regenerators 58 and a ROADM 30 to
provide optical regeneration of lightpaths 60 as needed according
to an exemplary embodiment of the present invention. The
directionless architecture 50 enables any lightpath 60 from client
to client to be routed through any arbitrary path through the
network. The one or more optical regenerators 58 can be used as
needed based on the lightpath 60 routing for optical regeneration
or for wavelength conversion (if there is wavelength blocking on a
path) based on low optical signal-to-noise ratio (OSNR) or the
like. This function will become increasingly of interest as data
rates increase and non-regenerated optical transmission distances
drop accordingly. Providing access to the dedicated bank of
regenerators 58 that are part of the transmission function and not
an integral part of a sub-wavelength switch or router will be both
efficient and cost effective. Additionally, the bank of
regenerators 58 can include wavelength conversion to switch a
channel from one wavelength to another. For example, this can
include tunable lasers of the regenerators 58.
[0063] Referring to FIG. 12, directionless architectures for a
ROADMs 62,64,66 are illustrated with redundant directionless
switches 52 for multiple degrees according to an exemplary
embodiment of the present invention. For carrier-grade availability
reasons, the present invention can use a redundant pair of
directionless wavelength switches 52 at each ROADM 62,64,66
location. Each individual directionless switch 52 connects to each
WSS 14 associated with each ROADM direction. For example, the WSS
14 can connect to a WDM multiplexer/demultiplexer which is
connected to the directionless switch 52. Thus, for a degree-2
ROADM 62, each directionless switch 52 connects to two directions
on the WSSs 14. For a degree-3 ROADM 64, each directionless switch
52 connects to three directions on the WSSs 14. For a degree-4
ROADM 66, each directionless switch 52 connects to four directions
on the WSSs 14. Note, the present invention can be utilized with
any arbitrary number of degrees.
[0064] Referring to FIGS. 13-16, a network 70 includes two IP
routers 12d,12e connected through ROADMs 62 64,66 configured with a
directionless architecture according to an exemplary embodiment of
the present invention. The network 70 is similar to the network 20
in FIGS. 3-4 and 7, but includes the directionless architecture of
the present invention with redundant directionless switches 52 at
each ROADM 62,64,66 site.
[0065] In this example, the routers 12d,12e are connected through a
single optical transceiver on each router 12d,12e. For illustration
purposes, a unidirectional path is shown from the router 12d to the
router 12e. At the router 12d, an optical splitter 22 is configured
to split an output from a transceiver on the router 12d into two
identical signals with each signal separately provided to redundant
switches 52 at a ROADM 64 node. The redundant switches are
connected to each of three WSS 14 at the ROADM 64 node. At the
router 12e, a tail end switch 24 is configured to receive outputs
from redundant switches 52 connected to three different WSSs 14 at
a ROADM 64 node. The switch 24 is configured to switch between the
redundant switches 52 responsive to a condition, such as loss of
signal.
[0066] Advantageously, introducing the directionless wavelength
switch 52 functionality on top of a WSS layer provides the
opportunity to achieve high availability connectivity as is
illustrated in the network 70 through an exemplary fault condition
72 in FIGS. 14-16. As illustrated in this example with the splitter
22 and tail end switch 24 for layer one optical tail-end
protection, connecting over a WSS layer is a flexible extension to
what some carriers are already doing today over static WDM links.
Each path of a 1+1 broadcasted router signal is switched through
each of the redundant directionless wavelength switches 52 and
forwarded to separate directions out of the ROADM node 64.
[0067] Note that, in this case, the expensive WDM interface
associated with the very high speed router connections is
integrated into the core router 12d,12e device and that no other
costly very high speed interfaces exist in the path. For example,
the core router 12d,12e can include WDM interfaces, and the core
router 12d,12e can alternatively include short-reach interconnects
to a WDM transceiver. Again, upon link failure, rapid `tail-end`
protection switching (with the switch 24) provides fast optical
layer protection without the need for router 12e reconfiguration.
Additionally, the network 70 can provide an alternative `third`
path in the network, the direction-less switch 52 associated with
the now failed link can reconfigure to create a new backup
path.
[0068] In FIG. 14, the network 70 is illustrated after the fault
condition 72 is experience on a working route 74. Accordingly, the
tail end switch 24 is configured to switch to a protect route 76 to
receive a protect signal transmitted by the splitter 22 on a
diverse path. Here, the working path 74 is now down, and the
network 70 is running unprotected only on the protect route 76.
