Method and device for photonic pulse modulation

Abstract

The invention relates to a method for generating, starting from a pulse-type optical signal, a train containing a predetermined number of time-shifted photonic pulses with amplitude modulatable as a function of an encoding to be adopted. The method is characterized in that it comprises the steps of: - splitting the optical signal into two optical subsignals; - circulating the optical subsignals in a closed-loop optical circuit in opposite directions so that the photons undergo different delays and / or phases as a function of the path that they follow from the input to the output of the loop, at least one section of said circuit having a multiplicity of routes of different length so as to induce different time delays with respect to the photons traveling along said routes; - acting on the drive signal of at least one phase modulator present in the circuit so as to induce a different phase to the photons having different time delays; - obtaining the desired amplitude modulation from the interference between the optical subsignals exiting the closed-loop circuit. The invention also relates to a device adapted to implement said method.

Description

[0001] METHOD AND DEVICE FOR PHOTONIC PULSE MODULATION

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to the technical field of quantum communications, in particular the encoding of quantum information by means of optical pulse modulation in quantum key distribution (QKD) systems.

[0004] BACKGROUND ART

[0005] Quantum key distribution is a protocol which allows the generation of a secure and private random sequence of (key) bits, shared between two parties by exchanging quantum information encoded through single quantum particles, such as photons. The security of the protocol is ensured by the laws of physics and quantum mechanics and not by assumptions on the computational complexity of a problem as in the case of public key cryptography, such as the Diffie-Hellman protocol, for example. The generated key is mainly applied in cryptography using it with cryptographic techniques, such as one-time-pad. For an overview, see S. Pirandola et al., 'Advances in quantum cryptography," Adv. Opt. Photon., AOP, vol. 12, n. 4, pp. 1012-1236, Dec. 2020, doi: 10.1364 / AQP.361502, which is to be considered an integral part of the present description.

[0006] The encoding of information in the time position of an optical pulse is typically referred to as PPM encoding in the context of traditional communications or timebin encoding in the context of quantum communications.

[0007] In quantum communications, i.e., in the cases in which the optical pulse has less than one photon per pulse, it is particularly important to ensure that both the amplitude and the relative phase of the light in the single time slots are accurately controlled and stable over time.

[0008] The typical approach for generating time-bin-encoded quantum states in quantum applications is to use a laser source pulsed in gain-switch mode, the light of which is processed by a device (e.g., an unbalanced interferometer), which distributes the light over various time slots (2 to d slots) and assigns different phases (<t>i, .., 4>d) to the different pulses. The case of 2 time slots corresponds to the generation of a qubit (i.e., a 2-level quantum system), the case of d slots corresponds to the generation of a qudit (i.e., a d-level quantum system). At the output of the interferometer, the amplitude of each slot can be controlled by an intensity modulator (see Fig. 1 ).

[0009] A commonly used type of unbalanced interferometer is the Mach-Zehnder, with which the relative phase between the generated optical pulses can be controlled by acting with various phase modulators placed on the arms thereof. Alternatively, the relative phases between the pulses (shown in the figure as <J>1 , <J>2, <t>3 and <J>4) can be controlled by a phase modulator placed in cascade with the interferometer. A method alternative to the preceding one is the use of a DC laser and the modulation of the amplitudes in single time slots using an intensity modulator and then of the phase with a phase modulator as shown in Fig. 2. Such a system has the disadvantage that the amplitude of the single pulses depends on the electrical signal applied to the intensity modulator, and therefore the fluctuations in amplitude of the electrical signal affect the optical amplitude. In addition, the phase between optical pulses belonging to different rounds of the protocol must be random to ensure system security. In this configuration, given that the source laser is operated in DC and not in gain-switch mode (which allows the phase to be randomized between each pulse) it is necessary to act on the end phase modulator or add a second phase modulator to actively randomize the phase between successive round pulses.

[0010] The intensity modulator is the element that most affects the performance of these two schemes. Two technologies are used in this area of interest, i.e. , electro-optic effect interference modulators and electro-absorption modulators. Commercial devices of the first type are based on Mach-Zehnder interferometers (Figure 3a). In such interferometers, the relative phase between the arms is subject to fluctuations due to variations in temperature and mechanical stress, and thus they need external hardware to compensate for the relative phase fluctuations and maximize the quenching capability. The latter, on the other hand, are based on materials that vary their absorption coefficient as a function of an external electrical stimulus, as described for example in R. Amin, J. B. Khurgin, and V. J. Sorger, "Waveguide-based electro-absorption modulator performance: comparative analysis," Opt. Express, OE, vol. 26, n. 12, pp. 15445-15470, June 2018, doi: 10.1364 / OE.26.015445. Both are characterized by limited quenching capabilities, typically around 35 dB. On the other hand, an interferometric scheme not requiring compensation is that based on the Sagnac interferometer (Figure 3b). A phase modulator is inserted into the loop in a position generally not symmetric to the center, thereby the pulses traveling clockwise and counterclockwise meet the modulator at different time instants.

