Abstract: A method and apparatus for switching optical signals by selectively routing such signals through a packet switch is disclosed. Depending upon predetermined traffic conditions in data being received and transmitted, the system may be provisioned to provide circuit like optical switching, or a mixture of optical and packet switching. An exemplary growable optical switch architecture is disclosed in the preferred embodiment.

Abstract: A data network routing apparatus and method are presented. The routing apparatus comprises a packet engine, which itself comprises a switch, a forwarding engine and a queueing processor. The queueing processor tracks individual input port to output port flows, and assigns packets to these flows. Flows are assigned to queues. Each queue can accommodate a large number of packets. Each queue is assigned to a subclass, and a number of subclasses are assigned to a class. The apparatus and method thus support numerous differentiable classes of data as well as further differentiable subclasses within each class. While queues within a given subclass are served with equal priority by the routing apparatus, each subclass can be assigned a different weight to differentiate the priority within a subclass. In turn, each class can be assigned a different weighting as well, to allow different treatment before reaching an output port. Thus, a wide spectrum of service differentiation is supported.

Abstract: A data network routing apparatus and method are presented. The data switch comprises a packet engine integrated with an optical engine. The optical engine and packet engine are under common software and/or hardware control, and communicably interconnected. As a result, the physical state of the optics is continually and immediately available to the packet engine, in substantially real time. The packet engine implements a routing algorithm which operates on at least the data traffic in the network and the optical transport topology. The routing apparatus periodically updates the routing algorithm's operands under normal operation, and in the event of the failure of a communications link in the network, the routing apparatus immediately updates the routing algorithm operands in response. A network comprising a plurality of nodes with such immediate rerouting capability is achieved by utilizing a plurality of the disclosed data switches as its nodes.

Abstract: The present invention is a portable base station (PBS) switching device for use in a wireless ATM local area network (LAN). The PBS includes a plurality of high speed interfaces which operate, for example, in the Gb/s range and which are adapted to couple to other PBS s in the network. The high speed interfaces are coupled to an ATM switching fabric adapted to selectively route incoming ATM cells to an appropriate output port in said PBS without any VPI (virtual path identifier) translation at transit PBSs. The PBS includes wireless interface and processor unit for coupling to mobile users which may include laptop and notebook computers. Memory is included within the processor unit for selectively storing ATM cells according to their VPI. In one preferred embodiment of the PBS a passive optical unit is used to transmit and receive data between PBSs. The passive optical unit may include a laser transmitter and optical receiver.

Abstract: A packet-based telecommunications architecture is disclosed that, like virtual-circuit networks, preserves the sequential order in which packets are presented to the network, but does not require node-by-node call set-up or tear-down, unlike virtual-circuit networks. Further, the packet switches which compose the architecture can be more simple than those used in datagram or virtual-circuit networks. An illustrative embodiment of the present invention comprises determining the topology of a network of packet switches that are connected by communication links, associating at least two names with at least one of the packet switches, and populating the router tables in the packet switches so that for each name the packet switches and communication links form a elemental network with the topology of a sink tree with the named packet switch at the root of the tree.

Abstract: The present invention concerns a method and apparatus for memory and channel sharing in a communication switching network which uses packet switching. In one form of the invention memory and channel sharing is provided within a packet switch. The packet switch includes a memory for storing a plurality data packets and a router that routes data packets from the memory to a particular output port. The router is operable to route data packets to either a dedicated output port, in other words, an output port defined by the virtual circuit path of the data packet, or a shared output port. In another form of the invention memory and channel sharing is provided by a packet switch in conjunction with a terminal. Protection lines which are provided in transmission systems can be used as the shared channels.