[0069] In FIG. 15, the network 70 is illustrated with a third
alternative route 78 between the routers 12d,12e. Accordingly, the
network 70 can now provide connection diversity between the routers
12d,12e even though the original working route 74 is down. Here,
the directionless wavelength switch 52 can be configured to route
the lines from the working route 74 to the third alternative route
78. This is done automatically at the switches 52 without the need
for manual patching of connections since there is no hard-wired
connection from the routers 12d,12e to the WSSs 14. Further,
because the new path 78 re-forms the 1+1 broadcast connection
between the head-end splitter 22 and tail-end protection switch 24,
rapid optical protection capability is regained. Thus, upon a
second failure, a sub-50 ms protection switch is guaranteed.
[0070] In FIG. 16, the network 70 is illustrated with regenerators
58 at a WSS 64 node on the third alternative route 78. This
illustrates the functionality depicted in FIG. 11. Here, on the
third alternative route 78, a ROADM node 64 is configured to route
the third alternative route 78 to a regenerator 58. Regeneration
may be required on a long optical link, or wavelength conversion if
the wavelength on route 78 is being used on a particular span. At a
ROADM node 64, the directionless wavelength switch 58 provides
flexible access to a pool of optical regenerators 58 when
necessary. Note, the regenerators 58 may include tunable wavelength
transmitters allowing any wavelength to be utilized. Tunable
wavelength transmitters are configured to set a center wavelength
of a transmitted signal to any particular value within a set
range.
[0071] Referring to FIGS. 17-20, a network 80 includes two IP
routers 12a,12b configured through ROADMs 62,64,66 configured with
a directionless architecture according to an exemplary embodiment
of the present invention. The network 80 includes the directionless
architecture of the present invention with redundant directionless
switches 52 at each ROADM 62,64,66 site.
[0072] Not all carriers choose to implement optical layer
protection between adjacent core router ports, such as in FIGS.
13-16. As described earlier, when failures occur, the IP layer
reorganizes IP flows using techniques such as IP/MPLS
(Multi-Protocol Label Switching) Fast Re-route. FIGS. 17-20 show
the same core router connectivity as described earlier for the
static WDM scenario but with a flexible WSS photonic layer.
However, as in the previous example, expensive WDM interfaces may
be integrated into the core routers to reduce transmission
costs.
[0073] In FIG. 17, each router 12a,12b includes two optical
transceivers connected to the redundant directionless switches 52.
At the router 12a,12b, the optical transceivers can be configured
to each provide working traffic with a fill rate of around 50%. The
unused capacity can be used for layer three rerouting between the
routers 12a,12b if a link 82,84 is down. In FIG. 18, the network 80
is illustrated after a fault condition 86 is experience on the link
82. Accordingly, each router 12a,12b is configured to utilize layer
three mechanisms to reroute the 50% fill traffic from link 82 onto
link 84. Here, the link 82 is now down, and the network 80 is
running unprotected with 100% fill on the link 84.
[0074] In FIG. 19, the network 80 is illustrated with a third link
88 between the routers 12a,12b. Accordingly, the network 80 can now
provide connection diversity between the routers 12a,12b even
though the first link 82 is down. Here, the directionless
wavelength switch 52 can be configured to route the lines from the
link 82 to the third link 88. This is done automatically at the
switches 52 without the need for manual patching of connections
since there is no hard- wired connection from the routers 12a,12b
to the WSSs 14. The link 88 can now be provisioned as a working
link and approximately 50% of the traffic from link 84 can be moved
to link 88 providing excess capacity on a diverse link for
protection.
[0075] In FIG. 20, the network 80 is illustrated with regenerators
58 at a WSS 64 node on the third link 88. This illustrates the
functionality depicted in FIG. 11. Here, on the third link 88, a
ROADM node 64 is configured to route the third link 88 to a
regenerator 58. Regeneration may be required on a long optical
link, or wavelength conversion if the wavelength on link 88 is
being used on a particular span. At a WSS node 64, the
directionless wavelength switch 58 provides flexible access to a
pool of optical regenerators 58 when necessary. Note, the
regenerators 58 may include tunable wavelength transmitters
allowing any wavelength to be utilized. Tunable wavelength
transmitters are configured to set a center wavelength of a
transmitted signal to any particular value within a set range.