[0011] Defining the optical route time difference as T and the beamsplitter transmittance as n, the expression of the transmitted field b(t) in response to an input mode a(t) is

[0012] The modulus and phase of the pulse transmitted by the asymmetric Sagnac interferometer depend on the difference in the value of the phase applied by the phase modulator between two time instants separated by the interferometer unbalance. Such a behavior is because the light injected at the input of the beamsplitter is split by such an element into two components which propagate in the two opposite directions and, because of the unbalance, cross the phase modulator at different times. Instead, in the Mach-Zehnder interferometer, the amplitude of the output signal is a function of the instantaneous phase difference applied between the two arms. For such reasons, the Mach-Zehnder scheme is often used to control the shape of optical pulses from a continuous signal with accuracy, given that the temporal shape of the control signal is mapped in amplitude as the sine square thereof, whereas the Sagnac-type scheme is more easily applicable for the amplitude modulation of pulsed optical signals because the modulation acts on both the co-propagating and counter-propagating modes within the loop as described in G. L. Roberts et al., “Patterning-effect mitigating intensity modulator for secure decoy-state quantum key distribution”, Optics Letters, vol. 43, n. 20, page 5110, Oct. 2018, doi: 10.1364 / ol.43.005110.

[0013] For the generation of 2-dimensional time-bin states (qubits), it is also possible to exploit the conversion from polarization coding, as described for example in D. Scalcon et al, "Cross-encoded quantum key distribution exploiting time-bin and polarization states with qubit-based synchronization". Advanced Quantum Technologies, October 17, 2022, doi: 10.1002 / qute.202200051 . The scheme needs a polarization beamsplitter, an optical device which propagates the vertical and horizontal components of the light it receives as input into two different outputs. A Mach-Zehnder-style scheme is used to achieve the polarization-to-time- bin conversion, in which the input beamsplitter is replaced by a polarization beamsplitter (see Fig. 4). The amplitude and phase of the two polarization components are thus mapped directly into the amplitude and phase of the two pulses.

[0014] An approach that does not make use of phase and intensity modulators is that described, for example, in T. K. ParaTso et al, 'A modulator-free quantum key distribution transmitter chip," npj Quantum Inf, vol. 5, n. 1 , Art. no. 1 , May 2019, doi: 10.1038 / s41534-019-0158-7, which exploits the effect of optical injection between two lasers, also referred to as "optical injection locking". A first laser (master) injects light into a second laser (slave) by means of an optical circulator or beamsplitter (Fig. 5). Due to the light from the master, the phase of the light produced by the slave remains constant even if the diode is completely switched off and on again. Moreover, a change in the amplitude of the optical signal of the master over time induces a change in the phase of the slave.

[0015] In time-bin state generation, the intensity in the single time slot is controlled by the electrical drive pulse of the slave laser, while the relative phase is modified by varying the current of the master around a set point.

[0016] This approach has the simplest optical scheme but requires an additional laser not present in the other schemes.

[0017] It is the object of the present invention to provide a method and a device for the preparation and encoding of quantum states which allow improving the performance of QKD transmitters according to the prior art, in particular by reducing the quantum bit error rate (QBER), defined as the probability of detecting a photon in a quantum state orthogonal to the prepared one.

[0018] It is a further object to provide a device which is intrinsically stable over time and does not require the control or compensation of some parameter to ensure the functionality thereof so as to reduce the complexity of time-bin QKD systems in which it is used. BRIEF DESCRIPTION OF THE INVENTION

[0019] The invention achieves the objects with a method for generating, from an optical pulse, a train containing a predetermined number of photonic pulses having determined time separation and optical amplitude and phase as a function of an encoding to be adopted. The method comprises:

[0020] - splitting the optical pulse into two optical subsignals;

[0021] - circulating the optical subsignals in a closed-loop optical circuit in opposite directions so that the optical pulses undergo different delays and / or phases as a function of the path that they follow from the input to the output of the loop, at least one section of said circuit having a multiplicity of routes of different lengths to induce different time delays with respect to the signal portion traveling along said routes;

[0022] - acting on the drive signal of at least one phase modulator present in the circuit so as to induce a different phase to the optical pulses having different time delays;

[0023] - obtaining the desired amplitude and phase modulation from the interference between the optical subsignals exiting the closed-loop circuit.

