Non-blocking, scalable optical router architecture and method for routing optical traffic
First Claim
1. A router coupled to a plurality of data lines, comprising:
- a plurality of egress edge units, wherein each egress edge unit is coupled to at least one egress port;
a plurality of ingress edge units, wherein each ingress edge unit receives a plurality of optical data packets, each optical data packet is destined for a destination port at one of the plurality of egress edge units, aggregates the plurality of optical data packets into a plurality of super packets wherein each super packet comprises optical data packets intended for a particular destination egress edge unit and is to be routed to that particular destination egress edge unit;
an optical switch fabric that receives the plurality of super packets from the plurality of ingress edge units and routes each super packet through the optical switch fabric to the particular destination egress edge unit for which the super packet is intended, and further wherein the routing through the optical switch fabric is performed in a non-blocking manner; and
a core controller that controls the arrival of the plurality of super packets at the optical switch fabric so as to avoid contention among the plurality of super packets between the optical switch fabric and the plurality of egress edge units; and
wherein the plurality of egress edge units receive the plurality of super packets, de-aggregate the plurality of super packets into the optical data packets, and transmit each of the plurality of optical data packets to at least one of the at least one egress port.
11 Assignments
0 Petitions
Accused Products
Abstract
A system and method for providing non-blocking routing of optical data through a telecommunications router that allows full utilization of available capacity. The router includes a number of data links that carry optical data packets to and from an optical router. The optical router includes a number of ingress edge units coupled to an optical switch core coupled further to a number of egress edge units. The ingress edge units receive the optical data packets from the data links and aggregate the optical data packets into “super packets” where each super packet is to be routed to a particular destination egress edge unit. The super packets are sent from the ingress edge units to an optical switch fabric within the optical switch core that routes each super packet through the optical switch fabric to the super packet'"'"'s particular destination egress edge unit in a non-blocking manner (i.e., without contention or data loss through the optical switch fabric). This routing is managed by a core controller that monitors flow information at each ingress edge unit to control the super packet generation and transmission to the optical switch fabric and schedules each super packet to exit the optical switch fabric so as to avoid contention among the plurality of super packets in the transmission between the optical switch fabric and the egress edge units. The egress edge units receive the super packets, de-aggregate the super packets into the original optical data packets, and transmit the optical data packets to the data lines.
286 Citations
73 Claims
-
1. A router coupled to a plurality of data lines, comprising:
-
a plurality of egress edge units, wherein each egress edge unit is coupled to at least one egress port;
a plurality of ingress edge units, wherein each ingress edge unit receives a plurality of optical data packets, each optical data packet is destined for a destination port at one of the plurality of egress edge units, aggregates the plurality of optical data packets into a plurality of super packets wherein each super packet comprises optical data packets intended for a particular destination egress edge unit and is to be routed to that particular destination egress edge unit;
an optical switch fabric that receives the plurality of super packets from the plurality of ingress edge units and routes each super packet through the optical switch fabric to the particular destination egress edge unit for which the super packet is intended, and further wherein the routing through the optical switch fabric is performed in a non-blocking manner; and
a core controller that controls the arrival of the plurality of super packets at the optical switch fabric so as to avoid contention among the plurality of super packets between the optical switch fabric and the plurality of egress edge units; and
wherein the plurality of egress edge units receive the plurality of super packets, de-aggregate the plurality of super packets into the optical data packets, and transmit each of the plurality of optical data packets to at least one of the at least one egress port. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25)
a switch controller in communication with the optical switch fabric; and
a super packet scheduler in communication with the switch controller and further in communication with each of the plurality ingress edge units via a plurality of control packet links; and
wherein the super packet scheduler monitors the plurality of ingress edge units to determine a scheduling pattern for each of the plurality of ingress edge units, wherein the scheduling pattern causes each ingress edge unit to transmit super packets to the optical switch fabric so that no two super packets destined for a single egress edge unit arrive at the optical switch fabric in an identical switching time interval; and
wherein the switch controller creates a unique path through the optical switch fabric for each super packet arriving at the optical switch fabric during the identical switching time interval.
