Abstract: A scheme for guaranteeing network tunnel path (NTP). e.g., Internet Protocol (IP), service levels in optical (e.g., wavelength-division multiplex (WDM)) networks. The invention, in one embodiment, accounts for both network topology (e.g. switches, cross-connects, and links between nodes) and resource usage (e.g., available/provisioned link bandwidth and available/used wavelength paths), which information may be available from the IP and optical WDM protocol layers, in particular (i) router capacities and (ii) presence or absence of wavelength conversion capability of optical cross-connect at each node. A determination is made whether to route an arriving request for an NTP over existing topology by computing “good” routes in accordance with a defined metric, or to open a new, available optical wavelength path by provisioning nodes of the topology for, and computing “good” routes for new wavelength paths. As many requests as possible are identified without a priori information of future requests.

Abstract: A source node, along with transmitting an optical path setup request, notifies towards a destination node available resource information related to itself, pre-assigns the available resources which its own node has notified. Each of transit nodes, along with relaying towards the destination node the optical path setup request which has been received from a previous hop node, notifies towards the destination node available resource information related to itself, and pre-assigns the available resources which each of their own nodes has notified. The destination node, along with reserving a resource which is to be used for setting up an optical path based upon an optical path setup request which has been received, transmits a resource reservation request towards the source node. The transit nodes and the source node actually reserve an available resource which has been pre-assigned, based upon a resource reservation request from a next hop node.

Abstract: A method and apparatus for an optical reroutable redundancy scheme. According to one embodiment of the invention, a wavelength division multiplexing optical network includes a number of nodes each having an optical cross connect and each having stored therein a database representing conversion free connectivity from that node to others of the nodes. In addition, each of the nodes employs a source based scheme with the database to provision optical circuits in real time and an optical reroutable redundancy scheme.

Abstract: An optical switch router performs both optical switching and traditional routing. The optical switch router includes optical interfaces for coupling to one or more incoming optical fibers and to one or more outgoing optical fibers, and also includes a number of traditional router ports. Individual incoming optical data streams received over the incoming optical fiber(s) can be selectively passed through to one or more of the outgoing optical fibers or “dropped” from the optical communication path for traditional routing. Routed traffic, which can be received over the router ports or from the “dropped” optical data streams, can be forwarded over optical data streams that are “added” to the outgoing optical fiber(s). The “added” optical data streams may be added to the outgoing optical fiber(s) at any unused wavelengths, including, but not limited to, the wavelengths of the “dropped” optical data streams.

Abstract: In this application, the invention is a novel heuristic mechanism of working path and backup path reservation for highly utilizable WDM mesh networks, which we call dynamic-network adapted cost selection (DAC-selection) mechanism. One of its characteristics is simple traffic distribution mechanism, and the performance of DAC-selection mechanism is superior to that of Random selection (R-selection) and Advanced Combined Min-cost selection (ACM-selection) mechanism. DAC-selection mechanism provides simple cost function by assigning proper weights to each component of the cost function and the total cost is obtained by just summing up the individual cost. Therefore, it is possible to select a best pair of working path and backup path which consumes least wavelength for reservation.

Abstract: A main optical signal is wavelength-multiplexed with an optical identifier which is different in wavelength from this main optical signal, before a wavelength-demultiplexing is take place at a desired point so as to isolate the optical identifier from the main optical signal for the purpose of monitoring establishment of correct or incorrect transmission route between transmission and receiving sides based on the detected optical identifier with reference to the stored corresponding data to the relationship between the main optical signal and the optical identifier which belongs to the main optical signal, wherein the monitoring is made without any photoelectric conversion of the main optical signal and any deterioration in quality of the main optical signal.

Abstract: The present invention relates to an optical communication network and to a network element for use in such a network. The network element comprises a plurality of receivers (70-76) for receiving optical communication signals, a plurality of transmitters (54-62, 82-86) for transmitting optical communication signals, and a plurality of network connections, each network connection having an individual signal impairment characteristic. The pluralities of receivers (70-76) and transmitters (54-62, 82-86) are adapted to employ a plurality of different modulation schemes (64, 66). Furthermore, the pluralities of receivers (70-76) and transmitters (54-62, 82-86) are assigned to the network connections as a function of the individual signal impairment characteristics.

