A quantum computing module and a method for processing quantum propagation modes to perform quantum computing.

Abstract

The present invention relates to a quantum computing module, comprising an input section (2) equipped with a mode separator (15) configured to separate quantum transverse electric modes TE and transverse magnetic modes TM into two parallel paths, each path comprising an input section (2) characterized by a single-mode waveguide (4) and a mode processing means (20) including at least a first phase shifting means (22), and an output section (3) including a mode combining means (25), wherein at least the first path (5) includes a first conversion means (30) that converts the transverse magnetic mode TM into a transverse electric mode TE located upstream of the processing means (20), and a second conversion means (32) that converts the transverse electric mode TM into a transverse magnetic mode TM located downstream of the processing means (20).

Description

[Technical Field]

[0001] The present invention relates to a quantum computing module and a method for processing quantum propagation modes to perform quantum computing.

[0002] This module includes a photon quantum circuit, enabling it to perform the fundamental operations of quantum computing. It is a versatile, programmable quantum computing module that can be placed at each node of a general quantum circuit to perform computer-based operations (quantum and classical operations using a Toffoli configuration). This module can also be hybridized or connected to other classical or quantum devices besides photon devices.

[0003] Quantum computation is performed by rotating the base of the photon mechanism state and by developing the quantum state as a Heisenberg modality (the Heisenberg picture of quantum mechanics). [Definition]

[0004] OAM=orbital angular momentum

[0005] SAM = Spin Angular Momentum

[0006] STATES = quantum states associated with photons

[0007] Modes are the decomposition of an electromagnetic field into the fundamental modes of a waveguide; TE modes are transverse electric modes, and TM modes are transverse magnetic modes.

[0008] A vortex is a mode accompanied by non-zero angular momentum, i.e., non-zero or non-one.

[0009] 1(+1) and 1(-1): These are the OAM modes used herein. These are literally modes with orbital angular momentum of 1=+1 and 1=-1. In its broadest sense, OAM1(+1) refers to the OAM mode with 1=+1, and OAM1(-1) refers to the OAM mode with 1=-1. These modes are indicated using OAM(1=+1) and OAM(1=-1), or alternatively, OAM +1 and OAM -1 This can also be shown using [the appropriate method].

[0010] The basis (BASIS) corresponds to a set of four photon states TE, TM, 1(+1), and 1(-1), which form the orientation in four-dimensional Hilbert space. The term "basis" is a mathematical term for the fundamental vectors that constitute the space. For example, a vector v in a plane is given by v = a*vx + b*vy, where vx and vy are the basis vectors associated with the x and y axes. They are called unit vectors when chosen as unit lengths. Thus, a vector in space is given by the sum of projections on the basis vectors. In this case, a vector in Hilbert space, with at least four dimensions, represents the quantum state of a photon in the output section of a computing module, which can be one of at least four states, or a preferred combination thereof.

[0011] Active trench / active trench equivalent: These terms are sometimes used as synonyms for general active phase-shifting means, which are equivalent to adjustable trenches (thermal or otherwise). They are achieved by driving controlled changes in a given region, such as changes in the geometry of a waveguide.

[0012] Active phase variation means: These are means that can be controlled, and the active trench described above is one example; in a preferred mode of operation, they may include thermo-optic devices controlled to heat a given region of a waveguide (4), which is preferably single-mode. These are means that perform a function when activated by either control or piezoelectricity.

[0013] Passive phase variation means: These are non-operational means, and for example, they have a fixed shape. Therefore, they always perform the same function regardless of activation. [Background technology]

[0014] Prior to this invention, the applicant’s earlier Italian Patent Application No. 102022000022368, which claims priority and is incorporated herein by reference, introduces the concept of performing quantum computation in a four-dimensional Hilbert space using a four-mode cudit and implemented by a selective waveguide compatible with four fundamental photon modes OAM(1=+1), OAM(1=-1), TM, and TE. Discontinuities in trenches along the waveguide allow for mode tuning.

[0015] Even if this system were to function perfectly, it would be connected to waveguides and their trenches. Since these are not readily available on the market as standard components, their manufacture is expensive. [Prior art documents] [Patent Documents]

[0016] [Patent Document 1] Italian Patent Application No. 102022000022368 Specification [Overview of the Initiative] [Problems that the invention aims to solve]

[0017] Therefore, the applicant of the present application has deepened the research with the perspective of overcoming this problem.

[0018] Another object of the present invention is to provide a multi-purpose circuit that can perform main quantum operations while minimizing photon loss as much as possible. Photon loss causes calculation errors.

[0019] Another object of the present invention is to use standard manufacturing waveguides and techniques widely used in the assembly of standard circuits, such as those for manufacturing silicon and silicon oxide components, which are parts used daily to significantly reduce costs.

Means for Solving the Problems

[0020] These objects are solved by a quantum computing module and method according to the appended claims.

[0021] Other features and advantages of the present invention will be best understood from the following detailed description of the preferred embodiments, by reference to the accompanying drawings which are given by way of illustration and not limitation.

Brief Description of the Drawings

[0022] [Figure 1] It is a diagram schematically showing a quantum computing module according to the present invention. [Figure 2] It is a diagram showing the conversion mapping from one mode to another based on the applied phase change (π / 2 or its multiple). [Figure 3] It is a diagram showing one of two examples of passive phase change means. [Figure 4] It is a diagram showing the other of two examples of passive phase change means. [Figure 5] It is a diagram showing OAM, TE, and TM modes with corresponding polarizations.

Embodiments of the Invention

[0023] Figure 1 shows the quantum computing module according to the present invention, with the entire module shown by reference numeral 1.

[0024] Module 1 is: -Hereafter referred to as a polarization beam splitter, the input unit 2 is shown as an example, not an limitation, and is equipped with a mode separator 15, configured to split an incident photon beam characterized by a transverse electric field TE and a transverse magnetic field TM into two sub-beams, one characterized by the transverse magnetic field TM and the other by the transverse electric field TE, and each of these two modes TM and TE is equipped with polarizations Ex and Ey, the input unit 2 and; - Downstream of the polarizing beam splitter (PBS) 15, there are two parallel paths 5 and 10, one for each sub-beam, each of which is as follows: For example, a single-mode waveguide 4 made of silicon on an insulator (also known as SOI, SI on an insulator), preferably of a standard size of 220 × 480 nanometers for a wavelength of 1550 nanometers, where losses are minimal in these waveguides; Hereafter, a mode processing means 20 is provided with a Mach-Zehnder interferometer, also referred to as MZI1 and MZI2, which is shown as an example rather than an limitation, and includes a first phase shifting means 22, indicated as PS1 and PS2 in the figure; Hereafter also referred to as C, the output section 3 includes a mode combination means 25 arranged to join two parallel paths 5 and 10. Characterized by two parallel paths 5 and 10, Includes, Here, the two parallel paths 5 and 10 are distinguished from each other by the fact that at least the first path 5 includes first and second mode-converting means 30, 32 (also called R), shown as examples and not limited to, by their respective rotating means of polarization (and thus associated modes), one located upstream of the respective processing means 20 and the other downstream.