-
-
4. The router of claim 1, further comprising:
-
a plurality of ingress super packet links, wherein each super packet link connects one of the plurality of ingress edge units to the optical switch fabric;
a plurality of ingress super packet links, wherein each super packet link connects one of the plurality of edge units to the optical switch fabric; and
a plurality of egress super packet links, wherein each egress super packet link connects one of the plurality of egress edge units to the optical switch fabric; and
wherein the plurality of super packets aggregated at the plurality of ingress edge units are transmitted to the optical switch fabric over the plurality of ingress super packet links and further wherein the plurality of super packets are transmitted to the plurality of egress edge units from the optical switch fabric over the plurality of egress super packet links.
-
-
5. The router of claim 4, further comprising:
-
a plurality of ingress control packet links, wherein each ingress control packet link connects an ingress edge unit to the core controller; and
a plurality of egress control packet links, wherein each egress control packet link connects an egress edge unit to the core controller; and
wherein the core controller receives a plurality of pattern data from the plurality of ingress edge units that the core controller uses to establish a pattern that is used to route the plurality of super packets from the plurality of ingress edge units, through the optical switch fabric, to the plurality of egress edge units; and
wherein the core controller receives a plurality of control data from the plurality of egress edge units that the core controller uses to control overflow of a plurality of egress edge unit buffers and to synchronize data flow in the router.
-
-
6. The router of claim 5, wherein the core controller monitors a plurality of synchronization data at the plurality of ingress edge units via the ingress control data links and at the plurality of egress edge units via the egress control data link to synchronize data flow through the router;
- and
wherein the core controller monitors a plurality of time information at each of the plurality of egress edge units via the egress control data links to verify data intended for each egress edge unit arrives at an appropriate time.
- and
-
7. The router of claim 1, wherein the optical switch fabric comprises an optical cross-bar switch, comprising:
-
a plurality of inputs, each input connected to one of the plurality of ingress edge units;
a plurality of outputs, each output connected to one of the plurality of egress edge units; and
a plurality of switching elements configured to create a plurality of unique paths through the optical cross bar switch from each input to each output.
-
-
8. The router of claim 7, wherein the switching elements are N to 1 switching elements, where N is equal to the number of the plurality of ingress edge units, for switching a super packet received on any of the N inputs to one the plurality of outputs.
-
9. The router of claim 8, wherein each of the N to 1 switching elements comprises an N to 1 semiconductor optical amplifier operable to switch from N inputs to one output.
-
10. The router of claim 7, wherein each switching element has sufficiently broad bandwidth to accommodate all wavelengths within an optical fiber upon which the plurality of super packets are transported.
-
11. The router of claim 7, wherein the core controller, for each super packet arriving at the optical cross bar switch, connects an input of the optical switch fabric that is associated with the ingress edge unit from the super packet arrived to an output of the optical switch fabric that is associated with the egress edge unit for which the super packet is destined.
-
12. The router of claim 1, wherein each ingress edge unit further comprises:
-
a plurality of ingress super packet processors, wherein each ingress super packet processor receives a portion of the plurality of optical data packets arriving at the ingress edge unit and creates a plurality of partial super packets wherein each partial super packet is destined for a particular egress edge unit; and
an ingress super packet factory in communication between the super packet processor and the optical switch fabric, wherein the ingress super packet factory receives the plurality of partial super packets from each of the plurality of ingress super packet processors and creates a plurality of super packets by combining partial super packets having a common destination egress edge unit.
-
-
13. The router of claim 11, wherein the ingress super packet factory is in communication between the super packet processor and the optical switch fabric via one of a plurality of ingress super packet links, and further wherein the ingress super packet factory is in communication between the super packet processor and the core controller via one of a plurality of ingress control packet links.
-
14. The router of claim 11, wherein the super packet processor places data contained in each of the plurality of partial super packets on one or more wavelengths.
-
15. The router of claim 11, wherein each egress edge unit further comprises:
-
an egress super packet factory, wherein the egress super packet factory receives the plurality of the super packets destined for the egress edge unit and disassembles the plurality of super packets into a plurality of partial super packets intended for a common destination port; and
a plurality of egress super packet processors, wherein each egress super packet processor is coupled to a single destination port, and wherein each of the plurality of egress super packet processors the plurality of partial super packets intended for the destination port and transmits data contained within the plurality of partial super packets to the destination port.
