Non-blocking, scalable optical router architecture and method for routing optical traffic

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.

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.

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.

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.