[0025] Optionally, the two paths 5 and 10 can also be distinguished by the second path including a second phase-changing means 70 inserted between the mode processing means 20 and the second conversion means 32. However, this is not necessarily required if the first phase-changing means has sufficient performance, for example, to enable changes from -180° to +180°.

[0026] The mode processing means 20 of each pass preferably includes an active power splitting device 50 at the exits of the MZI1 and MZI2 interferometers, respectively; this splitter is preferably driven by the phase induced by the phase shifting means 22 and performs at least the following operations: - All radiation is sent directly to the output unit of the processing unit 20; - Sending a first portion of the radiation to the output unit of the processing unit 20, and a second portion to the respective reflecting units 51 and 52 (e.g., mirrors) that reflect it back, and eliminating the second portion because this drive configuration corresponds to a closed port for its photon state; - All radiation is eliminated by sending it to the respective reflection means (51, 52) described above, which corresponds to a complete closure for any photon state.

[0027] The splitter is, for example, a 50 / 50 type. That is, it divides the incident radiation into a first and second part, each being 50% of the incident radiation.

[0028] The splitting device 50 can be of a known type.

[0029] Generally, couplers combine beams, while dividers divide them; often, the same device can be used for both functions simply by reversing its orientation. In the literature, they are known as coupling / dividing devices.

[0030] The output patterns achieved by driving the two dividers / couplers 50 of MZI1 and MZI2 are as follows, for example: 1. The output pattern allows output only from MZI1 and blocks output from MZI2; 2. The output pattern allows output only from MZI2 and blocks output from MZI1; 3. The output pattern allows for output with equal power from both MZI1 and MZI2; 4. The output pattern allows for output with equalized power from MZI1 and MZI2 (e.g., equalization of the right / left channels in a stereo music player).

[0031] The input section 2 preferably includes at least one input register having at least one qubit of four states or a qubit register of four levels for a photon quantum, where these four quantum states correspond to the following four photon propagation modes and their superpositions: A = OAM 1(-1) B = TM C=OAM 1(+1) D=TE

[0032] The output unit 3 also preferably includes at least one register having at least one quid of four states, or a qubit register of four levels, characterized by the quantum states and their superpositions. These four states are obtained by combining two modes processed by the processing means. The processing means imposes a phase shift of π / 2 or a multiple thereof on the modes corresponding to those states (see phase shift mapping in Figure 2), thereby combining them again in the combining means C, resulting in the following states: D= Ex=TE B =Ey =TM A = Ex + iEy = 1(-1) C = Ex - iEy = 1(+1) Here, i = -1 is the square root.

[0033] We have confirmed that "states" correspond to quantum states of photons or their equivalents related to electromagnetic fields, while "modes" relate to waveguides. In this context, these terms have the same meaning, and waveguide modes are used to generate photon quantum states. In fact, these states are photon states, and as is known from Maxwell's equations and from Majorana-Wigner quantization to semiclassical systems, photon states correspond to modes carried by waveguides and the entire path.

[0034] In a single-mode waveguide, it has also been confirmed that OAM and polarization are coupled behind the mode separator 15. TE and TM modes exist, as well as their superpositions TE+iTM and TE-iTM, where i is an imaginary number, indicating a phase shift of plus or minus 90°, or half the waveguide. The TE mode is associated with horizontal polarization, and the TM mode is associated with vertical polarization.

[0035] The mode separator 15 separates the TE and TM modes, then decomposes the vortex into TE and TM, while preserving the half-wavelength phase shift. Thus, the vortex, i.e., the state with 1=+1 or 1=-1 in this case, is converted into a coupled horizontal mode by the phase shift. This is a very special case that allows the use of a single-mode waveguide to transfer the TE mode to the TE mode with the smallest possible loss. For this reason, the TM mode is temporarily converted to TE, and once the processing required by calculation has been done, they are converted back to TM and reassembled to generate either a vortex, or a superposition of the TE and TM modes, as well as their vortices.

[0036] It has been confirmed that a multimode waveguide with, for example, four modes can be present before the mode separator 15 and after the output section 3. The input section 2 may correspond to the output section of the preceding circuit, and / or the output section 3 may correspond to the input section of the subsequent circuit.

[0037] Output unit 3 essentially recombines the two subspace TE and TM to construct the four modes used in the calculation. Behind output unit 3, various possibilities can be envisioned. For example, a final module 60 with the function of a mode (and their superposition) tuner can be inserted, including active phase shifting means 40, such as an adjustable trench (either thermal or otherwise), which will be referred to hereafter as an active trench. The active trench shifts the results in beamcoupler module C in the desired manner, realizing other state transformations such as Pauli z and controlled z (CZ). The final module 60 is a multimode waveguide and may comprise either two other consecutive calculation modules, or a separator directed toward a detector, or another single module.

[0038] The phase change means can be of various types, such as active or static.

[0039] An active example includes a controlled thermo-optical device for locally heating a single-mode waveguide 4.

[0040] Among the static examples, we have a trench 40b directly etched into the single-mode waveguide 4 shown in Figure 4. This phase shifter will hereafter be collectively referred to as the waveguide trench 40b. It modifies each of the photon states by transforming the states according to a predetermined scheme, as in Figure 2. Note that depending on how the programming language is defined, we can also choose a different sequence involving the transformation of the ABCD states.

[0041] Figure 4 shows a support wafer 12 made of SiO2. A single-mode waveguide 4 made of Si, with a rectangular cross-sectional area H×W=220×480 nanometers for a wavelength of 1550 nanometers, is placed on the wafer 12.

[0042] The trench is realized as a rectangular cross-sectional recess starting from the edge of the waveguide. Trench 40b has an etching depth of 70 nm; width w = ff * W with coefficient ff = 0.25. The waveguide width W is selected from 1 or 1.1 μm depending on the frequency.