-
-
16. The router of claim 1, wherein each ingress edge unit further comprises:
-
an ingress super packet processor, comprising;
a packet classification queue, comprising a plurality of sub-queues wherein each sub-queue is assigned to contain data of a particular data type that is intended for a particular egress edge interface unit;
a packet classification controller, wherein the packet classification controller receives the plurality of optical data packets, wherein each of the plurality of optical data packets comprises data of at least one data that is intended for one of the plurality of egress edge units, and routes each data within each optical data packet to the sub-queue that is assigned to contain the data, thereby building a partial super packets in each sub-queue wherein each partial super packet contains data of a particular data type intended for a particular egress edge unit; and
an edge unit destination controller, wherein the edge unit destination controller transmits each partial super packet from the packet classification queue to an ingress super packet factory; and
an ingress super packet factory, comprising;
a super packet ingress queue comprising a plurality of lambda/destination queues wherein each lambda/destination queue is assigned to contain data on a particular wavelength that is intended for a particular egress edge interface unit;
a partial super packet controller, wherein the partial super packet controller receives each partial super packet, wherein each partial super packet comprises data on at least one wavelength that is intended for one of the plurality of egress edge units, and routes each data within each partial super packet to the lambda/destination queue that is assigned to contain the data, thereby building a super packet in each lambda/destination queue, wherein each super packet contains data on a particular wavelength intended for a particular egress edge unit; and
a super packet transmit controller that forwards each super packet to the optical switch fabric based on input received from the core controller.
-
-
17. The router of claim 16, wherein the packet classification controller routes the data contained within the plurality of optical data packets by examining a header information for each of the plurality of optical data packets to determine for each data (i) the egress edge unit for which the data is intended and (ii) the data type of the data.
-
18. The router of claim 17, wherein the number of sub-queues is equal to the number of egress edge units multiplied by the number of data types, and further wherein the at least one data type comprises a TDM data type and at least one packet data type.
-
19. The router of claim 17, wherein the ingress super packet factory is in communication with the core controller via a control packet link to enable the core controller to monitor a plurality of input/output packet flow data at each ingress edge unit to determine super packet generation and control the transmission of super packets to the optical switch fabric.
-
20. The router of claim 16, wherein each egress edge unit further comprises:
-
a plurality of ports;
an egress super packet factory, comprising;
a super packet egress queue comprising a plurality of port/lambda queues wherein each port/lambda queue is assigned to contain data on a particular wavelength that is intended for a particular port;
a super packet egress queue controller, wherein the super packet egress queue controller receives each super packet from the optical switch fabric and routes each data within each super packet to the port/lambda queue that is assigned to contain the data, thereby deconstructing each super packet into a plurality of egress partial super packets where in the plurality of port/lambda queues, wherein each egress partial super packet contains data on a particular wavelength intended for a particular port; and
a super packet port selector forwards each egress partial super packet to an egress super packet processor;
an egress super packet processor, comprising;
a packet declassification queue comprising a plurality of egress sub-queues wherein each egress sub-queue is assigned to contain data on a particular wavelength that is intended for a particular port;
a packet declassification controller, wherein the packet declassification controller receives the plurality of egress partial super packets, wherein each of the plurality of optical data packets comprises data of at least one data type that is intended for one of the plurality of ports, and routes each data within each optical data packet to the egress sub-queue that is assigned to contain the data, thereby deconstructing the egress partial super packets so that each egress sub-queue contains data of a particular data type intended for a particular port; and
a packet selector that transmits the data in each egress sub-queue to an intended port.
-
-
21. The router of claim 1, wherein each ingress edge unit further:
-
collects a set of classification information for each of the plurality of optical data packets;
creates a classification index for each of the plurality of super packets, wherein each classification index contains the set of classification information for each of the optical data packets that comprise the super packet; and
places each classification index in an overhead of the super packet associated with the classification index; and
wherein the particular destination egress edge unit to which a particular super packet is destined receives the particular super packet, extracts from the classification index the classification information associated with each optical data packet within the particular super packet and routes each optical data packet to a port of the particular destination egress edge unit based on the classification information for each optical data packet.
