Abstract: In a wide-coverage network comprising electronic edge nodes interconnected by bufferless core nodes, where each edge node comprises a source node and a sink node, both sharing an edge-node controller and having means for data storage and managing data buffers, the transfer of bursts from source nodes to sink nodes via the core nodes requires precise time coordination to prevent contention at the bufferless core nodes. A core node preferably comprises a plurality of optical switches each of which may switch entire channels or individual bursts. Each source node has a time counter and each core node has at least one time counter. All time counters have the same period and time-coordination can be realized through an exchange of time-counter readings between each source node and its adjacent core nodes. The time-counter readings are carried in-band, alongside payload data bursts destined to sink nodes, and each must be timed to arrive at a corresponding core node during a designated time interval.

Abstract: An optical channel switch comprising independent optical space switches interconnects electronic data-switch modules to form a high-capacity wide-coverage network. The optical channel switch connects incoming wavelength-division-multiplexed (WDM) links each originating from a data-switch module to outgoing WDM links each terminating in a data-switch module. The optical space switches are of equal dimension, each having the same number of dual ports (input-output ports). The channels of incoming WDM links are assigned to the input ports of the space switches and the output ports of the space switches are assigned to channels of the outgoing WDM links in a manner which permits a switched path from any data-switch module to any other data-switch module even when the number of data-switch modules substantially exceeds the number of dual ports per space switch. A method of reconfiguring the channel switch in response to varying traffic demand is also disclosed.

Abstract: A scaleable high-capacity network comprises several non-uniform composite-star networks interconnected by a number of lateral uniform composite-star networks, thus forming an irregular two-dimensional network. Each non-uniform composite star network comprises electronic edge nodes, possibly of significantly different capacities, interconnected by optical core nodes. The optical core nodes are not connected to each other, and each may be configured differently and have a different reach index, where the reach index of a core node is the number of edge nodes to which the core node directly connects. The selection of a route through a core node within a non-uniform composite-star network is based on a composite index determined according to the reach index of the core node and the propagation delay along the route.

Abstract: A method and apparatus are provided for low latency loss-free burst switching. Burst schedules are initiated by controllers of bufferless core nodes and distributed to respective edge nodes. In a composite-star network, the burst schedules are initiated by any of a plurality of bufferless core nodes and distributed to respective edge nodes. Burst formation takes place at source nodes and a burst size is determined according to an allocated bitrate of a burst stream to which the burst belongs. An allocated bitrate of a burst stream may be modified according to observed usage of scheduled bursts of a burst stream. A method of control-burst exchange between each of a plurality of edge nodes and each of a plurality of bufferless core nodes enables burst scheduling, time coordination, and loss-free burst switching. Both the payload bursts and control bursts are carried by optical channels connecting the edge nodes and the core nodes.

Abstract: A fast optical switch is needed to realize an economical and scaleable optical-core network. In the disclosed optical switch, switching is effected by rapid wavelength conversion. Either channel switching, Time Division Multiplex (TDM) switching or both may be provided by the fast optical switch. The operation of the fast optical switch is enabled by a fast scheduler. The throughput of the optical switch may be increased through a process of bimodal pipelined connection-packing. An in-band exchange of control signals with external nodes may serve to minimize the control overhead. Such control signals may include time-locking signals and connection-requests. A modular structure may be configured to comprise several fast optical switches to yield a high-speed, high-capacity, fully-connected optical switch.

Abstract: The present invention enables fast reconfiguration of paths associated with bufferless core nodes in a network where any two edge nodes may interconnect through a core node. In order to enhance network agility and performance, edge nodes associated with a core node are separated into a plurality of groups based on, for example, round-trip delay from the core node. This separates the short-haul paths from the long haul paths, thereby enabling the short-haul paths to be configured more frequently than the long-haul paths. It is preferable that the periods be set such that each additional period is an integer multiple of the previous one. This allows for simultaneous configuration of groups of paths. Although the methods and apparatus of the present invention are most effective for bufferless optical switches, they are still effective for electronic switches which may be bufferless or equipped with input data buffers.