[0043] The same effect can be achieved by depositing material onto the same single-mode waveguide 4, or by appropriate active changes in the waveguide's shape, for example, by a thermal actuator.

[0044] A second static example is the trench 40a in Figure 3, which is the same single-mode waveguide 4 described above, but with two slits of different depths.

[0045] The trenches 40a and 40b essentially function as π / 2 polarization rotors, as shown in Figure 1, and thereafter these two trenches can be used as examples of polarization rotating means 30 and 32.

[0046] Figure 2 shows the transformation mapping between states in a four-dimensional Hilbert space, based on the phase change applied to module 1 and the shift imposed by the phase change means of the final module 60.

[0047] 1) The first advantage of this circuit is its versatility in performing major quantum operations with minimal potential loss. This is because it is small, has no junctions that could cause photon loss, and therefore does not introduce computational errors. In contrast to those already existing in the literature and previous patents, this small circuit makes it possible to do almost everything required by on-site quantum computing techniques. Others, on the other hand, require connecting to many circuits to move photons through longer and more complex paths, resulting in photon loss, and thus loss of information, and the introduction of errors during computation.

[0048] 2) The second advantage is that widely used, standard manufacturing waveguides and techniques are used for assembling the silicon and silicon oxide components, which are parts used in everyday applications. Therefore, costs are significantly reduced.

[0049] 3) Preferably, the calculations are performed on a single-mode silicon-on-insulator (SI on an insulator, also known as SOI) waveguide of standard dimensions, e.g., 220 × 480, for a wavelength of, for example, 1550 nanometers. Losses are minimized in these waveguides. Other materials and other wavelengths have their own configurations, as given by known literature.

[0050] 4) Each of the four quantum states used is decomposed into TE and TM modes, including superposition, at the module's input by a polarization beam splitter (or equivalent device).

[0051] 5) If the waveguide is significantly curved, the TM modes, which are easily dispersed in the circuit, are converted to TE modes by mode rotation means R, such as a suitable optical element, which rotates the polarization and therefore the associated quantum state. The TE modes, which propagate with very little loss, are processed for quantum computation and converted back to TM modes, and then superimposed onto the TE modes.

[0052] 6) The processing of photon quantum states is performed via MZI tuning means, which are activated by a programmer in a predetermined manner to perform various transformations of quantum states, thereby enabling various operations via so-called quantum logic gates and quantum state registers. Activation includes, for example, local heating of the waveguide, and performing phase changes using active phase-shifting means. This mode operates like a trench in the waveguide, but in this case it is active, i.e., maneuverable, so its shape can be freely changed. To achieve this phase shift, the present invention also includes other active phase-shifting means known in the literature, such as piezoelectric means or certain materials inserted into the waveguide, which are modified via an electromagnetic field or the like.

[0053] 7) Processing of quantum states, particularly phase changes, can also be performed by passive optical elements.

[0054] 8) A third advantage is that by using these standard manufacturing techniques, errors in the manufacturing process itself are reduced, ensuring not only high quality but also the aforementioned low cost.

[0055] 9) The fourth advantage is the versatility of this computing module, which utilizes four quantum states related to photons. It not only simplifies calculations, as in the circuit in the applicant's previous patent application, but also performs major quantum operations without the use of other circuits.

[0056] 10) One of these modules can be used as a distribution node (photon distributor) for computation to other modules or other parts of the circuit.

[0057] 11) Each of these modules can be connected to another component of the circuit to act as a photon focuser.

[0058] 12) Each module can also be used as a computing module, a distributor to one or more modules, or a general circuit element.

[0059] 13) A single module, or a network of such modules, can construct quantum and classical computing circuits of universal characteristics, which can be coupled to or to other quantum or classical computers or circuits.

[0060] 14) The computer modules presented herein, or more general universal computing circuits, can be used by processing the quantum states of a single photon (i.e., a single photon), or multiple photons or packets of photons (referred to as "continuations" or "continuation modes") that are distributed to the circuit itself.

[0061] 15) Computation can be enhanced by using pairs or groups of entangled photons.

[0062] 16) Quantum computation errors can also be reduced by using entangled photons. One example is to use "herald" photons, i.e., one or more entangled photons, to signal the arrival of one or more of them at a point in the photon circuit, which is determined during the computation phase.

[0063] 17) Further developments of the circuit can also utilize increasing photon frequencies, generally obtained by beating waves to achieve frequencies close to the fundamental frequency, which also propagates within the circuit. This results in multiple states and increases the size of the cuid, as illustrated in the example of additional "coloring" configurations.

[0064] 18) The calculation is formally performed via a rotation of the base of the four-dimensional Hilbert space of the quantum states that each photon can assume. This rotation can be arbitrary and is given by superimposing the four states.

[0065] It has been confirmed that general photon quantum circuits can be realized by connecting such computing modules to each other to generate a network, i.e., a circuit for quantum and classical universal computing, and by connecting them to other optical elements that can distribute and / or process the quantum states of photons used in the computation.

[0066] By appropriately implementing each module, the main quantum and classical logic circuits are generated, and the processing means PS1 and PS2 are: Identity All three Pauli x, y, x Hadamard Controlled Not(CNOT) Controlled Z (CZ) Toffoil Swap Phase shift gates: e.g., Pi / 8 and Phase(s,p) And NOT Ure And combinations thereof.

[0067] This is achieved by linking polarization (SAM) to vortex (OAM), which is a different method from the one proposed in the applicant's previous patent application, which uses separated SAM and OAM polarization.

[0068] As in previous patents, while using vortices (OAM states) and four states, the new circuit uses the electromagnetic field in the fundamental modes of the waveguide at the input of module 1 of this quantum circuit: TE mode - i.e., transverse electric, and TM mode - i.e., transverse magnetic, The technique employs a decomposition method, and these are stored by appropriate time shifts within the circuit. In this way, information about the input state, i.e., one of the four states associated with the photon (and their combinations according to the laws of quantum mechanics), is stored and processed in order to carry out all the transformations associated with quantum computation.

[0069] The four fundamental modes in this case are TE polarized along the x-axis, and TM identified using the corresponding TM magnetic mode with perpendicular polarization. The OAM modes are given by phase-shifted superpositions 1=+1=TE+iTM and 1=-1=TE-iTM, where i is an imaginary number (i.e., the square root of -1), and correspond to a phase shift of one wavelength plus or minus in the superposition of the TE state with the TM state.