[0024] In practice, the pulse signal exiting from a laser, after being split by a beamsplitter with a defined transmission-to-reflection ratio, e.g., of the 50:50 type, is processed in a closed-loop optical circuit which is typical of Sagnac-type intensity modulators in which an unbalanced Mach-Zehnder interferometer is inserted with a number of ways equal to the number of temporally offset pulses to be generated.

[0025] Multiple configurations are possible. For example, in a first embodiment, the method includes:

[0026] - splitting one of the two optical subsignals in its path from the input to the output of the loop into two or more signal parts, subjecting each signal part to different time delays, recombining the signal parts, modulating the signal as recombined, and completing the path up to the loop output;

[0027] - applying a phase modulation to the other of the two optical subsignals, splitting the subsignal as modulated into two or more signal parts, subjecting each signal part to different time delays, recombining the signal parts, and completing the path up to the loop output.

[0028] Instead, another configuration includes: - splitting one of the two optical subsignals in its path from the input to the output of the loop into two or more signal parts, applying a phase modulation to each signal part, and subjecting each signal part as modulated to different time delays or conversely subjecting each signal part to different time delays and applying a phase modulation to each signal part as delayed, recombining the signal parts and completing the path up to the loop output;

[0029] - applying a phase modulation to the other of the two optical subsignals, splitting the subsignal as modulated into two or more signal parts, subjecting each signal part to different time delays, recombining the signal parts, and completing the path up to the loop output.

[0030] The various configurations depend on the position of the phase modulator(s) inside the loop so as to ensure the suitable phase shifting of the pulses upstream or downstream of the delay lines which create the time-bins.

[0031] In an advantageous form, the use of expensive N-way splitters can be avoided by:

[0032] - splitting one of the two optical subsignals in its path from the input to the output of the loop into two signal parts in a recursive manner so as to obtain the same splitting result of the subsignal into more than two signal parts, applying a phase modulation to each signal part and subjecting each signal part as modulated to different time delays or conversely subjecting each signal part to different time delays and applying a phase modulation to each signal part as delayed, recombining the signal parts and completing the path up to the loop output;

[0033] - applying a phase modulation to the other of the two optical subsignals, splitting the subsignal as modulated into two signal parts, in a recursive manner so as to obtain the same splitting result of the subsignal into two or more signal parts, subjecting each signal part to different time delays, recombining the signal parts and completing the path up to the loop output.

[0034] Moreover, as for the drive signal of the modulator(s), it advantageously has the form of a train of rectangular waves of different amplitudes and time-shifted so as to induce a different phase with respect to the photons that cross said modulator at different time instants. Other waveforms to drive the modulator are also possible.

[0035] If there are multiple drive signals from different modulators, such drive signals can have, for example, a rectangular waveform with the same time delay but with different amplitudes so as to induce a different phase to photons that cross said different modulators with the same time delay. It is thus possible to drive modulators also present in the branches dedicated to the creation of pulse bursts, i.e. , in the section containing the Mach-Zehnder interferometer.

[0036] According to another aspect, the invention further relates to a device for generating, from an input pulse-type optical signal, an output train containing a predetermined number of optical pulses that are time-shifted and of modulatable amplitude as a function of an encoding to be adopted. The device comprises:

[0037] - a first beamsplitter capable of splitting an input optical signal on one port into two output optical subsignals on two ports and vice versa of merging two input optical signals on two ports into a single output signal on one port;

[0038] - an optical loop circuit arranged downstream of the first beamsplitter and comprising two circuit branches connected at one end with the opposite ends connected to two ports of the first beamsplitter so that the input signal to the first beamsplitter is split into two optical subsignals which circulate in the closed-loop optical circuit in opposite directions from one end to the other end and vice versa so that the two subsignals can undergo different conditioning in their path from one end of the loop to the other end and then merge into a single output signal when they reach the first beamsplitter again;

[0039] - a second and a third beamsplitter capable of splitting an input optical signal on one port into two or more parts of an output signal on two or more ports and vice versa of merging two or more parts of an input signal on two or more ports into a single output signal on one port, said second and third beamsplitters being connected in series on a section of the loop so as to split the subsignal which circulates in one direction or the opposite direction into two or more signal parts in subsections of the loop and merge said signal parts into a single signal;

[0040] - one or more delay lines connected in series between the second and third beamsplitters so that the signal parts undergo different delays;

[0041] - a phase modulator capable of modulating the phase of the optical signals which cross it as a function of a drive pulse, said phase modulator being connected in series on the other branch of the loop.

[0042] Additional phase modulators can be present in the subsections of the loop either upstream or downstream of the delay lines so as to modulate the phase of the signal parts that pass through said subsections.