-
-
22. The router of claim 21, wherein the classification information for each optical data packet comprises a destination egress edge unit and a destination port within the destination egress edge unit.
-
23. The router of claim 1, wherein the core controller further initializes the router based on either a predetermined scheduling pattern or a scheduling pattern based on an expected data flow.
-
24. The router of claim 5, wherein each of the plurality of super packet links and each of the plurality of control packet links are WDM fibers.
-
25. The router of claim 1, wherein each super packet is routed using slot deflection routing to route the super packet form an ingress edge unit to an egress edge unit.
-
26. A router for routing optical data, comprising:
-
an egress edge unit comprising a plurality of egress ports;
an ingress edge unit comprising a plurality of ingress ports, wherein the ingress edge unit receives an optical data packet intended for a destination port at the egress edge unit from one of the plurality of ingress ports, and further wherein the ingress edge unit aggregates the optical data packet into a super packet containing a plurality of optical data packets intended for the destination port;
an optical switch fabric that receives the super packet from the ingress edge unit and routes the super packet through the optical switch fabric to the egress edge unit in a non-blocking manner; and
a core controller that controls the arrival of the super packet at the optical switch fabric in such a manner that the super packet flows between the optical switch fabric and the egress edge unit without contention; and
wherein the egress edge unit receives the super packet, extracts the optical data packet from the super packet, and transmit the optical data packet to the destination port in the egress edge unit. - View Dependent Claims (27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47)
wherein the particular destination egress edge unit to which a particular super packet is destined receives the particular super packet, extracts from the classification index the classification information associated with each optical data packet within the particular super packet and routes each optical data packet to a port of the particular destination egress edge unit based on the classification information for each optical data packet.
-
-
28. The router of claim 27, wherein the classification information for each optical data packet comprises a destination egress edge unit and a destination port within the destination egress edge unit.
-
29. The router of claim 27, wherein the classification index is placed within a super packet header in a super packet containing the optical data packets associated with the classification index.
-
30. The router of claim 26, wherein the ingress edge unit further converts each optical data packet into an electrical data packet prior to aggregating the electrical data packet into a super packet.
-
31. The router of claim 26, wherein ingress edge unit further places the data within each super packet onto at least one wavelength.
-
32. The router of claim 26, wherein ingress edge unit further places the data within each super packet onto a plurality of wavelengths.
-
33. The router of claim 26, wherein the core controller comprises:
-
a switch controller in communication with the optical switch fabric; and
a super packet scheduler in communication with the switch controller and further in communication with the ingress edge unit via a control packet link; and
wherein the super packet scheduler monitors the ingress edge unit to determine a scheduling pattern for the ingress edge unit, wherein ingress edge unit transmits the super packet to the optical switch fabric according to the scheduling pattern so that the super packet arrives at the optical switch fabric at a switching time interval when no other super packet destined for the egress edge unit arrives at the optical switch fabric; and
wherein the switch controller creates a unique path through the optical switch fabric for the super packet.
-
-
34. The router of claim 26, further comprising:
-
an ingress super packet link that connects the ingress edge unit to the optical switch fabric;
an ingress super packet link that connects the edge unit to the optical switch fabric; and
an egress super packet link that connects the egress edge unit to the optical switch fabric; and
wherein the super packet is transmitted to the optical switch fabric over the ingress super packet link, and further wherein the super packet is transmitted to the egress edge unit from the optical switch fabric over the egress super packet link.
-
-
35. The router of claim 26, further comprising:
-
an ingress control packet link that connects the ingress edge unit to the core controller; and
an egress control packet link that connects the egress edge unit to the core controller; and
wherein the core controller receives a plurality of pattern data from the plurality of ingress edge units that the core controller uses to establish a pattern that is used to route the plurality of super packets from the plurality of ingress edge units, through the optical switch fabric, to the plurality of egress edge units; and
wherein the core controller receives a plurality of control data from the plurality of egress edge units that the core controller uses to control overflow of a plurality of egress edge unit buffers and to synchronize data flow in the router.