Abstract: Embodiments of the present invention provide a system and method for routing optical data. In one embodiment of the present invention, data can routed in a non-blocking fashion from a plurality of the ingress edge units to a plurality of egress edge units by selectively connecting the ingress edge units to the egress edge units at a photonic core. Switching at the photonic core can be driven by a clock signal and does not require an electrical to optical conversion to occur at the core. This can allow switching on the nanosecond level, substantially faster than many prior art systems.

Abstract: A path routing computation method enables reduction of the memory capacity for path routing computation. The method is characterized in that a wavelength convertible subnetwork in which paths are connected in a mesh form; a first and second wavelength inconvertible subnetworks have a starting point node and an end point node, respectively, and include a plurality of nodes and connected via the wavelength convertible subnetwork, and out of the nodes constituting the first and second wavelength inconvertible subnetworks, a node has a port connected to the wavelength convertible subnetwork is defined as a border node, and the method includes the steps of: obtaining, for the first wavelength inconvertible subnetwork, a path from the starting point node to a border node in the first subnetwork; and obtaining, for the second wavelength inconvertible subnetwork, a path from the end point node to a border node in the second wavelength inconvertible subnetwork.

Abstract: A modular reconfigurable multi-server system with hybrid optical and electrical switching fabrics for high-speed networking within a wavelength-division-multiplexed based photonic burst-switched (PBS) network with variable time slot provisioning. An optical high-speed I/O module within the multi-server system includes an optical switch with the control interface unit. A server module of the multi-server system statistically multiplexes IP packets and/or Ethernet frames to be transmitted over the PBS network, generate control and data bursts and schedule their transmission. Then, the server E-O converts the bursts, and then transmits the optical bursts to the optical I/O module. The optical I/O module then optically transmits the bursts to the next hop in the optical path after processing the optical control burst to configure the optical switch, which then optically switches the optical data burst without performing an O-E-O conversion.

Abstract: A bit-rate-transparent electrical space-division switching matrix is employed in an optical cross-connect and the input/output stage is constructed from simple, broadband optical receivers and transmitters. Since the switching matrix operates in unclocked manner, i.e. its switching function is not based on internal bit timing and frame timing, arbitrary signals can be switched though transparently at almost any bit rate, independently of the protocol-type being used. The inputs and outputs likewise operate fully independently of bit rate and protocol, since they only implement an O/E conversion or O/E conversion. By virtue of this structure, a simply constructed but extremely powerful optical cross-connect is created that can be employed equally for all types of optical signals within the stipulated wavelength-range.

Abstract: The present invention relates to a node used in an optical communication network, and comprises functions for transferring and receiving data, and a unit for establishing and releasing a cut through path to a node of the next stage, the establishing and releasing unit having a detecting unit which detects the arrival of a request packet for establishing a cut through path to the node stage.

Abstract: A method and system for providing tandem protection in a communication system. Path protection is provided using at least two redundant communication paths and selecting the communication having higher signal quality. Interface protection is provided through a protection transceiver thus implementing M:N equipment protection and 1+1 optical protection.

Abstract: A network is protected against interruption of service while one or more faulty switches or optical fiber transmission lines are repaired or replaced, by an interconnecting configuration of small N×N optical input/output switches, where N is 2 or greater than 2. The switches are configured among protection and working transmission lines. The small number of fibers for each switch improves repair and installation connection reliability and permits configurations that flexibly meet differing requirements. Also the fault is monitored with a fault check signal.

Abstract: A wavelength allocation method and apparatus that is capable minimizing the number of wavelength converters is provided. In a dynamic wavelength multiplexing division network structure having a limited maximum transmission distance, an optical path is established to maximally suppress activation of a 3R wavelength converter, thereby maximizing efficiency of the 3R wavelength converter. In a dynamic wavelength multiplexing division network structure having a limited maximum transmission distance, a new optical path is established to minimize 3R wavelength conversion and suppress use of unnecessary expensive 3R wavelength converters and thereby increase network efficiency.

Abstract: A method and apparatus are disclosed for temporally shifting one or more packets using wavelength selective delays. The header information associated with each packet, together with a routing algorithm, routing topology information and internal OPTR state, is used to route each packet to the appropriate destination channel and to make timing decisions. A wavelength server generates optical control wavelengths in response to the timing decisions. A generated optical control wavelength is used to adjust the wavelength of a given packet tray and thereby introduce a wavelength selective delay to the packet tray to align packet trays or to shift one or more packet trays to avoid a collision. The wavelength of the packet tray is converted to a control wavelength corresponding to an identified delay, irrespective of the initial channel upon which the packet tray was received. At the output stage of the packet tray router, the packet tray wavelength can be converted to any desired output channel wavelength.