[0076] Referring to FIGS. 21-24, a network 90 includes three IP
routers 12a,12b,12c configured through ROADMs 62, 64,66 configured
with a directionless architecture according to an exemplary
embodiment of the present invention. The network 90 is similar to
the network 10 in FIGS. 1-2, but includes the directionless
architecture of the present invention with redundant directionless
switches 52 at each site 62,64,66. As described herein, the network
90 utilizes IP layer protection to reorganize IP flows using
techniques such as IP/MPLS Fast Re-route. As such, links 92,94
between routers 12a,12b,12c are provisioned with approximately 50%
fill.
[0077] In FIG. 21, each router 12a,12b,12c includes two optical
transceivers connected to the redundant directionless switches 52.
The switches 52 are configured to direct signals from the two
optical transceivers to different WSSs 14. At the router
12a,12b,12c the optical transceivers can be configured to each
provide working traffic with a fill rate of around 50%. The unused
capacity can be used for layer three rerouting between the routers
12a,12b,12c if a link 92,94 is down. In FIG. 22, the network 90 is
illustrated after a fault condition 96 is experience on the link
92. Accordingly, each router 12a,12b is configured to utilize layer
three mechanisms to reroute the 50% fill traffic from link 92 onto
link 94. Here, the link 92 is now down, and the network 90 is
running unprotected with 100% fill on the link 94.
[0078] In FIG. 23, the network 90 is illustrated with a third link
98 between the routers 12a,12b. Accordingly, the network 90 can now
provide connection diversity between the routers 12a,12b even
though the first link 92 is down. Here, the directionless
wavelength switch 52 can be configured to route the lines from the
link 92 to the third link 98. This is done automatically at the
switches 52 without the need for manual patching of connections
since there is no hard-wired connection from the routers 12a,12b to
the WSS 14. The link 98 can now be provisioned as a working link
and approximately 50% of the traffic from link 94 can be moved to
link 98 providing excess capacity on a diverse link for
protection.
[0079] In FIG. 24, the network 80 is illustrated with regenerators
58 at a ROADM 66 node on the third link 98. This illustrates the
functionality depicted in FIG. 11. Here, on the third link 98, a
ROADM node 66 is configured to route the third link 98 to a
regenerator 58. Regeneration may be required on a long optical
link, or wavelength conversion if the wavelength on link 98 is
being used on a particular span. At a ROADM node 66, the
directionless wavelength switch 58 provides flexible access to a
pool of optical regenerators 58 when necessary. Note, the
regenerators 58 may include tunable wavelength transmitters
allowing any wavelength to be utilized. Tunable wavelength
transmitters are configured to set a center wavelength of a
transmitted signal to any particular value within a set range.
[0080] Referring to FIGS. 25-29, a high availability router
interconnect 100 is illustrated with two routers 12f,12g
interconnected over an optical network 102 utilizing a
directionless architecture according to an exemplary embodiment of
the present invention. FIGS. 25-29 provide a different schematic
view of the same functionality described herein in previous
diagrams, focusing at interconnections at each node. The
interconnect 100 includes a degree three ROADM node 64 and a degree
four ROADM node 66, such as described in FIG. 12. Each ROADM node
64,66 includes multiple WSSs 14 which can connect to WDM
multiplexers, demultiplexers, optical amplifiers, DCMs, and the
like. A redundant directionless switch 52 is included before each
ROADM node 64,66.
[0081] FIG. 25 illustrates an equipment configuration of the
interconnect 100. Each router 12f,12g includes a single optical
transceiver, such as a 10 G, 40 G, 100 G, or the like transceiver.
The single optical transceiver at the router 12f is connected to a
1.times.2 optical splitter 22 which is configured to replicate the
signal and provide the signal to two separate switches in the
redundant directionless switch 52. The switch 52 provides the
duplicate signals to separate WSSs 14 at the ROADM node 64 for
transmission over the optical network 102.
[0082] The ROADM node 66 receives the duplicate signals at separate
WSSs 14 after transmission over the optical network 102. The
outputs from the various WSSs 14 are connected to another redundant
directionless switch 52 which connects to a 1.times.2 tail-end
switch 24. The tail-end switch 24 is configured to receiver both
duplicate signals and to provide one output to the optical
transceiver on the router 12g. The tail-end switch 24 is configured
to switch between signals based on a predetermined condition, such
as loss-of-signal, alarm indication signal, signal degrade, and the
like.