-
-
36. The router of claim 26, wherein the core controller monitors a plurality of synchronization data at the plurality of ingress edge units via the ingress control data links and at the plurality of egress edge units via the egress control data link to synchronize data flow through the router;
- and
wherein the core controller monitors a plurality of time information at each of the plurality of egress edge units via the egress control data links to verify data intended for each egress edge unit arrives at an appropriate time.
- and
-
37. The router of claim 26, wherein the optical switch fabric comprises an optical cross-bar switch, comprising:
-
a plurality of inputs, each input connected to one of the plurality of ingress edge units;
a plurality of outputs, each output connected to one of the plurality of egress edge units; and
a plurality of switching elements configured to create a plurality of unique paths through the optical cross bar switch from each input to each output.
-
-
38. The router of claim 37, wherein the switching elements are N to 1 switching elements, where N is equal to the number of the plurality of ingress edge units, for switching a super packet received on any of the N inputs to one the plurality of outputs.
-
39. The router of claim 38, wherein each of the N to 1 switching elements comprises an N to 1 semiconductor optical amplifier operable to switch from N inputs to one output.
-
40. The router of claim 38, wherein each ingress edge unit further places the data within each super packet onto a plurality of wavelengths, and further wherein each switching element has sufficiently broad bandwidth to accommodate all wavelengths within an optical fiber upon which the plurality of super packets are transported.
-
41. The router of claim 38, wherein the core controller, for each super packet arriving at the optical cross bar switch, connects an input associated with the ingress edge unit from the super packet arrived to an output for which the super packet is destined.
-
42. The router of claim 26, wherein each ingress edge unit further receives a portion of the plurality of optical data packets arriving at the ingress edge unit and creates a plurality of partial super packets wherein each partial super packet comprises data having a particular set of routing characteristics;
- and
creates a plurality of super packets by combining partial super packets having a common destination egress edge unit.
- and
-
43. The router of claim 42, wherein each egress edge unit further receives the plurality of the super packets destined for the egress edge unit and disassembles the plurality of super packets into a plurality of partial super packets intended for a common network port, and transmits data contained within the plurality of partial super packets to the common network port.
-
44. The router of claim 26, wherein each ingress edge unit further comprises:
-
a packet classification queue, comprising a plurality of sub-queues wherein each sub-queue is assigned to contain data of a particular data type that is intended for a particular egress edge interface unit;
a packet classification controller, wherein the packet classification controller receives the plurality of optical data packets, wherein each of the plurality of optical data packets comprises data of at least one data that is intended for one of the plurality of egress edge units, and routes each data within each optical data packet to the sub-queue that is assigned to contain the data, thereby building a partial super packets in each sub-queue wherein each partial super packet contains data of a particular data type intended for a particular egress edge unit;
an edge unit destination controller, wherein the edge unit destination controller transmits each partial super packet from the packet classification queue to an ingress super packet factory;
a super packet ingress queue comprising a plurality of lambda/destination queues wherein each lambda/destination queue is assigned to contain data on a particular wavelength that is intended for a particular egress edge interface unit;
a partial super packet controller, wherein the partial super packet controller receives each partial super packet, wherein each partial super packet comprises data on at least one wavelength that is intended for one of the plurality of egress edge units, and routes each data within each partial super packet to the lambda/destination queue that is assigned to contain the data, thereby building a super packet in each lambda/destination queue, wherein each super packet contains data on a particular wavelength intended for a particular egress edge unit; and
a super packet transmit controller that forwards each super packet to the optical switch fabric based on input received from the core controller.
-
-
45. The router of claim 44, wherein the packet classification controller routes the data contained within the plurality of optical data packets by examining a header information for each of the plurality of optical data packets to determine for each data (i) a destination egress edge unit for which the data is intended and (ii) a destination port contained within the destination egress edge unit for which the data is intended.
-
46. The router of claim 41, wherein the ingress super packet factory is in communication with the core controller via a control packet link to enable the core controller to monitor a plurality of input/output flow information and control super packet generation and transmission to the optical switch fabric.