Abstract: An exemplary embodiment of the invention is a network element in an optical communications network. The network element includes a data collection application program interface (API) for receiving a request from a client for network topology information along with a topology gatherer module, a topology change module and a gateway node module, each in communication with the data collection API. One of the topology gatherer module, topology change module and gateway node module provides the network topology information requested by the client to the data collection API. The data collection API then provides the network topology information to the client. Alternate embodiments include a method of collecting optical communications network.

Abstract: A method capable of optical packet switching even in a case where physical optical data in an optical packet switch has not the shape of a packet but the shape of continuous data, and a system therefore are provided. The optical switching method includes the steps of (a) detecting input terminals into which an optical packet is input, of an optical switch and switching the optical packet to destination output terminals of the optical packet, and (b) when there are the input terminals into which the optical packet is not inputted in step (a), detecting other input terminals into which the optical packet, which direct an output terminal among the plurality of output terminals connected to the input terminals, is inputted, and searching a blank output terminal and switching dummy data from the input terminal to the blank output terminal.

Abstract: An optical router is disclosed having at least one frequency router. The frequency router includes a plurality of input ports and a plurality of output ports. At least one input port simultaneously receives at least two optical signals to be frequency routed, while at least one output port simultaneously presents at least two frequency routed optical signals. Each optical signal to be frequency routed is colored in response to destination information.

Abstract: An optical fiber system is provided which allows customers to access the system at points between huts (facilities where regional optical fiber cables are available for connection). A long-haul optical fiber system may be disposed between a first long-haul node and a second long-haul node. A regional optical fiber may then be disposed between the first long-haul node and the second long-haul node, and may have an access point at a hut disposed between the first long-haul node and the second long-haul node. An access optical fiber may be disposed between the first node and the second node and may have an access point at the hut, as well as at one or more points between the hut and the first long-haul node. The access optical fiber and the regional optical fiber may be coupled to a switch at the hut.

Abstract: In an optical packet switch, NW wavelengths, over which inputted optical packets may be switched, are grouped into KG groups of wavelengths, where NW and KG are integers greater than one. The KG groups of wavelengths are characterized in that each of the KG groups of wavelengths is allocated to optical packets distinguished from other optical packets by at least one attribute of at least one packet characteristic. Each one inputted optical packet is switched over a wavelength having an available transmission resource selected from among wavelengths in one of the KG groups of wavelengths that is matched to the one inputted optical packet by correspondence of attributes of the at least one packet characteristic. Related apparatus and methods are also described.

Abstract: A method and apparatus provide a network element and associated control system in a broadcast optical communications network, including an optical switch (switch fabric), to cross-connect selected wavelengths input to the switch to pre-determined output ports, under the control of a control (supervisory) channel for efficient upstream and downstream delivery of optical communications services. The network element includes an optical switch controller to monitor the communications traffic input to and output from the optical switch ports in order to dynamically change output port connections for the input wavelengths as controlled by the control channel.

Abstract: There is provided a wavelength division multiplex transmission system with satisfactory transmission characteristics and extensive functions for avoidance of defects. Transmission signals to be transmitted by an optical transmission device are distributed among a plurality of wavelength components, converted into WDM signals, and sent to a WDM transmission network, and WDM signals from the WDM transmission network are restored to the transmission signals by an optical receiving device. This system includes a wavelength component-specific route setting device which sets routes for transmission on the WDM transmission network for each wavelength component.

Abstract: A host station apparatus (10) generates band allocation information including identifications of slave station apparatuses (20-1 through 200-n) and types of data to be transmitted by the slave station apparatuses and posts the information to the plural slave station apparatuses (20-1 through 20-n). The plural slave station apparatuses (20-1 through 20-n) identify as to whether or not the band allocation information is band allocation information about the data types of the slave station apparatuses respectively, and when the band allocation information is band allocation information about the data types of the slave station apparatuses, they control to transmit data to the host station apparatus (10) according to the data types represented by the band allocation information.