[0083] FIG. 26 illustrates a work path 104 and a protect path 106
through the optical network 102. The duplicate signals are sent on
each of the paths 104,106 based on a 1+1 optical broadcast through
the optical splitter 22. Advantageously, the interconnect 100
minimizes expensive optical transceivers on the routers 12f,12g
while maintaining a highly available architecture through optical
1+1 protection and route diversity.
[0084] FIG. 27 illustrates a rapid protection response of the
optical 1+1 protection based on a fault 108 on the working path
104. The router 12f does not implement any action here since it is
sending duplicate signals through the splitter 22. Instead, at the
router 12g, the tail-end switch 24 realizes the fault 108 and
performs a switch to the protect path 106 restoring connectivity
between the routers 12f,12g.
[0085] FIG. 28 illustrates a reconfiguration to provide a third
path 110 to maintain route diversity during the fault 108 on the
working path 104. Here, the switch 52 at the router 12f is
configured to move the signal from the working path 104 to another
WSS 14 on the third path 110. Also, the switch 52 at the router 12g
is also configured to move the output from the working path 104 to
the third path 110. Now, the interconnect 100 maintains route
diversity despite a failure on the working path 104.
[0086] Advantageously, the optical 1+1 and directionless
architecture provide protection and diversity despite faults while
minimized transceiver costs. Note, as rates increase to 40 G, 100
G, etc., the optical transceiver costs dominate capital expense.
However, the optical 1+1 does not protect against transceiver
failures at the routers 12f,12g. FIG. 29 illustrates a 1:N
protection mechanism 112 for optical transceivers on the routers
12f,12g. 1:N means one protection transceiver can provide equipment
redundancy for up to N working transceivers. Note, the 1:N
protection mechanism 112 for optical transceivers can also be
utilized on Ethernet platforms, SONET/SDH platforms, OTN platforms,
etc.
[0087] At the router 12f, each working and protect transceiver is
connected to the switch 52 through an optical splitter 22. In the
event a working transceiver fails, the protect transceiver is
configured to take over transmitting the working transceiver's
signal through the optical splitter 22 and the switch 52 and onto
the optical network 102. Advantageously, the 1:N protection
mechanism 112 minimizes transceiver costs by only using one
protection transceiver for N working transceivers. For example, the
1:N protection mechanism 112 can be utilized with mesh restoration
to provide similar resiliency as provided with ring configurations
while providing better efficiency than rings.
[0088] Referring to FIGS. 30-34, a high availability router
interconnect 120 is illustrated with two routers 12h,12i
interconnected with dual ports 122 over an optical network 102
utilizing a directionless architecture according to an exemplary
embodiment of the present invention. Each of the routers 12h,12i
include the dual ports 122 with each port connected to a separate
switch in the redundant directionless switch 52. The switch 52
provides signals from each port to separate WSS 14 at the ROADM
node 64 for transmission over the optical network 102. FIG. 30
illustrates an equipment configuration of the interconnect 120.
[0089] The ports 122 can each be configured to carry working
traffic between the routers 12h,12i. FIG. 31 illustrates a link
sharing on the ports 122, such as utilizing MPLS Traffic
Engineering (MPLS-TE). Here, two paths 124,126 each are configured
to provide traffic between the routers 12h,12i with excess capacity
to provide for rerouting in the event of a failure on one of the
paths 124,126. Note, this configuration provides both route and
transceiver protection.
[0090] FIG. 32 illustrates a rapid protection response of the
interconnect 120. Here, the first path 124 has a failure 128 on it.
After the failure 128, all traffic from the path 124 is rerouted to
the path 126. For example, this can be accomplished using MPLS-FRR
(fast reroute). Now, the path 126 is utilized up to full capacity
leaving no additional route protection in the interconnect 120.
FIG. 33 illustrates a third path 130 for restoring path diversity
while the path 124 is down. The third path 130 is configured by
rerouting the optical transceiver from the first path 124 using the
switch 52 to the third path 130. FIG. 34 illustrates a 1+1/1:1
protection mechanism 132 on the dual port transceivers 132
according to an alternate embodiment of the present invention.
1+1/1:1 protection 132 provides layer one protection against faults
at both the dual port transceivers 122 and on the paths.