[0070] The use of waveguides in the above example, i.e., implemented using SOI technology, and measuring 220 × 480 nanometers for a wavelength of 1550 nanometers, offers the advantage of a simpler and less expensive procedure; however, the waveguide is not rectangular like the main example in the applicant's previous patent application, and TE and TM modes cannot easily create and propagate OAM modes. The solution proposed by the present invention is to perform calculations in Hilbert 4D space by decomposing the computer-generated space into subspaces, obtained by decomposing the initial 4D field into two superimposed 2D subspaces, one representing TE modes and the other representing TM modes. This decomposition is performed by a polarizing beam splitter (PBS). Due to the shape of the waveguide, if the photon path in the waveguide is curved, TM modes may exhibit some loss or alteration of the photon state. To avoid these losses, the present invention performs a transformation in the portion of the circuit corresponding to the subspace TM in the TE mode (corresponding to one of the two paths 5, 10), performs local calculations according to the superimposed subspace TE, converts the TE mode back to the TM mode, and finally superimposes it onto the TE mode in the "combination" zone. This is defined as the subbeam combination means C, which regenerates the required OAM or TE / TM mode, or a superposition thereof. Naturally, the other TE modes (the other paths 5, 10) undergo other transformations according to the calculations.

[0071] To achieve this, polarization, also known as SAM from now on, is no longer independent of OAM, and the two are coupled.

[0072] Module 1 is generated in a 1-to-2 cascade structure, which connects modules that already act as CCNOT gates in themselves (and drives the final module 60 in a preferred manner to obtain a circuit configuration for controlled Z and Pauli z, which rotates the eigenstates by 180°), and is based on two Mach-Zehnder interferometers, one for each polarization state (Ex and Ey correspond to TE and TM modes). They include very high-precision phase-shifting means 22 for the TE and TM modes, each with variable phases (φ1 and φ2), and are connected in parallel behind a polarization beam splitter (PBS) used as an input.

[0073] Optionally, the output sections of two parallel MZIs can be phase-adjusted by acting on the phase-shifting means 70 (φm) in the Ex path behind the x-polarization interferometer of MZI1.

[0074] In fact, if the two phase shifting means 22 associated with the MZI1 and MZI2 interferometers are sufficiently accurate, the phase shifting means 70 is unnecessary and can be omitted. Field tests have already shown that it is possible to achieve a very accurate phase shift from -180° to +180° using only the phase shifting means 22.

[0075] The output field and its calculation are determined by the phase adjusters (φ1, φ2, and φm).

[0076] The TM field is rotated by 90° and sent to its Mach-Zehnder interferometer (MZI2) for calculation, reducing the loss of TM modes in the transport; then the two fields Ex and Ey are combined again. At this point, the combined field can be sent, for example, to the active phase shifter 40 in the final module 60.

[0077] The final module includes active phase shifting means, such as active trenches, for generating Pauli z and CZ, which shift the phase from 0° to 360°, thereby achieving a desired shift of photon modes (A, B, C, D).

[0078] Figures 3 and 4 show two examples 40a and 40b of the TE-TM rotor 30 configured as a fixed trench, as is well known in the literature.

[0079] The circuit can operate in continuous mode, that is, with many photons or laser pulses.

[0080] To avoid signal and mode loss, or to avoid the possibility of unintended photon state mixing and subsequent coherence loss of the carried quantum state, which is known to be prominent in curved multimode rectangular waveguides, the present invention employs a single-mode (monomode) waveguide with a known shape and structure in the literature, of the type commonly used in SOI wafers having a Si thickness of 220 nm and a width of 480 nm for a wavelength of 1550 nm, as mentioned in the applicant's previous patent application. To obtain equivalent results, the size and shape of the waveguide will obviously vary depending on the material and wavelength used, so the general condition is that the waveguide is single-mode.

[0081] In this configuration, and at this frequency of 1550 nanometers, the waveguide can be bent without significant loss, and the field modes Ex and Ey can be compressed, which in this case correspond to the TE and TM modes, and the corresponding TM mode, respectively.

[0082] In the transport of TE and TM modes, in order to avoid the effects of birefringence in a 220 × 480 nm waveguide, it is preferable to convert the TM mode to the TE mode via a mode rotation means (rotating polarization) located behind the polarizing beam splitter (PBS) used as the input port, especially when the waveguide is curved or in coupled mode. After the necessary conversion and processing of the photon states, the two passes are sent to the beam coupler C and the active mode tuner 60.

[0083] Throughout the circuit, SAM and OAM are coupled, resulting in OAM modes being given as a superposition of SAM modes. The mode rotation means R converts pure SAM and mixed SAM / OAM states into SAM or OAM modes. This is an example of the classical correlation (known as classical entanglement) between SAM and OAM.

[0084] Therefore, OAM mode depends on SAM as follows: -TE+iTM gives 1=-1 (clockwise). This means that they are realized using the polarization state H+iv, i.e., the clockwise elliptic polarization Ex+iEy. -TE-iTM gives 1=+1 (counterclockwise). This means they are realized using polarization state H-iv, i.e., counterclockwise rotating elliptical polarization, and Ex-iEy will be decomposed by PBS.

[0085] Any OAM mode is realized using the superposition of TE=Ex and TM=Ey, which, upon passing through PBS, decomposes into two fundamental components TE and TM, while maintaining the phase shifts that characterize the positive and negative OAM modes, and preserving the OAM mode in the combination of two passes without a phase adjuster such as a trench waveguide, either when no phase shift is added or when it is reversed when a π phase is added; the other two cases involving π / 2 and 3 / 2π (=-π / 2) result in a superposition of the TM and TE mode phases, which means that oblique polarization occurs.

[0086] PBS acts as a projector of Hilbert's 4D space into a 2D subspace (corresponding to each of the Mach-Zehnder MZI interferometers). The routes must be synchronized.

[0087] Quantum computation is represented in the table of states and transformations.

[0088] As known from international literature, when an OAM mode passes through a polarized beam splitter called a PBS, each OAM mode is decomposed into two polarization states (SAMs), TE and TM.

[0089] As an example, if the OAM mode passes through PBS, it is determined that the TE=Ex mode has passed through each phase shift means 22 without any change, and this is associated with the |0> state level (or symbol 0 in the truth table). This means that means 22 has been deactivated, and therefore the original phase shifts that the OAM modes with ±π / 2 (i.e., ±i) TM modes are generated from the 1=±1 OAM mode are preserved.