[0043] According to an improvement, for configurations with a number of time-bin states n greater than two, the second and third beamsplitters, instead of being of the n-way type, are of the two-way type capable of splitting a signal into two parts and conversely merging two signals into one part. Additional phase modulators and two-way beamsplitters are further present in series in the subsections of the loop so as to further split the signal parts present in the subsections into additional signal subparts and modulate the phase of such signal subparts obtaining the same result that would be achieved using multiway beamsplitters capable of operating the splitting / merging of a signal into more than two parts.

[0044] According to a further aspect, the invention relates to the use of the aforesaid device for generating quantum states qudits, particularly qubits, with time-bin encoding for quantum applications, in particular for quantum key distribution.

[0045] According to an aspect, the invention also relates to a quantum key distribution transmitter comprising a laser source of pulsed optical pulses, connected to a device according to the invention.

[0046] Further features and improvements are the subject of the dependent claims.

[0047] BRIEF DESCRIPTION OF THE FIGURES

[0048] Further features and advantages of the invention will be apparent from reading the following detailed description, given by way of a non-limiting example, with the aid of the figures shown in the accompanying drawings, in which:

[0049] Fig. 1 shows an example of a block diagram of a qudit generator with time-bin encoding according to the prior art;

[0050] Fig. 2 shows another example of a block diagram of a qudit generator with timebin encoding according to the prior art;

[0051] Fig. 3 shows two alternative schemes of an intensity modulator in the preceding figures;

[0052] Fig. 4 shows a generator scheme with time-bin coding according to the prior art which exploits the polarization coding;

[0053] Fig. 5 shows an additional generator scheme with time-bin coding according to the prior art which employs a pair of lasers instead of phase and intensity modulators; Fig. 6 shows the exemplified principle scheme of the modulator according to the invention (Fig. 6b) compared to the scheme of a modulator according to the prior art (Fig. 6a);

[0054] Fig. 7 diagrammatically shows the phase pattern of optical pulses at various points in a circuit according to an embodiment of the invention which involves using a two-arm Mach-Zehnder interferometer with a single phase modulator in which the input optical pulse is equally split between clockwise and counterclockwise propagating;

[0055] Fig. 8 shows the same scheme as in the preceding figure with a general N-arm Mach-Zehnder interferometer with two examples of drive signals of the phase modulator highlighted;

[0056] Fig. 9 shows another embodiment of the modulator according to the invention in which there are multiple phase modulators with two possible drives highlighted;

[0057] Fig. 10 shows two variants in which multiple 2-way Mach-Zehnder interferometers are used recursively and, more in general, multiple two-way beamsplitters and delay lines with or without phase modulators in series are used instead of using a 4-way Mach-Zehnder interferometer to obtain 4-state qudits;

[0058] Fig. 11 is a variant of Fig. 10 in which the time slots on which each phase modulator can act are highlighted with numbers from 1 to 4.

[0059] Figs. 12-17 show some variants of the preceding schemes in which semiconductor optical amplifiers (SOAs) are used in place of and / or in addition to at least some of the phase modulators.

[0060] The following description of exemplary embodiments relates to the accompanying drawings. The same reference numbers in the various drawings identify the same elements or similar elements. The following detailed description does not limit the invention. The scope of the invention is defined by the appended claims.

[0061] DETAILED DESCRIPTION OF THE INVENTION

[0062] With reference to Fig. 6a, in a typical optical circuit of the prior art, a train of pulses is generated by an unbalanced Mach-Zehnder interferometer 1 and the intensity in the single time-bin is controlled by an intensity modulator 2 based on a Sagnac scheme.

[0063] In contrast, the idea behind the present invention is to place the Mach-Zehnder 1 interferometer inside the Sagnac interferometer 2, as shown in Fig. 6b. This solution provides several advantages including simplicity in controlling the generated states as explained below.

[0064] The operating principle is as follows. A different phase applied by the modulator 3 (shown as <t>-mod in the figure) to the signals propagating clockwise and counterclockwise in the Sagnac loop 2 generates constructive or destructive interference to the input beamsplitter 4, modulating the ratio of transmitted to reflected light. Let us see in detail how this occurs.