Abstract: A wavelength-selective routing capability is provided in a single network element by configuring the network element using a combination of add/drop network elements to route individual optical channels of WDM signals among a plurality of optical transmission paths coupled to the network element. Any optical channel of any WDM signal received at the network element can be selectively added, dropped, or routed among the multiple optical transmission paths within and external to the node.

Abstract: The present invention provides an optical system for decoding, switching, demultiplexing, and routing of optical encoded data symbols, including: a plurality of optical paths having first and second terminals; a splitting mechanism for directing the encoded data symbols to each of the first terminals; a plurality of decoding devices for producing decoded signals in response to the encoded data symbols; and each of the optical paths includes, between the first and second terminals, at least one of the decoding devices to produce one of the decoded signals at one of the second terminals in response to one of the encoded data symbols.

Abstract: A system, device, and method for supporting cut-through paths in an optical communication system involves obtaining hop count and quality of service information by an initiating device and using the hop count and quality of service information by the initiating device to make decisions relating to a cut-through path. The hop count and quality of service information is provided in a reply message that is sent by a terminating device and modified by intermediate devices between the initiating device and the terminating device. The terminating device sends the reply with a hop count equal to one and quality of service information for a first link (hop) toward the initiating device. Each intermediate device increments the hop count in the reply and adds quality of service information for a next link (hop) into the reply.

Abstract: A network is provided that includes a plurality of antennas coupled over the network to a plurality of base stations. The network can be optical or constructed with RF microwave links. The base stations are configured to provide cellular transmission. A plurality of links couple the plurality of antennas and the plurality of base stations. At least one link of the plurality of links provides multiple transmission paths between at least a portion of the base stations with at least a portion of the antennas. In one implementation, at least one link of the plurality of links is shared by at least two cellular operators on different transmission paths. In another implementation, at least a portion of the plurality of base stations are in a common location and at least a portion of the antennas are geographically disbursed.

Abstract: An optical communication network, an optical information transmission method and network node devices for use in an optical communication network, in which optical signals are exchanged via a first data link between a first network node device and a second network node device with interposition of a number of further interconnected network node devices, in which, after a disturbance on the first data link, a third network node device sends a signaling signal to a fourth network node device connected to the third network node device for setting up a second data link which acts at least partially as a standby for the first data link, which signaling signal contains a parameter (NRR, n) determined by the third network node device on the basis of which it is determined whether the fourth network node device is responsible for setting up the second data link or not.

Abstract: An optical communication equipment comprises shared optical sources 88a–88d to be shared by communication nodes 100a–100d, the wavelengths of optical signals 76a–76d are converted into desired wavelengths ?a–?d according to the addressed information of the corresponding optical label signals 77a–77d by using the shared optical sources 88a–88d, and routed to the addressed communication nodes without being converted into electrical signals by using the wavelength routing function of the cyclic-wavelength arrayed-waveguide grating (AWG) 120. The load of each communication node can be reduced by incorporating the multi-wavelength optical sources, which can be shared among individual communication nodes, into the router 80.

Abstract: An optical switch includes a first port, a second port, a first component group, a second component group, and a switching component group. The switching component group includes a reflector, a polarization beam splitter coupled to the reflector, and a polarization switch coupled to the polarization beam splitter. The first component group is coupled between the first port and the reflector in the switch component group. The second component group is coupled between the second port and the polarization beam splitter in the switch component group.

Abstract: An optical communication equipment comprises shared optical sources 88a–88d to be shared by communication nodes 100a–100d, the wavelengths of optical signals 76a–76d are converted into desired wavelengths ?a–?d according to the addressed information of the corresponding optical label signals 77a–77d by using the shared optical sources 88a–88d, and routed to the addressed communication nodes without being converted into electrical signals by using the wavelength routing function of the cyclic-wavelength arrayed-waveguide grating (AWG) 120. The load of each communication node can be reduced by incorporating the multi-wavelength optical sources, which can be shared among individual communication nodes, into the router 80.

Abstract: A technique for establishing automatic service connectivity in a network between multiple network elements is disclosed. Each network element utilizes routing and distribution protocols to discover its neighbors and establish a topology. Optical fibers connect the network elements. Each optical fiber carries multiple wavelengths of signals, wherein the network elements communicate with a server. The method comprises: storing information pertaining to each of the network elements at the server; registering network elements by collecting information about each network element; receiving a connectivity request from a first registered node for connection with a second registered node; determining compatibility of the first and second registered node; and instructing network elements upon verifying compatibility to search for an end-to-end wavelength path and establish a connection between the first registered node and the second registered node.