[0090] After passing through PBS15, the TE=Ex mode reaches the input of MZI1 and remains invariant in each phase shifting means 22 (preserving the TE), thus preserving the initial phase shift that generates the OAM with ±π / 2 (i.e., ±i) TM modes from the 1=±1 OAM mode. Meanwhile, the other mode TM=Ey passes through each polarization rotation means 30 and is converted to the TE mode for the input to MZI2. This is because this mode in the single-mode waveguide propagates without difficulty and can be appropriately tuned with almost no loss, then converted back to the TM mode by the polarization rotation means 32, and combined again in the mode combination means 25, thus generating the desired output state by quantum computation prepared by MZI1 and MZI2 in a way that minimizes loss.

[0091] To implement quantum computation in Module 1, phase shift means 22, and optionally optional phase shift means PS-EX70, are used in MZI1 and MZI2. For example, when passing through phase shift means 22, 70, and 40, if they are deactivated, Module 1 returns to output 3, which is the same input unit that receives input 2. This corresponds to the identity operator. This operation is associated with the state label |0> (or the symbol 0 in the truth table).

[0092] If superimposed TE+TM modes corresponding to 45° linear polarization exist in input 2, they are decomposed by PBS15 into a TE mode advancing toward MZI1 and a TM mode advancing toward MZI2. When the phase shifting means 22 in MZI2 is activated to prepare the output state, the TM mode has a phase shift of +π / 2 and, after the rotation means 32, becomes a +iTM mode with an invariant component coming out of MZI1 in the coupler 3, which becomes the 1=-1 mode. The 1=-1 mode is TE+iTM with clockwise rotating elliptical polarization. The 1=+1 mode is obtained from this result by shifting it by 180° using the active trench 40.

[0093] In quantum computing, there are combinations of modes of transverse electric field TE and transverse magnetic field TM in a waveguide. These include the invariant phase superposition in the case of the 1=-1 mode and TM mode, which result in the superposition mode TE+iTM±TM. This corresponds to the vortex mode with polarization rotation, the superposition of 1=-1, and TM. The same thing happens in the 1=+1 mode, which results in a superposition between 1=1 and TE.

[0094] If the input direction of the field is set to the V direction for PBS, then |0> means that the polarizations do not match, which is the case when V=TM mode.

[0095] If TM does not change, the state remains |0>.

[0096] When TE corresponds to |1>, a vortex mode with vortex 1=-1 is generated after passing through the phase change means 22 and 70 in Figure 1.

[0097] Polarization here is a superposition of the i-phase-shifted modes of TE and TM, polarized clockwise or counterclockwise by the sign of TM.

[0098] 1=-1 and 1=+1 generate superpositions of (TE+-TM)+iTM and (TM+-TE)+iTE, respectively. These are superpositions of linearly polarized and circularly polarized light, and are elliptically polarized.

[0099] Mode change: Obtain a superposition of modes, from linearly polarized to elliptically polarized, or from circularly polarized (a special case of elliptical polarization) (left or right) to elliptically polarized.

[0100] The phase shifting means 70 functions as a regulator or compensator module between MZI1 and MZI2.

[0101] Circuit: Rectangular waveguide of a silicon chip on an insulator on a 220nm thick silicon layer, which is a standard SOI wafer. The frequency is 1550nm. The single-mode TE waveguide has H=220nm × W=480nm.

[0102] Waveguides: Like standard SOI wafers, they consist of silicon (n=3.48@1550nm) with an SiO2 coating (n=1.445@1550nm), and a silicon thickness of 220nm.

[0103] The flat structure of the waveguide means that the OAM is given by the superposition of the Ex and Ey fields with a phase difference of ±π / 2. With the SAM-OAM coupling, the OAM beam starts as a coupling of linearly polarized light and becomes elliptically polarized. OAM 1=+1 is given by Ex-iEy. OAM 1 = -1 is given by Ex + iEy.

[0104] In a preferred example, mode adjustment is performed via an active mode adjuster, the phase shifting means 22 being one example (e.g., a waveguide heater to achieve polarization rotation; a real example is a controlled thermo-optic device). The phase shifting means 70 mentioned above is used to adjust two MZIs when something is out of phase or when it is desired to reverse the modes.

[0105] Essentially, this example employs a 220×480nm single-mode TE waveguide instead of using a trench waveguide to reduce losses, and adjusts the polarization using active changes to the waveguide, such as thermal activation and phase shifting.

[0106] Regarding the realization of (polarization) rotation means using CMOS-compatible SOI technology and based on the breaking of symmetry in the waveguide cross-section, they can be created using a 220 nm thick waveguide and active components that modify the waveguide properties. In this case, it is necessary to achieve a controlled rotation of the polarization vector to superimpose the TE and TM modes using the respective phase delays acting as mode conversions.

[0107] Module 1 is a universal computing module, which can be implemented in CCNOT mode (Tofoli), and the CNOT subunit includes an Hadamard gate. Each Module 1, accompanied by PS1 and PS2, can become a unit of a quantum processor. The active trench 40 in Module 60 prior to PBS, when a 180° rotation phase is added, can obtain controlled Z and Pauli z in the combined zone output section C, where 1=+1 becomes 1=-1 and vice versa in the 4D configuration space, and TE becomes TM and vice versa. Module 60 preferentially rotates the state from 0° to 360°.

[0108] Each CCNOT module 1 generates a mode that can be split into two other consecutive CCNOT modules 1, which can accept or reject photon states (when both MZIs are out of phase), or the field can be modulated using MZI1 and MZI2 with phase-shifting means 70 (PS-EX).

[0109] The final module 60, with its active trench 40, implements operations through the z-axis as a parity operator when the mode is rotated 180°. This language represents changes based on Hilbert space, and the photon state remains invariant, as in the Heisenberg picture of quantum mechanics.

[0110] Regarding operations through the z-axis, we have confirmed that they involve abstract rotations in three-dimensional space and do not require the direct use of the circuit's outline. Therefore, z is not an axis of the circuit. In fact, Pauli matrices represent rotations of vectors in 3D space. In this case, the spatial rotations of Pauli x and Pauli y are obtained using the submatrices of CNOT and, of course, CCNOT, while Pauli z and CZ require another degree of freedom associated with transformations in Hilbert space. This corresponds in this case to a shift of two positions in Hilbert space, or rather a double rotation of the basis; they are called parity operators because they resemble reflections in a mirror.