[0065] In Fig. 7 there is depicted the phase variation of the optical pulses at various points in one of the possible implementation schemes, with two-arm Mach- Zehnder interferometer and single phase modulator. The input optical pulse is split between clockwise and counterclockwise propagating, indicated as Ecw and Eccw, respectively. If the input beamsplitter 4 is 50:50, if Eo is the complex amplitude of the input pulse, the two pulses will have the form:

[0066] The Ecw component, traveling clockwise, will first meet the phase modulator 3, which will apply a certain phase D, and then the Mach-Zehnder interferometer 1 , which will distribute the light in the two time slots, resulting in a pair of pulses with identical phase spaced by the unbalance T of the interferometer. The Eccw component, which travels counterclockwise instead, will first be processed by the interferometer 1 and will then cross the phase modulator 3. In this case, the latter can apply a different phase E to the time-bin traveling early (early component) and <t>L to that traveling late (late component).

[0067] The additional factor 1 / 2 and the minus sign of the late component are both introduced by the Mach-Zehnder interferometer 1 .

[0068] By recombining at the input beamsplitter 4, the light from the two routes interferes, thus allowing the modulation of the signal intensity in the two time-bins. If the width of the optical pulse is smaller than the optical route difference in the interferometer, which is always true in the case of time-bin state generation, the two do not interfere and the respective amplitudes can be analyzed separately, since E(early)cw interferes only with E(early)ccw, while E(late)cw interferes only with E(|ate)ccw.

[0069] A clearer form that the preceding ones can be found by means of a simple manipulation. Taking, for example, the early pulse:

[0070] Taking the square module, the response becomes:

[0071] The advantages over the prior art largely depend on the specific configuration, in terms of the number of time-bins of the produced state and the position of the phase modulators. In particular, the invention allows decreasing the size of the device and increasing the stability and accuracy of the generated quantum states. We will now see some constructional variants.

[0072] Topology with single phase modulator

[0073] In the case of implementations with commercial optical fiber components, it can be cost-effective to minimize the number of components when implementing the setup. The phase modulator is a component with an often high cost. Hence the scheme in Fig. 8, which diagrammatically depicts a solution employing a single phase modulator 3 and an N-way Mach-Zehnder interferometer 1 .

[0074] The train of electrical pulses 5 applied to the phase modulator 3, synchronous with the passage of light which has first crossed the Mach-Zehnder interferometer 1 , selects the amplitude value for each time-bin. It is also possible, by acting on the light crossing the phase modulator 3 in a clockwise direction, to apply a common phase to all time-bins, which will then be reflected in a common offset in the amplitude of the output pulses.

[0075] In the present description, the electrical drive signals are assumed to be square or pulse signals. This simplifies the description of the device operation, but it does not mean that the drive signals can take different forms.

[0076] In the figure, the signal that generates the offset is denoted by Vo while Vi, V2,..VN denote the modulator control signals acting on time-bins 1 , 2, ..N, respectively. In the example in Fig. 8a, Vo is 0 so there is no offset in the output signal 6, while in the example in Fig. 8b, Vo is different from 0 while the signals Vi, V2,..VN are 0. As a result, in Fig. 8a the output signal 6 has substantially the same trend as the drive signals Vi, V2,..VN while in Fig. 8b the output signal has pulses all of the same amplitude due to the offset introduced by the drive signal Vo.

[0077] This feature is useful for encoding two-dimensional states for QKD applications. Indeed, it results that in such a configuration it is very easy to generate the three states required for the 3-state efficient BB84 protocol, i.e., the early (light present in the first time slot only), late (light present in the second one only) or superimposed state of the two.

[0078] That seen above is the particular single phase modulator configuration which provides a greater advantage over the prior art when with a single electrical signal is not possible to control all the time-bins. Some alternative configurations are shown below.

[0079] Topology with phase modulators in the single arms of the interferometer

[0080] In setups obtained with integrated optics, the cost difference in adding or not adding an element, such as a phase modulator, is less significant than in discrete- element setups. In this case, it is possible to add a phase modulator 3', 3" in each branch of the Mach-Zehnder interferometer 1 as shown in Fig. 9.

[0081] Such an approach simplifies the logic for the generation of drive signals and allows controlling the pulses, with the same available electronic bandwidth, at a shorter time interval, thus increasing the possible generation frequency. This is because in the case with single modulator, the pulses are serialized at the modulator input, whereas in this case they must be generated in parallel, which relaxes the electronic bandwidth requirements. Moreover, the single modulators can be driven with fixed-amplitude signals so as to reduce the complexity of the electronic drive system.

[0082] The examples shown in Fig. 9a and 9b relate to the generation of the same output signals 6 as in Figs. 8a and 8b. In this case, having inserted N phase modulators into the arms of the Mach-Zehnder interferometer 1 , the drive signal splits into N drive signals that take the same form as the Vi, V2,..VN pulses (not shown in Fig. 9b because null) but with the same time offset. The modulator 3 is also driven by a single-pulse signal Vo which provides the desired offset (not shown in Fig. 9a because null).