Abstract: A bit-rate-transparent electrical space-division switching matrix is employed in an optical cross-connect and the input/output stage is constructed from simple, broadband optical receivers and transmitters. Since the switching matrix operates in unclocked manner, i.e. its switching function is not based on internal bit timing and frame timing, arbitrary signals can be switched though transparently at almost any bit rate, independently of the protocol-type being used. The inputs and outputs likewise operate fully independently of bit rate and protocol, since they only implement an O/E conversion or O/E conversion. By virtue of this structure, a simply constructed but extremely powerful optical cross-connect is created that can be employed equally for all types of optical signals within the stipulated wavelength-range.

Abstract: A method for validating a path through a switched optical network is disclosed. A bit error rate at the end of the path is determined. The path is validated if the bit error rate is within a predefined range.

Abstract: An optical wavelength switch having a planar wave guide formed on a substrate is disclosed that comprises a wave-guide-type diffraction grating which includes an input/output wave guide having an under-clad layer on a sacrificial layer formed on the substrate, a core layer formed on the under-clad layer and an over-clad layer formed on the core layer, a first slab wave guide connected with the input/output wave guide, an array wave guide whose one side is connected with the first slab wave guide, and a second slab wave guide with which the other side of the array wave guide is connected; and a movable girder whose one end is firmly secured to the substrate, the movable girder having the same under-clad layer, core layer and over-clad layer as those of the wave-guide-type diffraction grating, wherein the optical wavelength switch has a reflecting mirror at the tip of the movable girder, the reflecting mirror facing an end face of the second slab waveguide, with the position of the reflecting mirror being set d

Abstract: A method is provided for establishing a service connection between first and second network nodes in a WDM optical network for a plurality of network users. The method begins by receiving, from each network user, user-preferences prioritizing a plurality of decision criteria defining preferable characteristics of the service connection. Next, a prescribed algorithm is used to select a path and a channel wavelength at which information is to be conveyed over the path between the first and second nodes. The selection is based on the plurality of decision criteria as prioritized in accordance with the user-preferences. Finally, the first and second network nodes are interconnected over the selected path with the selected channel wavelength.

Abstract: A method and apparatus that recognizes a portion of an address that would be unrecognizable to an intended associated switch or device and manipulates the portion of the address to make it recognizable. The apparatus and method manipulates a discontinuous address to provide the appearance to the associated device, switch or peripheral, that the address is continuous. This provides additional address capacity such that a new address is created within the switch itself for routing data within the switch. All or a portion of the switches in network are preassigned a chassis address, and each chassis also has a specific switch address that is different from the preassigned chassis address. An address adaptor provides translation of addresses and mapping within a switch so that in the event of a port failure, affected frames can be redirected from the failed port by employing the described translation and mapping operations.

Abstract: A technique for routing data within an optical network having a plurality of network nodes is disclosed. In one embodiment, the technique is realized by receiving data at a first network node via a first optical signal having a first wavelength. The first wavelength corresponds to a first optical frequency, and the first optical frequency is mapped to a first binary representation. The first binary representation is divided into a first plurality of fields, wherein at least one of the first plurality of fields corresponds to a routing label in a first label stack. A top routing label in the first label stack indicates a second network node. Based at least partially upon the top routing label, the data is transmitted from the first network node to the second network node via a second optical signal having a second wavelength. The first wavelength may be either the same as or different from the second wavelength.

Abstract: A method of establishing communications in an all optical wavelength division multiplexed network. This method is used in networks built from optically connected communication devices (20, 22) that are capable of converting an optical wavelength to at least one other given optical wavelength, and/or directing a given optical wavelength through an ingress and egress port. Communication is facilitated by devices utilizing a set of given optical wavelengths from the electromagnetic spectrum as a carrier to facilitate communication. A plurality of tables (FIG. 2–4) with varying information such as availability, status, functionality, usage, capabilities, and facilities provide a reference for device resources. Communication paths are created between devices by searching the tables (FIG. 2–4) for information regarding available resources needed to create a path, and then utilizing the available resources with or without optical wavelength conversions between devices (20, 22) to construct a communication path.