[0111] At the CCNOT end of the combined zone C in the two output fields, another rotation π can be obtained by a phase shift mapping in the constituent space in Figure 2, in addition to a phase rotation of π / 2, thereby obtaining a basis inversion. Thereafter, OAM 1=+1 is transformed to 1=-1 (and vice versa), which is the parity operation in the pseudovector, and the TE mode is transformed to the TM mode (and vice versa) according to the parity of the OAM state. This is a double rotation in the Hilbert space used for the calculations of the present invention.

[0112] Using 10 consecutive levels of branching, a readout output (detection of photon movement in the circuit, along with 4 states) is obtained, which is equivalent to a 1M pass of 10 quantum outputs. Using 20 levels results in a 1 terapass output.

[0113] The following table shows the truth tables for the Toffoli and CNOT gates obtained using the modules described above.

[0114] [Table 1]

[0115] The Hadamard gate is realized by switching two MZIs to the ON position and using the PS-EX phase change means 70.

[0116] For Pauli z and CZ gates, the phase change means 40 of the final module 60 must be activated to invert the last off-diagonal CNOT elements, Pauli z and Hadamard-Pauli x. First, the identity operation using MZI is set up, the phase is initially set using the phase change means 22, and then, if necessary, the phase change means 70 is used to activate the phase change means 40 when it is necessary to perform CZ or Pauli z.

[0117] From this point onward, the additional composition referred to as "coloring":

[0118] The OAM and MUX (multiplexing) frequencies in the OAM TE induction mode.

[0119] Both configurations can be extended by using a light source of entangled photon pairs and a technique for identifying photon states at frequency as well.

[0120] Thus, we obtain a quo-quad, i.e., (2^2)*(2^2)=16 states. This means a 16-state quid for each frequency in the circuit. By using both frequency and OAM multiplexing, we can utilize all possibilities of physical systems, such as optical and photon circuits, with frequency diversity, by multiplying the number of frequency bands obtained by beating both sides of the base frequency, thereby including the base frequency f identified by the fn parameter.

[0121] In this case, it is possible to adjust the frequency using these circuits that operate in TE mode, and to obtain a frequency close to the central frequency using optical-acoustic or similar techniques, and to increase the cudit with 400 states using entangled photon pairs given by an additional dimension added by the frequency multiplication fn, using two sidebands from a base frequency centered at 1550 nm. In this case, there are fn=5 and one quad for each frequency, the frequency and OAM are independent, and the superposition is given by a cudit with (fn*4)^2=(20)^2=400 states.

[0122] The conversion from optical frequency to OAM has been shown to be feasible in a controlled manner; it can reduce the complexity of the circuit by a function of log2(400) = 8.65. Its advantage is (log2(400))^2 = 75, obtained using five frequencies and four cudet states of entangled photon pairs. In the extreme case of fn=9 with four frequency beats in sequence, the maximum value that can be obtained is 362 = 1296 and log2(1296) = 10.34, and its advantage is (log2(1296))^2 = 106.91. This means that it has an advantage of the order of more than two orders of magnitude, while the complexity of the circuit increases.

[0123] A quantum computer equipped with the module according to the present invention is an optical quantum computer that uses cuids based on photon states carrying orbital angular momentum (OAM). Two different configurations exist using different configuration techniques at a given fixed frequency. Further extensions involve configuring OAM-based cuids and frequency multiplexing techniques that use ground frequency beating to obtain different photon states, where the number of frequency beats is fn times the number of cuid-based OAM states. The size of the computation depends on the dimensional correlation obtained through different paths and the circuit configuration in which one, two, or more photons are captured during the computation. This has been extensively investigated in the literature (see, for example, the 2014 literature by Krenn, Zeilinger et al.), where it is explained that the size of an entangled quantum state may increase with the number of particles, or in this case, with the number of relevant dimensions. This specification describes the properties of two photons in a 100-dimensional entangled state. These dimensions are represented by the size of the cuid state and by the variety of paths obtained in the computation.

[0124] This invention simplifies quantum circuit structures using quids. As is well known from the literature (see Wang et al., 2020), the use of quids reduces the complexity of quantum circuits; this is essential in the design and construction of optical circuits, where complexity is known to increase stepwise and exponentially. Indeed, speed, resource conservation, and deployment on physical platforms are reduced using multi-state qubits, also known as quids. Quids are characterized by having a Hilbert space larger than that of a qubit. By definition, a quid is a quantum version of the d-ary system (or d-ary sequence), and in a generalization of binary data sequences, a node has d child nodes instead of the 2 used in the data structure of a priority queue constructed from an array of d objects. A d-dimensional quid is represented as a quantum state of vectors embedded in a d-dimensional Hilbert space HD. The space is a set of d-dimensional orthogonal basis vectors {|q1>, |q2>, ···|qd>}, |Ψ>=α1|q1>+α2|q2>+···+αd|qd> is defined by, The standardization conditions are as follows:

[0125]

number

[0126] What is a Cupid? A qudit is an extension of the d=2 qubit concept to d dimensions. It represents a d-dimensional quantum system. Qutrit, ququart, and ququad represent 3-dimensional and 4-dimensional quantum states, respectively.

[0127] One of the main advantages of the cudit model over the qubit model is that it significantly reduces the number of cudits required to cover the state space used for computation. For example, at least tn_1 = log_2 N cudits are needed to represent an N-dimensional system of qubits. On the other hand, using cudits requires only n_2 = log_d N cudits, which has the advantage of reducing the complexity factor given by k = n1 / n2 = log2(d). The cudit method has a scaling advantage of (log2d)^2 over the qubit case. For a cuquad with d=2 and four cudit states, the advantage is (log2(4)^2 = 2^2 = 4. The n-th order complexity at a circuit scale of 2^(2*n) is four times that of the qubit case.

[0128] For entangled particles, the scale dimension of the entangled Hilbert space is d = g * n, where g is the entanglement dimension and n is the number of related components. Considering two entangled quads (g=4) obtained using a pair of entangled photons (g=2), they have a tensor product of d = 16. In that case, the circuit scale complexity is 2^(4*n).

[0129] When an additional dimension is added by multiplying the frequencies by fn, fn=5 and there is one quark for each frequency, the frequencies and OAM are independent, and the superposition is (fn*4) 2 =(20) 2 =400 is given by the state of the Qudit.

[0130] The configuration described in the applicant's prior patent application can be realized on a silicon wafer using a non-standard, multimode rectangular waveguide of 1 micrometer in size, and the calculation is obtained by checking the polarization state of each photon; this means that in this configuration, polarization is considered to be a quantity independent of the possible states of the photon's orbital angular momentum.