[0083] The advantage in terms of required electrical bandwidth is the main advantage of this embodiment over the prior art. In schemes based on intensity modulators, as well as in the single phase modulator scheme presented above, the electrical signal must be able to switch in times comparable to the time-bin. However, in this topology, the signal at each phase modulator will have the same period as the transmitter repetition rate. There also remains the advantage, presented above, of the generation of two-dimensional states using the phase modulator 3 not inserted into the Mach-Zehnder interferometer 1.

[0084] The N-way Mach-Zehnder interferometer described in the preceding sections can be made with different combinations of elements. Some of them are shown below. For fiber implementations, the beamsplitters commonly available on the market are of the 2x2 type, i.e. , with two inputs and two outputs. Using only such devices, it is however possible to make configurations generally capable of generating states with an arbitrary number of time-bins by placing a greater number of beamsplitters or interferometers in cascade, as shown in Fig. 10. If allowed by the manufacturing technology, the number of elements required can be reduced by using N-way or partially 2-way and N-way beamsplitters.

[0085] Such a configuration brings advantages in terms of device operability. In Fig. 11 is shows how in such a topology the same phase modulator can act on multiple time slots; numbering from 1 to 4 the time slots exiting from the interferometer, the corresponding number is associated with the phase modulator which can be used to act on the respective optical pulse. Again in this topology, the effect of the applied modulation can act on either the phase or the amplitude of the respective optical pulse, according to the driving method used.

[0086] In the configuration shown at the top in Fig. 10, the beamsplitters of the Mach- Zehnder interferometer 1 are six of the two-way type 11 , 11', 13, 13', 14, 14’ instead of being two of the N-way type. The signal, e.g., the one circulating counterclockwise, is first split into two parts which are in turn split into two additional parts, each after undergoing a different delay. In the example, the part passing through the top underwent a delay T due to the delay line 12 while the second one underwent no delay. At this point, the two signal parts undergo a further splitting into two further parts which, after having traveled a phase modulator (in those configurations in which it is provided because this is an entirely optional component), are subjected to further delay, e.g., T / 2 in the delay lines 15, 16 the first and zero the second, and then merge first into two and then into one as shown in the figure thus obtaining the same result as would be obtained by using two four-way beamsplitters capable of operating the splitting / merging of a signal into four parts.

[0087] In the variant shown at the bottom in Fig. 10, the signal parts are merged in the middle of the circuit of the Mach-Zehnder interferometer 1 to be split in two again so as to obtain a circuit topology with only two delay lines 12, 15 and three beamsplitters 11 , 13, 1 T. As usual, the modulators in the arms of the Mach- Zehnder interferometer 1 can or cannot be present according to the configuration adopted, i.e. , it is possible to use configurations similar to those shown in Fig. 8 in which there is only one modulator or only one part of modulators.

[0088] The modulator 3 outside the Mach-Zehnder interferometer 1 can also be provided in a different position than that seen so far. For example, it can be placed in series with the Mach-Zehnder interferometer 1 as shown in Fig. 11 for the case with 2x2 beamsplitters in cascade. In this figure, the numbers 1 to 4 denote the time slots on which each phase modulator can act.

[0089] Indeed, by acting on the drive pulses of the modulator(s), it is possible to induce different phase modulations to pulses having different time delays whatever technique is adopted to obtain said delayed pulses. These drive pulses can be in sequence as well as in parallel or partially in sequence and partially in parallel as a function of the position and number of modulators used.

[0090] Possible implementation technologies

[0091] The optical implementation of the invention can be achieved using fiber, integrated optical circuits or discrete components in a free space. In optical fiber, the scheme is conveniently made with polarization-maintaining fibers, which makes the response of the device more stable because the polarization of the light propagating in the phase modulator is thus fixed. In the integrated, free-space optics, the idea applies to any technology.

[0092] Applications

[0093] The invention mainly finds application in QKD for generating time-bin states, typically used in optical fiber channels. However, its use as an encoder for traditional optical communications, in modulations which exploit the temporal degree of freedom such as pulse position modulation (PPM), in combination with phase shift keying (PSK), is possible.

[0094] The invention can be widely varied. For example, configurations in which the input and output occur on the same port of the beam splitter 4 can be provided. This allows it to be more robust with respect to the imperfections of the beam splitter.

[0095] A further variant involves the use of one or more semiconductor optical amplifiers (SOA). These devices allow both attenuating the light if they are not subjected to electronic driving and amplifying it, as a function of the amount of current applied. Therefore, for our purposes, the SOA can be used as a light-intensity modulator.