[0091] Referring to FIGS. 35-37, a degree four ROADM node 66 is
illustrated with redundant directionless switches 52 for a
directionless architecture according to an exemplary embodiment of
the present invention. The ROADM node 66 is illustrated with the
optical 1+1 protection scheme formed with a splitter 22 and a
tail-end switch 24. An optical transceiver 140 is connected to the
splitter 22 on a transmitter and to the tail-end switch 24 on the
receiver. Note, the optical transceiver 140 can be in a router,
switch, optical cross-connect, WDM transponder, and the like, i.e.
although the present invention has been illustrated herein with
regard to an IP router, those of ordinary skill in the art will
recognize that the present invention can equally apply to any
device utilizing an optical transceiver and the like.
[0092] The ROADM node 66 includes two redundant directionless
switches 52 each includes two optical switches 142. The separate
optical switches 52 provide redundancy at the ROADM node 66, i.e. a
failure of one optical switch 52 is not a single point of failure.
Each of the optical switches 142 are separated for receive and
transmit sides (denoted as 142a for receive and 142b for transmit)
of the transceiver 140. The receive side switches 142a receive
inputs from demultiplexers 144. The demultiplexers 144 receive
outputs from four WSS 146 devices. The WSSs 146 receive inputs
external to the node 66 from optical fibers. As described herein,
the WSSs 146 are configured to dynamically route one or more
wavelengths to each demultiplexer. The receive side switches 142a
each provide an output to the tail-end switch 24.
[0093] The transmit side switches 142b each receive an output from
the splitter 22, and connect to multiple multiplexers 148 which are
configured to multiplex one or more outputs from the switches 142b.
Outputs from the multiplexers 148 can be combined with pass-through
signals 150 from the WSSs 146 for output from the node 66. For
example, the multiplexer 148 outputs and pass-through signals 150
can be combined with a multiplexer 152. Additionally, those of
ordinary skill in the art will recognize that this degree four
configuration can also be extended to different degree
configurations as are known in the art.
[0094] FIG. 36 illustrates an exemplary path through the ROADM node
66 for dual transmit/receive signals 160,162. In this
configuration, the four inputs/outputs from the ROADM node 66 are
labeled as East, West, North, and South. The transceiver 140
transmits a signal to the splitter 22 which provides a dual
transmit signal 160 to each of the transmit switches 142b. The
first transmit switch 142b sends one of the dual transmit signals
160 in the North direction, and the second transmit switch 142b
sends the other of the dual transmit signals 160 in the East
direction. The WSSs 146 for North and East directions receive dual
receive signals 162 and provide each to one of the receive switches
142a. Each of the receive switches 142a connect the dual receive
signal 162 to the tail-end switch 24.
[0095] FIG. 37 illustrates an alternative configuration of the
ROADM node 66 with one of the redundant directionless switches 52
configured as a combination of multiple small-scale switches 164.
Advantageously, the small-scale switches 164 minimize capital
expense allowing a configuration which fits the actual deployment
size of the ROADM node 66, and enables the ROADM node 66 to scale
as bandwidth is required.
[0096] Referring to FIG. 38, a directionless ROADM node 180 is
illustrated with a single large switch 182 according to an
exemplary embodiment of the present invention. The ROADM node 180
only shows a drop side for illustration purposes, and those of
ordinary skill in the art will recognize that the ROADM node 180
can include an add side with similar components. The ROADM node 180
is of degree two with an East and North direction 184,186. Each
direction 184,186 includes a WSS 146 which receives an input 188
and provides an express output 190. The input 188 is received from
a node external to the node 180, such as over optical fiber, and
the output 190 transmits from the node 180 to another node.
[0097] The WSS 146 is configured to reconfigurably drop different
wavelengths from the input 188 to a channel demultiplexer 192. The
demultiplexer 192 is configured to separate one or more channels.
In this exemplary embodiment, outputs from the various
demultiplexers 192 are connected to the switch 182. The switch 182
provides a directionless architecture similar to the switch 52
described herein. The switch 182 can route outputs from any of the
demulitplexers 192 to any receiver 194. This avoids the need to
hard-wire receivers 194 to the demultiplexers 192 allowing path
changes as described herein. Note, the switch 182 requires full
capacity up from for all receiver 194 connections. This could
require a lot of demultiplexers 192 and significant wiring upon
initial deployment.