Abstract: A universal electronic switching node serves as an edge node in a high-capacity network with an optical core. The universal edge node may handle a variety of traffic classes and may control traffic admission, connection definition, connection routing and core node configuration. The provided capabilities significantly simplify network operation and control. The universal edge node includes input ports for receiving data streams, output ports for transmitting the data streams though the optical core, a switching fabric for communicating these data streams between input and output ports and a controller for controlling this communicating. In particular, the controller can select a route through the optical core, schedule the communication between input and output ports and adaptively allocate the bitrate of this communication.

Abstract: A method and a network for a universal transfer mode (UTM) of transferring data packets at a regulated rate are disclosed. The method defines a protocol that uses an adaptive packet header simplify packet routing and increase transfer speed. The protocol supports a plurality of data formats, such as PCM voice data, IP packets, ATM cells, frame relay and the like. The network preferably includes plurality of modules that provide interfaces to various data sources. The modules are interconnected by an optic core with adequate inter-module links with preferably no more than two hops being required between any origination/destination pair of modules. The adaptive packet header is used for both signaling and payload transfer.

Abstract: Methods and apparatus for transferring data segments of a data stream across multi-channel links in a high-capacity network, so that the segments are equitably smeared across the channels of each multi-channel link, are described. The high-capacity network comprises a multiplicity of distributed high-capacity edge nodes interconnected by multi-channel links to a plurality of core nodes. Each edge node comprises a source node and a sink node which may share memory and control. A path from a source node of a first edge node to a sink node of a second edge node traverses two links, a first link from the source node to a selected core node and a second link from the selected core node to the destination sink node. Tandem switching through an intermediate edge node is not required, even for data streams of very low bit rate. Two types of core nodes are described. A core node of a first type is constructed as a high-capacity single-plane switch.

Abstract: The expanding IP data networks require address resolution mechanisms that are fast and able to handle a vast number of addresses. Novel fast address resolution mechanisms with a large address directory are described. The address resolution mechanisms make use of a compound indexing-searching technique to increase the speed and capacity of a translation apparatus. Further embodiments employ partial address scrambling to balance the loads of parallel search units.

Abstract: A simple access device provides access to a high-capacity network to a traffic source. Once connected to a traffic source, an edge module in the high-capacity network can assign a sub-network address to the traffic source so that the source may also serve as a sink for data traffic. Although the sub-network address assigned to the traffic source may bear little resemblance to a network address for the edge module, distant edge modules may route traffic to a traffic sink having such a sub-network address easily and flexibly. Features may be added to the access device, which would otherwise be little more than a multiplexer, but, for the most part, the edge modules make the routing decisions. Data transfer from the current Internet to the high-capacity network and vice versa can be facilitated by requiring that each edge module in the high-capacity network that acts as a gateway be bilingual, that is, understand both a simple high-capacity network-specific routing protocol and conventional Internet protocols.

Abstract: In a communication network comprising nodes and links between the nodes, a controller node disseminates link state information. A nodal routing table exists at each node comprising routes between pairs of nodes. The nodal routing table is either populated by the given node based on network information received from the controlling node or populated at the controlling node and received by the given node. Each node receives heartbeat signals from its neighbouring nodes. An unexpected delay between heartbeat signals may be perceived as a failure of a link. The perceived failure of that link is reported by the perceiving node to the controlling node. Upon receiving link failure information from a node, the controlling node may determine a subset of nodes in the network influenced by the link failure and indicate the link failure to the determined subset of influenced nodes.

Abstract: A fast optical switch is needed to realize an economical and scaleable optical-core network. In the disclosed optical switch, switching is effected by rapid wavelength conversion. Either channel switching, Time Division Multiplex (TDM) switching or both may be provided by the fast optical switch. The operation of the fast optical switch is enabled by a fast scheduler. The throughput of the optical switch may be increased through a process of bimodal pipelined connection-packing. An in-band exchange of control signals with external nodes may serve to minimize the control overhead. Such control signals may include time-locking signals and connection-requests. A modular structure may be configured to comprise several fast optical switches to yield a high-speed, high-capacity, fully-connected optical switch.