[0096] In a configuration, the SOA can be inserted into the inner branch of the Sagnac interferometer, as shown in Fig. 12, in both branches of the Mach-Zehnder as in Fig. 13, or in all three branches of the circuit as shown in Fig. 14.

[0097] The advantage of such configurations is that the SOA can be used to change the intensity of the optical pulses within the loop, allowing decoy states to be created more easily.

[0098] The decoy modulation is thus carried out by driving SOAs, while the time-bin creation is carried out by driving the phase modulators.

[0099] Given this decoupling, 2 advantages emerge:

[0100] - The electrical signals sent to the various phase modulators are easier to generate (they are only binary signals and not multi-level);

[0101] - Phase modulators only ever impart iT-modulation, equivalent to cancellation of the derivative of the transfer function, and therefore are more robust to noise and fluctuations in the electrical signal.

[0102] Another variant of the device that takes advantage of SOAs involves using the same port as input and output and inserting one or more SOAs, without phase modulators, as shown in Fig. 15. The advantage of this solution is that it does not require phase modulators, which are generally more unstable than SOAs and sometimes more expensive.

[0103] In this case, if the SOA is turned off during the passage of one of the two counterclockwise propagating pulses, one of the two pulses can be suppressed, creating the early or late states.

[0104] The modulation of the clockwise propagating pulse, on the other hand, allows the creation of the superimposed state, with intensity per pulse equal to half the early or late pulses.

[0105] In a specular manner, it is possible to have SOAs inside the Mach-Zhender interferometer as shown in Fig. 16.

[0106] Switching off one SOA at a time allows the creation of the early or late states, while modulating both SOAs simultaneously allows the creation of the superimposed state.

[0107] Any combination of these solutions is obviously possible, as shown in Fig. 17, for example.

[0108] Finally, by using multilevel modulations toward the SOAs it is also possible to create both time-bin and decoy states.

[0109] Obviously, this discussion can also extend directly to the multi-dimensional case, i.e., with n branches present within the Sagnac loop, one or more SOAs can be employed, indeed, with or without phase demodulators, to achieve the same advantages as the one-dimensional solution, this without abandoning the informing principle stated above and claimed below.

Claims

CLAIMS1 . A method for generating, starting from a pulse-type optical signal, a train containing a predetermined number of time-separated photonic pulses with relative phases and amplitudes modulatable as a function of an encoding to be adopted, the method comprising:- splitting the optical signal into two optical subsignals;- circulating the optical subsignals in a closed-loop optical circuit in opposite directions so that the photons undergo different delays and / or phases as a function of the path that they follow from the input to the output of the loop, wherein at least one section of said circuit has a multiplicity of routes of different lengths so as to induce different time delays with respect to the photons traveling along said routes;- acting on the drive signal of at least one phase modulator present in the circuit so as to induce a different phase to the photons having different time delays;- obtaining the desired amplitude modulation from the interference between the optical subsignals exiting the closed-loop circuit.

2. A method according to claim 1 , wherein it provides:- splitting one of the two optical subsignals in its path from the input to the output of the loop into two or more signal parts, subjecting each signal part to different time delays, recombining the signal parts, modulating the signal as recombined, and completing the path up to the loop output;- applying a phase modulation to the other of the two optical subsignals, splitting the subsignal as modulated into two or more signal parts, subjecting each signal part to different time delays, recombining the signal parts, and completing the path up to the loop output.

3. A method according to claim 1 or 2, wherein it provides:- splitting one of the two optical subsignals in its path from the input to the output of the loop into two or more signal parts, applying a phase modulation to each signal part, and subjecting each signal part as modulated to different time delays or conversely subjecting each signal part to different time delays andapplying a phase modulation to each signal part as delayed, recombining the signal parts and completing the path up to the loop output;- applying a phase modulation to the other of the two optical subsignals, splitting the subsignal as modulated into two or more signal parts, subjecting each signal part to different time delays possibly applying a phase modulation to each signal part as delayed, recombining the signal parts, and completing the path up to the loop output.

4. A method according to one or more of the preceding claims, wherein it provides:- splitting one of the two optical subsignals in its path from the input to the output of the loop into two signal parts in a recursive manner so as to obtain the same splitting result of the subsignal into more than two signal parts, applying a phase modulation to each signal part and subjecting each signal part as modulated to different time delays or conversely subjecting each signal part to different time delays and applying a phase modulation to each signal part as delayed, recombining the signal parts and completing the path up to the loop output;- applying a phase modulation to the other of the two optical subsignals, splitting the subsignal as modulated into two signal parts, in a recursive manner so as to obtain a same splitting result of the subsignal into two or more signal parts, subjecting each signal part to different time delays possibly applying a phase modulation to each signal part as delayed, recombining the signal parts and completing the path up to the loop output.