[0131] The second configuration of this invention application is based on an optical circuit with a waveguide provided by a 220 nm thick SOI layer, which is currently the most common, inexpensive, and standard construction process. The latter configuration is easier to construct, but the price to pay is the same as that of a standard single-mode waveguide, where a rectangular waveguide with a height of 220 nm and a width of 480 nm is used, and the field associated with the photon must be coupled with both OAM and polarization. In fact, the waveguide is birefringent, and the TE and TM modes are characterized by different propagation. The field is reconstructed at a fusion junction where the TE and TM modes (x and y polarization states) are coupled in synchronization using the resulting elliptic polarization LH or RH, depending on the type of OAM state. The phase shifting means 70 is used to synchronize the outputs of the two MZIs, ultimately optimizing the calculations made by the CCNOT module and thereby reducing any errors (local means performed by the module, then other means, and so on). Complex quantum circuits may feature cascades or networks of these CCNOT-CZ Pauli-z modules.

[0132] An additional advantage is that, in this case, the loss in transporting the photon state is very low, even if the waveguide is curved and the TE and TM modes are phase-shifted; for this reason, propagation must be converted mainly to the TE mode, and TM (y-polarized) is obtained again from another polarization rotor using a phase tuner for synchronization. This results in four independent states, including the OAM state and the TE-TM mode based on the combination of the TE and TM modes.

[0133] These quantum computer configurations represent modular units of quantum and classical optical computers, which operate using information units of four states identified by four labels (A, B, C, D), also known as quantum computation as four states or quads.

[0134] In standard quantum computing, the four states of a quad (A, B, C, D), |q>|Q1, Q2, Q3, Q4>, identify an orthogonal basis in Hilbert four-dimensional space and represent four orthogonal modes or quantum states allowed in a single-mode rectangular waveguide.

[0135] The electromagnetic (EM) field states propagating within a waveguide are TM and TE (vertical and horizontal polarization), where 1 = +1 and 1 = -1. The TM mode is related to the TE mode. The TE and TM modes correspond with sufficient approximation to Bessel modes or Hermit-Gaussian modes, with zero OAM. Si on SOI material in a waveguide selected to drive photon propagation at 1550 nm satisfies sufficient conditions for having independent TE and TM modes in a waveguide filled with a uniform and non-uniform lossless anisotropic medium.

[0136] OAM mode is then prepared as follows:

[0137] SAM and OAM are strongly coupled with elliptic polarization: 1=+1 is ccw-rotation and has RH polarization. 1=-1 represents cw-rotation and has LH polarization.

[0138] This situation can be used when the field is resolved in a waveguide circuit, and the xy symmetry is no longer preserved for a wavelength of 1550 nm, as in a standard single-mode waveguide of Si on a 220 × 480 nm SOI.

[0139] In this case, there are four possible states because the polarization depends on the beam's OAM state.

[0140] From these, each quad is given by a superposition of four independent eigenvectors |qi> with different polarization states. |Ψ>=α|Q1>+β|Q2>+γ|Q3>+δ|Q4> (1) Under normal conditions |α|2 +|β| 2 +|γ| 2 +|δ| 2 =1 (2) More precisely, |Q1> = |1,0,0,0>, |Q2> = |0,1,0,0>, |Q3> = |0,0,1,0>, and |Q4> = |0,0,0,1>.

[0141] Notably, these two configurations can also operate in a continuous mode for quantum computing of continuous variables using the intensity of the EM field and the squeezed state associated with each of the four states present within the configuration, the numerical value of which is adapted to a continuous interval at the input section; thus, the construction of quantum computing of continuous variables coincides with an infinite-dimensional Hilbert space, where in these configurations, a continuous interval can be obtained via a classical entanglement mechanism between polarization (spin angular momentum SAM) and the orbital angular momentum (OAM) of photons. When SAM and OAM are strongly coupled and the polarization discriminates a given OAM present in the quantum computer, the input photons are taken from an entangled pair of photons and the other is used as a "herald" to reduce the perturbation in the calculation due to the noise of the detector, and thus are appropriately manipulated. Another approach is based on the squeezed photon state; the chip is connected to a squeezed light source (infrared laser pulse and microresonator), and then is input into a first port that encodes the pulse into a superposition of four states (A, B, C, D), and then CV - quantum computing is performed. On the other hand, it is possible to entangle two entangled pairs to have three orthogonal Stokes operators and the corresponding SAM / OAM states between the beam pairs in two separate modules 1 according to the present invention. As a first basic language, the eigenvector |q i > and the waveguide modes (A, B, C, D) can be mapped. The mapping of the bit shift rotates each eigenvector in the local quantum state of the photon from |q i > to |q i+1>This shifts periodically, for example, from A to B, to C, to D. Any superposition of quantum states will rotate equally in d=4 Hilbert space. [General Terminology Explanation]

[0142] For understanding the purpose of the present invention, the term “comprehending” and its derivatives, when used herein, are intended to be open-ended terms specifying the existence of a particular characteristic, element, component, group, integer, and / or stage, but not excluding the existence of other characteristics, elements, components, groups, integers, and / or stages that are not specified. The foregoing also applies to words with similar meanings, such as the terms “including,” “having,” and their derivatives. In addition, when the terms “part,” “section,” “part,” “member,” or “element” are used in the singular, they may have a dual meaning of one part or more parts. When used herein to describe the modes of operation described above, the following directional terms, “forward,” “backward,” “upward,” “downward,” “vertical,” “horizontal,” “down,” and “cross,” as well as any other similar directional terms, refer to the modes of operation described in the operating position. Finally, terms of degree, such as “substantially,” “about,” and “approximately,” when used herein, mean a reasonable amount of deviation from the modified term, such that the final result is not substantially altered.

[0143] While only selected embodiments have been chosen to illustrate the present invention, it will be apparent to those skilled in the art from this description that various modifications and variations can be made without exceeding the scope of the invention as defined in the appended claims. For example, the size, shape, position, or orientation of various components can be changed as needed and / or desired. Components shown as being directly connected to or in contact with each other may have intermediate structures positioned between them. The function of one element may be performed by two, and vice versa. The structure and function of one embodiment may be adopted in another embodiment. Not all advantages are required to be present simultaneously in one particular embodiment. Any feature that is novel compared to known art should preferably be considered as a separate description of other inventions by the applicant, including the structural and / or functional concepts incorporated by those features, either by itself or in combination with other features. Accordingly, the foregoing description of embodiments in the present invention is provided for illustrative purposes only and not for the purpose of limiting the invention as defined by the appended claims and their equivalents.

