Tropic Networks Inc.

Rockport Networks Inc.

Nortel Networks Corporation

Mitel Corporation

Dan R Oprea

Roland A Smith

Udo Mircea Neustadter

Genzao G Zhang

Andrei Catana

Roger Ferland

Wen Liu

Niranjan Vethanayagam

Constantin Viorel Marian

Ping Wai Wan

Derrick Remedios

Patrick Chilufya Chimfwembe

Paul David Obeda

Thomas George Zboril

Jacob Cepelinski

Brian Michael Wirth

Yang Sup Lee

The present invention provides a method and apparatus to route data packets across a torus or higher radix topology that has low latency, increased throughput and traffic distribution to avoid hot spots development. Disclosed is a method of routing packets in a distributed direct interconnect network from a source node to a destination node comprising the steps of: discovering all nodes and associated ports; updating the database to include the nodes and ports in the network topology; calculating the shortest path from every output port on each node to every other node in the topology; segmenting each packet into flits at the output port of the source node; as the flits are segmented, distributing said flits along the shortest path from each output port on the source node to the destination node using wormhole switching, whereby the packets are distributed along alternate maximum disjoint routes in the network topology; and re-assembling and re-ordering the packets at the destination node so that the packets accord with their original order/form.

Method to route packets in a distributed direct interconnect network

The present invention provides a method for managing the wiring and growth of a direct interconnect network implemented on a torus or higher radix interconnect structure based on an architecture that replaces the Network Interface Card (NIC) with PCIe switching cards housed in the server. Also provided is a passive patch panel for use in the implementation of the interconnect, comprising: a passive backplane that houses node to node connectivity for the interconnect; and at least one connector board plugged into the passive backplane comprising multiple connectors. The multiple connectors are capable of receiving an interconnecting plug to maintain the continuity of the torus or higher radix topology when not fully enabled. The PCIe card for use in the implementation of the interconnect comprises: at least 4 electrical or optical ports for the interconnect; a local switch; a processor with RAM and ROM memory; and a PCI interface.

Method and apparatus to manage the direct interconnect switch wiring and growth in computer networks

Methods and apparatus are provided for bandwidth management within Automatically Switched Optical Networks. An Optical Connection Controller (OCC) provides threshold based load balancing using an upper and a lower threshold for each port of a network element, and once port usage exceeds the upper threshold no more connections are allowed through the port until the port usage falls below the lower threshold. The OCC also minimizes fragmentation of a port so as to reduce the probability of call rejection as a result of fragmentation. The OCC also maintains a database of shortest paths between a network element and each destination, so that connections can be re-established quickly through alternate paths in the event of failure of a link or node. The OCC also provides the entire network with notification of a fault, thereby enhancing CR-LDP fault notification messaging.

Control plane architecture for automatically switched optical network

The invention describes methods and systems for monitoring the performance of an optical network by marking a group of optical signals with a set of identification tags which are unique to network characteristics. In the preferred embodiments, fiber identification (FID) and bundle identification (BID) tags are encoded into optical signals by marking an optical signal with low frequency dither tones whose frequencies are unique to the fiber section and to a bundle of fibers respectively. Detecting of the FID and BID tones provides more effective and accurate monitoring of performance of the optical network and allows determining of the network topology, e.g. paths of optical channels and traffic load through different fiber sections in the network. Other sets of hierarchically arranged identifiers encoded into optical signals have also been proposed, including band, conduit, city, region, country, etc. identifiers, as well as identifiers related to network security and service characteristics.

Method and system for monitoring performance of optical network

A switch module for a communication network switch comprises first and second switching units each for communicating with one or more network communication interfaces, and a switch interface connected to each of the first and second switching units and having a pair of communication ports for enabling data to be transferred on to a local communication path which is connected between the local communication ports.

Switch and a switching apparatus for a communication network

An optical add/drop multiplexing (OADM) apparatus and method are disclosed for detecting pilot tones, removing ghost tones, and re-inserting pilot tones on WDM signals in an optical network. The OADM apparatus comprises a smart processing and control unit (PCU) for detecting pilot tones used for channel identifications, including identifying and removing ghost tones associated with respective optical channels, and re-inserting the processed pilot tones into the respective optical channels. The OADM apparatus and method may incorporate digital PCU or analog PCU for processing pilot tones and removing ghost tones.

Method and apparatus for optical add/drop multiplexing in optical networks

The invention describes methods and systems for monitoring the performance of an optical network by introducing a fiber identification (FID) tag and/or bundle identification (BID) tag which are unique to the fiber section and to the bundle of fibers respectively. The FID tag is introduced by marking an optical signal, traveling through a section of fiber, with a low frequency dither tone whose frequency is unique to the fiber section. Similarly, the BID tag is introduced by marking an optical signal, traveling through a section of fiber in a bundle of fibers, with another low frequency dither tone whose frequency is unique to the bundle section. Detecting of the FID and BID tones either alone or along with an optionally introduced channel identification (CID) tone is provided.

A method and system for automatic allocation of port addresses in a network is provided. The node performs a self-discovery after initial power-up, and for each port on the node, using unique values associated with the port in the network hierarchy, applies a function that allows for an inverse function to the set of values to generate a default unique address for the port. A typical hierarchical structure in a network is the network area, node, and, within the node, the shelf, the card and the port. Next, the node sends a frame including the default port address and node identifier, from the port to a connected node. The node polls the port for a frame with network information associated with the connected port on the other node. If a frame is received from the other node, the node applies a logical condition to the default and received port addresses, and, if the condition is satisfied, applies another function to the received port address to generate a new unique address for the port, and maintains the new port address within the node. The method is described for IP networks but can be equally applied to ATM or Frame Relay networks, and for any nodes that support Sub-network Access Protocol (SNAP).

Method and system for automatic address allocation in a network and network protocol therefor

A packet routing/switching system and method are provided by which label switched routing is performed within a hierarchy of LSP/FA-LSPs defined/provisioned in a network of interconnected nodes. Rather than transmitting an entire label stack with an LSP label for each LSP/FA-LSP in the hierarchy, at a given node in the network, a single LSP label is transmitted together with a components label which contains a list of component identifiers which do not inherently identify LSP/FA-LSPs, but from which the full LSP/FA-LSP labels can be determined locally at each node using previously distributed information described in detail below. The list of component identifiers is a shorthand way identifying to adjacent nodes the identity of the LSP/FA-LSP hierarchy without transmitting the entire LSP labels and thus significantly reducing overhead. The single label which is transmitted is the label of the LSP/FA-LSP of which the given node forms a part.

Label switched routing system and method

An improved local synchronization module which uses Frequency-phase adaptive double locked loop (FPADLL) to control a stable controllable oscillator is disclosed. A single physical feedback loop is implemented which can operate in either a phase locked loop mode or a frequency locked loop mode. The sync module includes a controller which determines in which mode the feedback loop operates. The controller also uses slipping information from a network reference recovery interface to reduce slipping. Also effects of ageing of the stable controllable oscillator are predicted and compensated for.

Growing community of inventors

Kanata, Canada

Dan R Oprea