[0098] Referring to FIG. 39, a ROADM configuration 200 is
illustrated with multi-X fixed filters 202 according to an
exemplary embodiment of the present invention. The ROADM
configuration 200 is similar to FIG. 38, but uses the multi-X fixed
filters 202 instead of the single X demulitplexers 192. For
example, the multi-X fixed filters 202 can include band splitter or
cyclic arrayed waveguide gratings (AWG). This configuration
provides fewer unused ports as the network grows since each port of
the filters 202 can provide multiple wavelengths. For example, each
filter 202 can provide a 4-channel passband on each of 10 ports to
provide a total capability of 40-channels for each filter 202. The
configuration 200 can leverage wavelength tunablity on optical
transceivers to re-optimize a network and reduce wavelength
blocking probability.
[0099] Advantageously, the multi-channel (either band-wide or
cyclic AWG) fixed filters under the WSS layer provide the WSS 146
single channel selectivity. The ROADM configuration 200 still
preserves some channel wavelength tunability, but with the cost of
fixed filters. The ROADM configuration 200 also allows a more
scalable growth in the switch size that gets used in the
directionless architecture. The ROADM configuration 200 provides a
balance between tunability and cost. Tunability is available within
the ranges of the filters 202 allowing wavelength assignments to be
dynamically changed to re-optimize a network and reduce wavelength
blocking probability.
[0100] Referring to FIG. 40, a directionless ROADM node 220 is
illustrated with one or more small scale switches 222 and multi-X
fixed filters 202 according to an exemplary embodiment of the
present invention. The ROADM node 220 only shows a drop side for
illustration purposes, and those of ordinary skill in the art will
recognize that the ROADM node 220 can include an add side with
similar components. The ROADM node 220 is of degree two with an
East and North direction 224,226. Each direction 224,226 includes a
WSS 146 which receives an input 188 and provides an express output
190. The input 188 is received from a node external to the node
180, such as over optical fiber, and the output 190 transmits from
the node 180 to another node.
[0101] In this exemplary embodiment, receivers 194a,194b can be
either directional or directionless. For example, receiver 194a can
be initially connected directly to the filter 202 for a directional
configuration. However, the node 220 can evolve to a directionless
configuration through the addition of the switch 220 and connecting
the receiver 194b directly to the switch 222 instead of the filter
202. Additionally, extra switches 222 can be added to provide
efficient growth and added redundancy.
[0102] Referring to FIG. 41, a network 300 illustrates connections
between two offices 302,304 utilizing a combination of mesh
restoration and optical ring restoration according to an exemplary
embodiment of the present invention. Each office 302,304 includes a
cross-connect 306 and a DWDM (dense-WDM) platform 308. The
cross-connect 306 can include an optical switch or multi-service
platform configured to provide SONET/SDH/OTN switching, and the
DWDM platform 308 is configured to transmit/receive multiple
wavelengths over an optical network 310. Alternatively, the
cross-connect 306 can also be an IP router, Ethernet switch, or the
like.
[0103] The cross-connects 306 are configured to provide mesh
restoration, such as using ITU-T Automatically Switched Optical
Network (ASON), IETF Generalized Multi-Protocol Label Switching
(G-MPLS) also known as Automatic Switched Transport Network (ASTN),
or the like. The DWDM platforms 308 are configured to provide
optical 1+1 protection such as described herein with the splitter
22 and tail-end switch 24.
[0104] Advantageously, the present invention combines the use of
mesh restoration and optical ring protection (i.e., not SONET/SDH)
to achieve super high availability performance. The optical ring
protection can utilized G.709/OTN to encapsulate SONET/SDH signals
while maintaining the mesh restoration on the underlying SONET/SDH
signals. Additionally, G.709/OTN provides a header with operations,
administration, maintenance, and provisioning (OAM&P)
capabilities on the optical ring, such as determining fault
conditions. The G.709/OTN optical rings provide high availability
connectivity for the mesh SONET/SDH connections.
[0105] The cross-connect nodes 306 that participate in mesh
restoration are connected together using cascaded protection rings
312,314 based on the DWDM platform 308. The use of cascaded
protection rings 312,314 provides protection against a single fiber
failure within the context of a small geographical domain. Thus,
multiple fiber failures may exist simultaneously (on different
rings 312,314) between a pair of mesh restoration nodes without the
need to implement mesh restoration.