-
47. The router of claim 46, wherein each egress edge unit further comprises:
-
a plurality of ports;
a super packet egress queue comprising a plurality of port/lambda queues wherein each port/lambda queue is assigned to contain data on a particular wavelength that is intended for a particular port;
a super packet egress queue controller, wherein the super packet egress queue controller receives each super packet from the optical switch fabric and routes each data within each super packet to the port/lambda queue that is assigned to contain the data, thereby deconstructing each super packet into a plurality of egress partial super packets where in the plurality of port/lambda queues, wherein each egress partial super packet contains data intended for a particular destination port;
a packet declassification queue comprising a plurality of egress sub-queues wherein each egress sub-queue is assigned to contain data on a particular wavelength that is intended for a destination port;
a packet declassification controller, wherein the packet declassification controller receives the plurality of egress partial super packets, wherein each of the plurality of optical data packets comprises data of at least one data type that is intended for one of the plurality of ports, and routes each data within each optical data packet to the egress sub-queue that is assigned to contain the data, thereby deconstructing the egress partial super packets so that each egress sub-queue contains data of a particular data type intended for a particular port; and
a packet selector that transmits the data in each egress sub-queue to an appropriate destination port.
-
-
48. A method of routing a plurality of incoming packets wherein each packet has a payload and a header, comprising:
-
classifying each incoming packet based on a destination port;
aggregating the classified packets into super packets based on the destination port;
transporting the super packets through an optical switch fabric in a non-blocking manner;
de-aggregating the super packets into de-aggregated packets; and
transporting the de-aggregated packets through the output ports. - View Dependent Claims (49, 50, 51, 52, 53, 54, 55, 56, 57)
determining a set of classification parameters for each incoming packet at an ingress edge unit, wherein the classification parameters comprise a packet destination egress edge unit;
transport the data packet to a destination egress edge unit; and
transporting the set of classification parameters for the data packet to the destination egress edge unit.
-
-
50. The method of claim 49, further comprising:
-
creating a classification index containing the classification parameters; and
transporting the classification index with the super packet to the destination egress edge unit.
-
-
51. The method of claim 50, wherein determining a set of classification parameters for a data packet at an ingress edge unit further comprises determining a destination egress edge unit and a destination port within the destination egress edge unit for the data packet the ingress edge unit.
-
52. The method of claim 49, further comprising:
-
accessing a look-up table to correlate destination IP address to destination egress unit and destination port; and
placing destination egress unit and destination port information within overhead of the super packet associated with the incoming packet.
-
-
53. The method of claim 49, wherein the set of classification parameters include a destination egress edge, a destination port, and a QoS parameter vector, and further wherein the QoS vector comprises at least one code point representing a set of QoS parameters for the data packet.
-
54. The method of claim 48, further comprising transporting each super packet over a plurality of bandwidths on a super packet link.
-
55. The method of claim 48, further comprising:
-
subdividing an available bandwidth on each of a plurality of super packet links; and
transporting a set of data contained in each super packet over at least one bandwidth.
-
-
56. The method of claim 48, further comprising:
-
allocating an available bandwidth to a destination egress edge unit based upon an analysis of an amount of data destined for the destination egress edge unit;
transporting each super packet intended for the destination egress edge unit over the available bandwidth.
-
-
57. The method of claim 48, further comprising deflection routing at least one super packet through a non-destination egress edge unit prior to routing the at least one super packet to the destination egress edge unit.