Abstract: An optical router includes a substantially transparent dielectric layer, a metal layer with a first side that forms an interface with the dielectric layer, an input optical waveguide, and an output optical waveguide. The input optical waveguide is positioned to illuminate a portion of the interface. The interface is configured to cause light from the input waveguide to produce surface plasmons that radiate light to the output optical waveguide.

Abstract: The photonic network of the present invention uses a cost-effective DWDM optimized switch architecture allowing the introduction of DWDM into the metro network. In this invention the optical carriers are all generated in the photonic layer at the edge photonic switching node and are allocated out to the photonic access nodes or central core data switch for modulation. This has the advantage of providing the optical carriers to be modulated from a centralized highly stable and precise source, thereby meeting the requirements for DWDM carrier precision, whilst generating these carriers in relatively close proximity to the modulators. Sparse WDM components can be used in the access portion of the network without adversely affecting the ability of the signal to transit the DWDM portion of the core network, since the optical carrier frequency is fixed at the centralized source and is unaffected by these components.

Abstract: A sub-network of a communication network includes four nodes, with each node having two input ports and two output ports. The first node and the fourth node each link both of their input ports and one of their output ports to other nodes of the sub-network, with each of their remaining output ports operable to send signals outside of the sub-network. The second and third nodes link both of their output ports and one of their input ports to other nodes of the sub-network while each of their remaining input ports is operable to receive signals from outside of the sub-network.

Abstract: A label switching routing protocol for establishing a datapath as a sequence of locally unique labels in an optical communications network, is provided. A wavelength on an optical cross-connect is considered as a label, or one portion of a label. Timeslots may be assigned to designated wavelengths so as to form the second portion of a composite label. An optical/time cross-connect (OTXC) capable of wavelength conversion from an input to an output interface creates the datapath based on wavelength to wavelength substitution, under the control of a multi-protocol label switching (MPLS) protocol.

Abstract: A technique for providing user domain constrained optical route flooding for multiple optical networking service modes is disclosed. In one embodiment, the technique is realized by a method for routing information over an optical network having multiple optical service models.

Abstract: An apparatus for performing optical routing includes a metal layer having first and second sides, a regular array of structures positioned along the first side, and an input optical waveguide positioned to illuminate a portion of the first side. The illuminated portion of the side is adjacent to ones of the structures. The apparatus also includes a plurality of output optical waveguides positioned to receive light radiated from portions of the metal layer not illuminated by the input optical waveguide.

Abstract: One of optical packets x copied by an optical copying mechanism 11 is converted into an electric packet by an optic/electric converting mechanism 13. An address information extracting mechanism 14 extracts address information from a header of the packet converted into the electric packet. A control light generating mechanism 15 generates a control light based on the extracted address information. The other of the copied optical packets x is delayed by an optical delaying mechanism 12 for a predetermined time, and entered to a third nonlinear optical effect device 1b. The third nonlinear optical effect device 1b switches a route of the optical packet x based on the control light. The optical packet x is outputted through an optic/electric converting mechanism 1h0, a buffer 1i0, a multiplexing mechanism 1j, and an electric/optic converting mechanism 1k.

Abstract: A large fiber optic switch system with a free-space optical interconnection configuration. The switch system comprises a plurality of individual switch units, each individual switch unit having a plurality of electronic multi-switch switches each multi-switch switch being re-configurable upon command of a computer processor and having a plurality of electronic input ports and electronics output ports. A first portion of these input ports and a first portion of these output ports are connected directly or indirectly to incoming and outgoing communication lines. A second portion of the output ports is connected to an electronic driver unit that drives an optical emitter array. Each emitter in the emitter array produces a light beam for carrying an optical communication signal.

Abstract: A path route modification method and a switching system using the method enable to prevent duplicated label setting in the route modification performed in a network. For switching system, a packet switching system disclosed includes a table for mapping either an input label value or a set of input label value and input port to an output port and an output label, wherein a state variable is set for either the input label value or the set of input label value and input port mapped in the table to represent either of the three states of ‘not reserved’, ‘reserved’, and ‘double reserved’; and a packet switch for packet-switching packet data according to the information mapped in the table.

Abstract: A photonic switching device for switching without contention data in the form of optical packets includes a space switching matrix with a plurality of input ports and a plurality of output ports. A unit external to the space switching matrix includes a buffer memory common to all the output ports of the matrix. Each of the output ports provides access to the buffer memory via a space switching stage consisting of switches having a 1-to-2 switching function.