5. A method according to one or more of the preceding claims, wherein the drive signal of at least one modulator has the form of a train of waves of arbitrary form and different amplitudes, time-shifted so as to induce a different phase with respect to the photons that cross said modulator at different time instants.

6. A method according to one or more of the preceding claims, wherein there are several different modulator drive signals having an arbitrary waveform with the same time delay but with different amplitudes so as to induce a different phase to the optical signals that cross said different modulators with the same time delay.

7. A method according to one or more of the preceding claims, wherein, for the generation of qubits, i.e. , two-state optical signals, it is provided making optical subsignals circulate in two paths having different lengths in a section of the loop so as to generate a pair of pulses (Eearly, Elate) time-shifted by a given delay (T), modulating the phase of said subsignals so that the pulses that circulate in one direction of the loop are subjected to the same phase modulation (<t>D) while the pulses that circulate in the opposite direction are subjected to a different phase modulation<fi) so that, called Ein the signal entering the loop and Eout the signal exiting the loop, the relation applies:

8. A device for generating, starting from an input pulse-type optical signal, an output train (6) containing a predetermined number of photonic pulses which are time-shifted and of amplitude modulatable as a function of an encoding to be adopted, which device comprises:- a first beamsplitter (4) capable of splitting an input optical signal on one port into two output optical subsignals on two ports, and vice versa of merging two input optical signals on two ports into a single output signal on one port;- an optical loop circuit (2) arranged downstream of the first beamsplitter and comprising two circuit branches connected at one end with the opposite ends connected to two ports of the first beamsplitter (4) so that the input signal to the first beamsplitter is split into two optical subsignals which circulate in the closed- loop optical circuit (2) in opposite directions from one end to the other end and vice versa so that the two subsignals can undergo different conditioning in their path from one end of the loop (2) to the other end and then merge into a single output signal when they reach the first beamsplitter (4) again;- a second (11 ) and a third (11 ') beamsplitter capable of splitting an input opticalsignal on one port into two or more parts of an output signal on two or more ports and vice versa of merging two or more parts of an input signal on two or more ports into a single output signal on one port, said second and third beamsplitters (11 , 11') being connected in series on a section of the loop so as to split the subsignal which circulates in one direction or the opposite direction into two or more signal parts in subsections of the loop (2) and merge said signal parts into a single signal;- one or more delay lines (12, 15, 16) connected in series between the second and third beamsplitters (11 , 11') so that the signal parts undergo different delays;- a phase modulator (3) capable of modulating the phase of the optical signals which cross it as a function of a drive pulse (5), said phase modulator being connected in series on the other branch of the loop (2).

9. A device according to claim 8, wherein additional phase modulators (3, 3') are present in the subsections of the loop (2) either upstream or downstream of the delay lines (12, 15, 16) so as to modulate the phase of the parts of the signal that pass through said subsections.

10. A device according to claim 8 or 9, wherein the second and third beamsplitters (11 , 11') are of the two-way type capable of splitting a signal into two parts and vice versa of merging two signals into one part there being additional phase modulators (3', 3") and two-way beamsplitters (13, 13', 14, 1 ') in series in the subbranches of the loop (2) so as to further split the signal parts present in the subbranches into further signal subparts and modulate the phase of these signal subparts obtaining the same result as using multiway beamsplitters capable of operating the splitting / merging of a signal into more than two parts.

11. A device according to one or more of the preceding claims 8 to 10, wherein the loop circuit is a Sagnac interferometer (2) in which an unbalanced Mach- Zehnder interferometer (1) with a number of ways equal to the number of timeshifted pulses to be generated is inserted on one branch.

12. A device according to one or more of the preceding claims, wherein at leastone semiconductor optical amplifier (SOA) is present in series with at least one phase modulator in the circuit to change the intensity of the optical pulses inside the loop.

13. A device according to one or more of the preceding claims, wherein at least one semiconductor optical amplifier (SOA) is present instead of at least one phase modulator of the circuit, which semiconductor optical amplifier can be switched off, to suppress at least one pulse traveling in one direction of the circuit so that the early or late states can be created, and controlled in modulation to modulate the pulses traveling in the opposite direction so that the superimposed state can be created, with pulse intensity equal to half of the early or late pulses.

14. Use of the device according to one or more of the preceding claims 8 to 13 for generating quantum qudit states, particularly qubits, with time-bin encoding for quantum applications, in particular for quantum key distribution.

15. A quantum key distribution transmitter comprising a connected laser source of pulsed optical pulses and a device according to one or more of claims 8 to 13.

Images