Abstract: A flexible global distributed switch adapted for wide geographical coverage with an end-to-end capacity that scales to several Petabits per second (Pb/s), while providing grade-of-service and quality-of-service control, is constructed from packet-switching edge modules and channel-switching core modules. The global distributed switch may be used to form a global Internet. The global distributed switch enables simple controls, resulting in scalability and performance advantages due to a significant reduction in the mean number of hops in a path between two edge modules. Traffic is sorted at each ingress edge module according to egress edge module. At least one packet queue is dedicated to each egress edge module. Harmonious reconfiguration of edge modules and core modules is realized by time counter co-ordination. The global distributed switch can be enlarged from an initial capacity of a few Terabits per second to a capacity of several Petabits per second, and from regional to global coverage.

Abstract: At a master controller of a space switch in a node in a data network, a request is received from a source node that requests a connection to be established through the space switch. This request is compared to other such requests so that a schedule may be established for access to the space switch. The schedule is then sent to the source nodes as well as to a slave controller of the space switch. The source nodes send data bursts which are received at the space switch during a short guard time between successive reconfigurations of the space switch. Data bursts are received at the space switch at a precisely determined instant of time that ensures that the space switch has already reconfigured to provide requested paths for the individual bursts. The scheduling is pipelined and performed in a manner that attempts to reduce mismatch intervals of the occupancy states of input and output ports of the space switch.

Abstract: A method is described for segmenting a data stream comprising variable-size packets, a data stream being defined by its source node, sink node, assigned network route, and other attributes. The segments are of equal size and the method concatenates the packets in successive segments in a manner that attempts to minimize segmentation waste without undue delay. The method facilitates the construction of efficient networks that scale to very high capacities while respecting service-quality specifications. Apparatus for implementing the method are also described.

Abstract: A multi-grained network includes edge modules that switch high-variance multi-rate data traffic, and independent core modules that switch paths having different granularities. The core may include core modules that switch fixed-size data blocks, core modules that switch channels or bands of channels, core modules that switch entire links, and core modules that cross-connect channels or links. To simplify the control functions, the core modules operate independently from each other. Direct link, band or channel connections may be established for selected ingress-egress edge module pairs, if traffic volumes warrant. The use of graded granularity in the core simplifies the control function and reduces network cost.

Abstract: A self-configuring distributed packet switch which operates in wavelength division multiplexed (WDM) and time division multiplexed (TDM) modes is described. The switch comprises a distributed channel switching core, the core modules being respectively connected by a plurality of channels to a plurality of high-capacity packet switch edge modules. Each core module operates independently to schedule paths between edge modules, and reconfigures the paths in response to dynamic changes in data traffic loads reported by the edge modules. Reconfiguration timing between the packet switch modules and the channel switch core modules is performed to keep reconfiguration guard time minimized. The advantage is a high-capacity, load-adaptive, self-configuring switch that can be distributed to serve a large geographical area and can be scaled to hundreds of Tera bits per second to support applications that require, very high bandwidth and a guaranteed quality of service.

Abstract: A packet switch includes ingress modules, egress modules, and a switch core. Packets of variable sizes may be divided into segments of predetermined sizes to facilitate switching within the switch core. The segmentation process may require null-padding which increases the flow rate within the switch. The segments are transferred from ingress to egress under flow-rate control based on inflating specified flow rates. Null padding is removed at egress. Ingress modules sort segments or packets according to their destination egress modules and egress modules sort segments or packets according to their ingress-module origin. Thus, segments of the same packet, and packets of the same stream, are properly concatenated as they appear in a logical buffer at egress. To enable growth of the packet switch to accommodate a very large number of ingress and egress modules, the switch implements a method of fast matching to schedule packet transfer across the switch core.

Abstract: An N-dimensional lattice network that scales to capacities of the order of a Yotta bits per second (1024 bits per second) includes a plurality of sub-nets of edge module switches interconnected by an agile switching core. The agile core may be distributed. In the N-dimensional lattice network, each edge module 408 is connected to N core stages, each core stage having an associated dimensional indicator. The input/output ports of each edge module are divided into (N+1) port groups. One of the port groups serves local sources/sinks while the remainder of the port groups are respectively connected to core stages in each of the N dimensions. This structure permits virtually unlimited capacity growth and significantly simplifies the routing and forwarding functions. The edge modules are addressed using logical coordinates, one coordinate being assigned for each of the N dimensions. This simplifies routing and permits each edge module to compute its own routing tables.