[0106] The present invention utilizes a protected transparent ring
solution between mesh restoration nodes, such as cross-connects
306. These transparent rings 312,314 can include simple optical
1+1/tail-end protection of optical signals over fiber or WDM or, if
a managed solution is preferred, then the rings can be defined
using OTN/G.709 framing. For example, DWDM platforms 308 can be
configured to frame incoming signals from the cross-connect 306
using G.709. This provides complete transparency while providing
OAM&P at the DWDM layer. The G.709 framing can be utilized to
indicate a protection switch at the DWDM layer.
[0107] In this scenario, the connection between a pair of
cross-connect nodes 306 simply looks like a highly available SDH or
SONET connection. The cross-connect nodes 306 communicate as if
each is an adjacent SONET or SDH line or multiplex section
terminating piece of equipment. Signaling and routing protocols for
mesh restoration can communicate unimpeded across the chosen Data
Communications Channel (DCC). The present invention cascades
multiple optical protection rings 312,314 and combines the
protection attributes of these rings 312,314 with mesh restoration
to achieve a higher availability solution.
[0108] For the present invention to work, a hold-off timer is
required at each cross-connect node 306. In the event of a failure,
the cross-connect node 306 detects the failure and attempt to
recover service. In this scenario, the cross-connect node 306 must
wait to confirm that the intermediate 1+1 optical layer protection
has not recovered the service before implementing its own recovery
efforts. Because the proposed 1+1 optical protection is based on
tail-end protection, it is expected to be rapid so the
cross-connect node 306 hold off timer can be quite short, i.e. to
ensure sub-50 ms restoration.
[0109] Referring to FIG. 42, the network 300 is illustrated with
two simultaneous fiber failures 320,322 according to an exemplary
embodiment of the present invention. Based on the hold-off timer,
the cross-connects 306 detect the failures 320,322 but wait to
implement mesh restoration. During this wait, the optical 1+1
protection is implemented after the DWDM platforms 308 detect the
failures within each ring 312,314 and provide 1+1 switching.
Accordingly, the cross-connects 306 have service restored prior to
the hold-off timer expiring, and therefore do not implement mesh
restoration. The DWDM platforms 308 can utilize OTN/G.709 to detect
the fault on the rings 312,314. After detecting the failure 320 on
the first ring 312, the DWDM platforms 308 switch on ring 312.
After detecting the failure 322 on the second ring 314, the DWDM
platforms 308 switch on the ring 314.
[0110] Referring to FIGS. 43-45, a network 340 illustrates a
large-scale deployment of cross-connects 306a-e with cascaded
protection rings 341-352 according to an exemplary embodiment of
the present invention. FIG. 43 illustrates an equipment
configuration of the network 340 with multiple DWDM platforms 308
forming the cascaded protection rings 341-352. Note, the rings
341-352 can include other DWDM platforms 308 and/or regenerators,
optical amplifiers, and the like. For illustration purposes, only
head-end nodes are illustrated on the rings 341-352.
[0111] FIG. 44 illustrates multiple fiber failures 360 on the
various cascaded protection rings 341-352. A logical view 362
illustrates how the cross-connects 306a-e view the rings 341-352,
i.e. as a pipe. A physical view 364 illustrates the physical
deployment of the cross-connects 306a-e and the rings 341-352. From
the cross-connects 306a-e viewpoint, the multiple failure 360 are
ignored since the optical 1+1 recovers a connection 370 before
expiration of the hold-off timer in the cross-connects
306e,306a,306b (i.e., the cross-connects affected by the failures
360). In the physical view 364, rings 341-343,352 perform 1+1
switching to recover based on the multiple failures 360.
[0112] FIG. 45 illustrates a second failure 360 on the ring 343
following the failures 360 in FIG. 44. Here, the connection 370 is
now down since there is no fiber path on the rings 352,342-344
between cross-connect 306e,306a,306b. Now, the cross-connects
306e,306a,306b detect the failure and implement the hold-off timer
again. However, the hold-off time expires without restoration of
the connection 370. Accordingly, the cross-connects 306a-e
implement mesh restoration to provide a connection 372 between
cross-connects 306e,306d,306c,306b.
[0113] Although the present invention has been illustrated and
described herein with reference to preferred embodiments and
specific examples thereof, it will be readily apparent to those of
ordinary skill in the art that other embodiments and examples may
perform similar functions and/or achieve like results. All such
equivalent embodiments and examples are within the spirit and scope
of the present invention and are intended to be covered by the
following claims.
* * * * *