-
58. An optical router, comprising:
-
a non-blocking optical switch core, comprising;
a non-blocking optical switch fabric; and
a core controller;
at least one ingress edge unit;
at least one egress edge unit;
at least one ingress super packet link linking the at least one ingress edge unit to the optical switch fabric;
at least one egress super packet link linking the at least one egress edge unit to the optical switch fabric;
at least one ingress control packet link linking the at least one ingress edge unit to the core controller, wherein each of the at least one ingress control packet link provides a control information regarding the plurality super packets; and
at least one egress control packet link linking the at least one egress edge unit to the core controller, wherein each of the at least one egress control packet link provides an egress control information; and
wherein each ingress edge unit aggregates a plurality of incoming optical data packets into a plurality of super packets, wherein each super packet comprises optical data from the plurality of incoming optical data packets that is intended for one of the at least one egress edge units;
wherein the optical switch fabric receives the plurality of super packets and routes each super packet through the optical switch fabric to the at least one egress super packet link linking the one of the at least one egress edge unit for which the super packet is intended to the switch fabric, wherein the routing through the optical switch fabric is performed so as to avoid contention within the optical switch fabric between the plurality of super packets; and
wherein the core controller uses the ingress control information received from each of the at least one ingress edge units to schedule the plurality of super packets to exit the optical switch fabric in a manner that avoids contention among super packets over the at least one egress super packet link to the at least one egress edge unit. - View Dependent Claims (59, 60, 61, 62)
the at least one ingress super packet link comprises a plurality of ingress super packets links, wherein each ingress super packet link links one of the ingress edge units to the optical switch fabric;
the at least one egress super packet link comprises a plurality of egress super packet links, wherein each egress super packet link links one of the egress edge units to the optical switch fabric;
the at least one ingress control packet link comprises a plurality of ingress control packet links, wherein each ingress control packet link links one of the ingress edge units to the core controller; and
the at least one egress control packet link comprises a plurality of egress control packet links, wherein each egress control packet link links one of the egress edge unit to the core controller.
-
-
62. The optical router of claim 59, wherein the ingress super packet link is capable of transmitting multiple wavelengths in a time-multiplexed manner.
-
63. A method of routing a plurality of packets wherein each packet has a payload and a header, comprising:
-
receiving a plurality of optical data packets each of a plurality of ingress edge units, each optical data packet destined for a destination port at one of a plurality of egress edge units;
aggregating the plurality of optical data packets into a plurality of super packets wherein each super packet comprises optical data packets intended for a particular destination egress edge unit;
transmitting each super packet to its associated destination egress edge unit through an optical switch, wherein the transmitting is controlled so as to avoid contention within the optical switch fabric and among the plurality of super packets between the optical switch fabric and the plurality of egress edge units;
de-aggregate the plurality of super packets into the constituent optical data packets; and
transmit each optical data packet to an egress port. - View Dependent Claims (64, 65, 66, 67, 68, 69, 70, 71, 72, 73)
monitoring the plurality of optical data packets to determine a scheduling pattern;
transmitting super packets to the optical switch fabric according to the scheduling pattern;
wherein the scheduling pattern prevents any two super packets destined for a single egress edge unit from arriving at the optical switch fabric in an identical switching time interval.
-
-
65. The method of claim 64, further comprising creating a unique path through the optical switch fabric for each super packet arriving at the optical switch fabric during the identical switching time interval.
-
66. The method of claim 65, further comprising monitoring data flow at each of the plurality of egress edge units to control overflow of a plurality of egress edge unit buffers and to synchronize data flow in the router.
-
67. The method of claim 63, further comprising configuring the optical switch fabric to create a plurality of unique paths through the optical switch fabric from each ingress edge unit to each egress edge unit.
-
68. The method of claim 67, further comprising:
- connecting an input of the optical switch fabric that is connected to the ingress edge unit from which the super packet arrived to an output of the optical switch fabric t hat is connected to the egress edge unit for which the super packet is destined.
-
69. The method of claim 63, further comprising placing data contained in each of the plurality of super packets on one or more wavelengths.
-
70. The method of claim 63, further comprising:
-
creating a classification entry for each incoming optical data packet; and
creating a classification index for each super packet, wherein each classification index comprises the classification entry for every optical data packet within the super packet;
placing the classification index in an overhead of the super packet prior to transmitting the super packet to the destination egress edge unit;
extracting from the classification index the classification entry associated with each optical data packet within the super packet; and
routing each optical data packet to an output port of the destination egress edge unit based on the classification information for each optical data packet.
-
-
71. The method of claim 70, wherein the classification information for each optical data packet comprises a destination egress edge unit and a destination port within the destination egress edge unit.
-
72. The method of claim 63, further comprising periodically determining and applying a new scheduling pattern based on the monitoring of optical data packets.
-
73. The method of claim 63, further comprising transmitting at least one of the plurality of super packets to its associated destination egress edge unit through an intermediate edge unit.
Specification