Abstract: Logic operations are carried out on binary optical signals using a non-linear optical amplifier having a front port and a rear port respectively receiving a front wave and a modulated rear wave. The output wave via the rear port is separated from the rear wave by exploiting the fact that they have opposite propagation directions. Applications include optical data switching and transmission.

Abstract: In an optical switching matrix, a common waveguide and a plurality of individual waveguides are connected by a selector that connects the common waveguide to a selected individual waveguide. The individual waveguides are doped and light from a pump is injected into the common waveguide towards the selector. The selector selectively transmits this light to the selected individual waveguide. The resulting amplification in this waveguide reduces crosstalk. One application of these optical switching matrices is in fiber optic telecommunication networks.

Abstract: The network includes composite switching matrices (M1, M2, M3), each of which is split between a plurality of nodes (N1, N2, N3) of the network. The nodes are connected together via optical fibers (L3A3, L1A4) conveying wavelength multiplexes and included in the matrices. The matrices include emitters having staggered wavelengths (E2A1,1, . . . , E2A2,4), star couplers (C2A1, C2A2), controlled space-division switching matrices (X2A1, . . . , X2A4), and wavelength-controlled filters (F2A1,1, . . . , F2A4,4). The invention applies to telecommunications.

Abstract: On the input side of the matrix, a set of emitters (E1,1, . . . , Em,n) receive data to be transmitted via the matrix. The set of emitters emit waves carrying the data at controlled wavelengths. A set of filters (F1,1, . . . , Fn,m) transmit the waves at the output of the matrix selectively as a function of the wavelengths of the waves. The waves are transmitted from the emitters to the filters firstly via a group of switching distributors (X1, . . . , Xm), each of which is capable of connecting each of its inputs to each of its outputs on command, and secondly via a group of star distributors (C1, . . . , Cn), each of which permanently connects all of its inputs to all of its outputs. A particular application of the invention is to a telecommunications network when a data collection operation is to be performed simply.

Abstract: In an optical frequency marking method and a frequency channel communication network using this method, an optical reference frequency is scanned across a scanning band to produce frequency coincidences with a monitored frequency which constitutes the frequency to be marked. During each scanning half-cycle it effects a go path and a return path according to a known law to produce a go frequency coincidence and a return frequency coincidence. A marker interval is measured which is the time elapsed between said two coincidences. This interval marks the monitored frequency and enables said frequency to be locked by comparison with a set point interval. The invention finds a particular application in the marking and stabilization of carrier frequencies of a closely-spaced different frequency channel network.

Abstract: According to the invention, messages to be transmitted between user terminals interconnected by a transmission network are classified into "heavyweight" messages and "lightweight" messages. The heavyweight messages, i.e. messages containing a large quantity of data, are transmitted over the network in message frequency channels attributed to respective heavyweight messages, while the lightweight messages are transmitted over a shared channel common to the lightweight messages.

Abstract: In a method of the invention an emission wave (F(2P-1)) which is to be positioned within the spectrum relative to an external wave (F(2P-2)B) is modulated. Said external wave is mixed with one of the side waves that results from said modulation (F(2P-1)A) and the emission wave is positioned in response to variations in the frequency of a beat signal (F(2P-2)BE) that results from said mixing. The invention applies in particular to implementing an optical fiber communications network.

Abstract: Calls are made in known manner between terminals in the form of information-carrying modulation applied to emission waves that travel along common transmission lines. An emission frequency (F(2P-1)) is allocated to each terminal for each call. According to the invention, a marking wave is supplied to all of the terminals while no call is yet being set up, thereby marking an emission frequency. When a new call is being set up, the marked frequency is allocated for the new call to at least one of the terminals that are to participate in the call. The invention is particularly applicable to communications by means of optical fibers.

Abstract: The emission frequencies (F(1) . . . F(2P)) used for simultaneous communications form a stack extending from a fixed based frequency (F(0)) to a top of stack frequency (F(2P-1), F(2-P)). At the beginning of each call, the emission frequency (F(2P-1)) used for the call is at the top of the stack. During the call, said emission frequency is maintained by being servo-controlled to a predetermined spectrum distance (DF) beyond a lower support frequency (F(2P-2)) providing such a frequency can be detected. When such a frequency cannot be detected, the emission frequency is shifted progressively towards the base frequency. The invention is particularly applicable to communications between the peripherals of computer systems.