Claims

1. A quantum computing module, An input section (2) is provided with a mode separator (15) and is configured to separate the transverse electric mode TE and transverse magnetic mode TM of one quantum into two parallel paths. Parallel to each other and located downstream of the mode separator (15), the first and second paths (5, 10) are for the transverse magnetic mode TM and the transverse electric mode TE, respectively, and each path is A single-mode waveguide (4) and A mode processing means (20) including at least a first phase changing means (22), The first and second paths (5, 10), characterized by the following, Output unit (3) includes mode combination means (25) arranged to join the two parallel paths (5, 10), Equipped with, The two parallel paths (5, 10) are distinguished from each other, and at least the first path (5) is A first conversion means (30) is located upstream of the processing means (20) and converts the transverse magnetic mode TM into the transverse electric mode TE. A second conversion means (32) is located downstream of the processing means (20) and converts the transverse electric mode TE into the transverse magnetic mode TM. A module that includes this.

2. The module according to claim 1, characterized in that the mode processing means (20) includes at least one Mach-Zehnder MZI interferometer equipped with the first phase shifting means (22).

3. Each mode processing means (20) of each path comprises an active power splitting device (50), the splitter (50) is driven by a phase induced by the phase changing means (22), and performs at least the following operations: Send all the emitting radiation to the processing means (20), The first portion of the emitting radiation is sent to the processing means (20), and the second portion is sent to the respective reflection means (51, 52) to be eliminated, and By sending the radiation to each of the aforementioned reflecting means (51, 52), all radiation is eliminated. The module according to claim 1 or 2, characterized by performing the following.

4. The two dividers (50) are arranged in the following scheme, namely, a. A scheme that allows output from only one of the paths (MZI1) and blocks output from the other (MZI2), and vice versa. b. A scheme that allows for outputs with equal power from both processing means in the two paths (MZI1, MZI2), and c. A scheme that allows an output with power equalized in a desired manner by the processing means of the two passes (MZI1, MZI2), The module according to claim 3, characterized in that it is operated to produce a matched output accordingly.

5. The module according to claim 3 or 4, as dependent on claim 2, characterized in that the active power splitting device (50) is located at the output section from each of the interferators (MZI1, MZI2).

6. The module according to any one of claims 1 to 5, characterized in that each of the modules is provided with polarization, and the conversion means (R) is provided with mode polarization rotation means.

7. The module according to claim 6, characterized in that the rotating means comprises at least one optical element configured to rotate the polarization by π / 2 or a multiple thereof.

8. The module according to any one of claims 1 to 7, characterized in that the waveguide (4) has a rectangular cross-section with dimensions of 220 × 480 nanometers, preferably for a wavelength of 1550 nanometers.

9. The module according to any one of claims 1 to 8, characterized in that the first and second phase-changing means (22, 70) are active because they are driven, i.e., they have a driving effect, while the conversion means (30, 32) are passive phase-changing means, i.e., they have a fixed effect.

10. The active phase-changing means includes at least one of a waveguide thermal control means, a piezoelectric driving means, or an electromagnetic field driving means, and is used to insert a material into the waveguide or deposit it on the waveguide, and The passive phase shifting means preferably includes at least one trench etched into a waveguide having a rectangular cross-section and dimensions H × W = 220 × 480 nanometers. The module according to claim 9, characterized by the following:

11. The input unit (2) includes at least one input register, which comprises at least one quudit having at least four quantum states of a photon quantum, or a qubit register having at least four levels, wherein the four quantum states are as follows: A=OAM 1(-1) B = TM C=OAM 1 (+1) D = TE A module according to any one of claims 1 to 10, characterized in that it corresponds to the four photon propagation modes and their superposition.

12. The output unit (3) includes at least one register, which comprises at least one qudit or qubit register having at least four levels, characterized by the quantum states and their superposition, the four states being processed by the following processing means, i.e. D = Ex = TE B = Ey = TM A=Ex+iEy=1(-1) C=Ex+iEy=1(+1) The module according to claim 11, obtained by combining two modes processed by (here, the square root of i = -1).

13. The module according to any one of claims 1 to 12, characterized in that the output unit (3) preferably includes an active third phase changing means (40).

14. A method for processing quantum propagation modes in order to perform quantum computation, To separate the transverse magnetic mode TM and the transverse electric mode TE of photon radiation, The two separate modes are tracked along two parallel paths (5, 10) to the combination point. While moving along parallel paths (5, 10), the transverse magnetic mode TM is converted to a transverse electric mode TE, the transverse electric mode TE is processed, and after the processing and before the combination, it is converted back to a transverse magnetic mode TM. A method characterized by the following.

15. The method according to claim 14, characterized in that the process of converting the TM mode to TE includes at least one phase change.

16. The method according to claim 14 or 15, characterized in that the transformation includes at least one polarization rotation associated with the associated modes.

17. The method according to any one of claims 12 to 14, characterized in that the phase change of the TE mode is performed in a path parallel to the phase change of the TM mode converted to the TE.

18. The process includes processing both the TE mode in one of the parallel paths and the TM mode converted to the TE in the other of the parallel paths, wherein the process is the following operation: Phase change and, It is a division, and the following is a division: To send out all radiation, The first part of the radiation is sent out, and the second part is sent to the respective reflection means (51, 52) to be eliminated. All radiation is eliminated by sending it to the respective reflection means (51, 52). The phase change described above drives the division and The method according to any one of claims 14 to 17, characterized by comprising the above.

19. The method according to claim 18, characterized in that each mode processing means (20), comprising a phase changing means (22), a splitting means (50), and a reflection means (51, 52), is arranged in each of the two parallel paths.

20. The method according to any one of claims 14 to 19, characterized in that at least one single-mode waveguide (4) is used for each of the parallel paths (5, 10).

21. The method according to claim 20, characterized in that a waveguide with a rectangular cross-section of 220 × 480 nanometers for a wavelength of 1550 nanometers is used as a single-mode waveguide.

22. The method according to any one of claims 14 to 21, characterized in that the mode is processed by commanding it to pass through a trench in a waveguide, and / or by controlling local heating of the waveguide in the path.

23. The method according to any one of claims 14 to 22, characterized in that a single photon, multiple photons, entangled photons, or a continuous signal are input to the two paths.

24. The method according to any one of claims 14 to 23, characterized in that a qudit having at least four quantum states, or a qubit register having at least four levels, is used as an input register at the mode separation point.

Images