Abstract: A high-speed, high-capacity, asynchronous transfer mode packet switching system comprises electronic and optical components. The switching system may be a high-capacity multiple-gigabit-per-second optical system controlled by a relatively low-speed electronic controller operating at one hundred megabit-per-second rates. The switching system comprises a plurality of N input lines. Each of the N lines carries a succession of data packets or cells to the switching system. Each of the cells is input to the switch in time slots or cell periods of predetermined duration. A laser transmitter producing a separate and distinct carrier frequency identifying its associated input is responsive to each of the N input lines. The laser transmitters are connected to the input of an optical star coupler which is connected to a series of tunable receivers. There is one tunable receiver for each output of the switching system.

Abstract: A data stream is converted from a synchronous transfer mode (STM) to an asynchronous transfer mode (ATM) by extracting data of a number of payload signals from the data stream, writing data of the payload signal into a random access memory (RAM), separately recording address locations of the data of each of the payload signals in separate buffers, directing the RAM to separately read the data of the payload signals, and attaching a cell header to the data read from the RAM to form ATM cells.

Abstract: An m.times.n (m>n) output Packet Switch Unit is implemented by using an n.times.n Packet Switch Module and an m:n Concentrator. The arriving packet cells are supplied from the m Concentrator inputs to the n Concentrator outputs in a "first-in first-out" (FIFO) sequence. The Concentrator provides for buffering of arriving packet cells on the m Concentrator inputs in excess of available packet cell positions in the n Concentrator outputs until they can be supplied to a concentrator output in the FIFO sequence. In turn, packet cells from the n Concentrator outputs are supplied to n inputs of the Packet Switch Module which supplies them to appropriate output destinations associated with the n outputs of the Packet Switch Module. A plurality of the Concentrator-Based output Packet Switch Units is readily employed to implement any "larger" Packet Switch architecture.

Abstract: Significant throughput improvement is achieved for an input queued packet switch using output port schedulers by permitting the output schedulers to recycle or reassign cell transmission times from input ports which are unable to use them. When an output scheduler assigns a cell transmission time to an input port and that input port is unable to use the assigned transmission time due to a scheduling conflict, for example, the input port makes a new request for the same output port during the next subsequent request period and then returns the unusable transmission time assignment back to the output scheduler. The output scheduler stores the returned transmission time in a separate queue for assignment to later requests for the particular output port. Throughput performance is improved from 58% (without time slot recycling) to 92% (with time slot recycling) for random packet cell traffic models.

Abstract: A plurality of M receivers are utilized to receive a plurality of N channels from various incoming multiplexed data signals, N.gtoreq.M, where each incoming data signal arrives on a separate wavelength and comprises several channels. Each of the M receivers, in sequence, receives a separate one of the N channels, thereby M of the N channels. The process then repeats for the next M channels, in the same sequence as the first M channels. The repetitions continue until all N channels are received.

Abstract: A modular growable packet switching arrangement is constructed from a plurality of packet switches and a novel interconnect fabric. The interconnect fabric includes a plurality of cell routers, for receiving data packets and routing them to a plurality of packet switches, where each cell router includes at least one connection to each packet switch. The interconnect fabric also includes a sorting network for receiving concurrently arriving data packets, sorting the data packets based upon their addresses, and sending the sorted data packets to the inputs of the cell routers. The sorted data packets are arranged such that all data packets which must be transmitted to the same packet switch are sent to different cell routers. Then, all packets which are destined for a common packet switch may be routed there, each from a different cell router.

Abstract: A cross connect network is utilized in order to switch packets of information among a plurality of network nodes. Each of the plurality of network nodes generates a multiplexed data signal, where each multiplexed data signal includes several channels. The cross connect network routes the data packets based upon which channel each packet is in when it arrives from the network node, rather than based upon the address in each packet. Thus, the need to read the address and switch the packet at each switching stage in the cross connect network is eliminated.

Abstract: A communications network comprising a plurality of "subnetworks" multiplexed onto a single communications medium. A separate group of NIU's communicate on each predetermined "subnetwork", and each NIU also includes a tunable transmitter for transmitting data to NIU's of other subnetworks. Data from a user equipment is transmitted via the tunable transmitter if it is destined for an NIU from another subnetwork, and via a fixed transmitter if it is destined for an NIU on the same subnetwork. In one embodiment, fiber optics is utilized in order to provide a high speed network with tunable lasers.

Abstract: An interconnect fabric is constructed from a plurality of fixed wavelength transmitters which are used to transmit arriving data packets through a star coupler, and a plurality of tunable receivers which tune to whatever frequency necessary to receive the desired data from the star coupler. A control network, constructed from a plurality of fixed wavelength receivers and a plurality of tunable transmitters, determines what frequencies the tunable receivers should tune to, and sends a signal to effectuate such tuning.

Abstract: A congestion control method and apparatus for use with a communications link comprising a plurality of N channels. A plurality of at most N-1 queues are sequentially polled, and data is output therefrom to the communications link, thereby leaving at least one remaining channel. After the at most N-1 queues are polled, a determination is made as to which of the at most N-1 queues is closest to a data overflow condition. The remaining at least one channel is then utilized to transmit data from the queues which are closest to overflow.

Abstract: This invention is large N.times.N packet switch, formed using a plurality of smaller packet switches. The invention comprises an N input, L output interconnect fabric (L>N), and a plurality of J.times.K smaller packet switches (J>K). Each of the J inputs to each packet switch is connected to a separate one of the L outputs of the interconnect fabric, and each of the K outputs from each packet switch is connected to a destination equipment. In operation, packets are received at the N inputs to the interconnect fabric, and each packet is routed to one of the inputs of the packet switch associated with the destination user equipment for the packet. Simultaneous packets, up to J in number, are routed to separate inputs of a particular packet switch for distribution to their respective destinations, while all other simultaneous packets destined for user equipments associated with the same packet switch are lost, the probability of such a loss being acceptably small.

Abstract: This invention is an inventive N input by L output interconnect fabric. In operation, packets comprising an information field and an address are received at the N inputs to the interconnect fabric, and the address in each packet is mapped to a group of outputs, rather than to any particular output. Each packet is then routed to any available one of the interconnect fabric outputs associated with the group to which the packet is mapped. If a number of packets destined for the same group simultaneously arrive at the interconnect fabric inputs and the group to which they are all destined does not comprise enough outputs to accept them, then all packets in excess of the number that the destined group can accept are simply discarded. The probability of lost jackets due to such discarded packets is acceptably small. In one exemplary embodiment, the invention can be utilized to build arbitrarily large packet switches.

Abstract: This invention is an inventive N input by L output interconnect fabric. In operation, packets comprising an information field and an address are received at the N inputs to the interconnect fabric, and the address in each packet is mapped to a group of outputs, rather than to any particular output. Each packet is then routed to any available one of the interconnect fabric outputs associated with the group to which the packet is mapped. If a number of packets destined for the same group simultaneously arrive at the interconnect fabric inputs and the group to which they are all destined does not comprise enough outputs to accept them, then all packets in excess of the number that the destined group can accept are simply discarded. The probability of lost .[.jackets.]. .Iadd.packets .Iaddend.due to such discarded packets is acceptably small. In one exemplary embodiment, the invention can be utilized to build arbitrarily large packet switches.

Abstract: This invention is large N.times.N packet switch, formed using a plurality of smaller packet switches. The invention comprises an N input, L output interconnect fabric (L>N), and a plurality of J.times.K smaller packet switches (J>K). Each of the J inputs to each packet switch is connected to a separate one of the L outputs of the interconnect fabric, and each of the K outputs from each packet switch is connected to a destination equipment. In operation, packets are received at the N inputs to the interconnect fabric, and each packet is routed to one of the inputs of the packet switch associated with the destination user equipment for the packet. Simultaneous packets, up to J in number, are routed to separate inputs of a particular packet switch for distribution to their respective destinations, while all other simultaneous packets destined for user equipments associated with the same packet switch are lost, the probability of such a loss being acceptably small.