Quantum processor, quantum logic gate, method for modulating a quedit in a quantum processor, quedit, and register of a quantum processor

Abstract

The present invention relates to a quantum processor comprising at least one input register (5) of information units corresponding to at least one quantum placed as an input in at least one waveguide (10) having a longitudinal extension in a predetermined direction L, defining a path for at least one quantum. Each waveguide (10) is considered to be configured to support a "d" propagation mode of at least one quantum. Each waveguide (10) is considered to define a continuous optical path between the input and output, comprising at least one modulation shape modulator (24) having at least one modulator in a cross section perpendicular to the longitudinal direction L. Each waveguide (10) is considered to be arranged in optical continuity with the rest of the waveguide. The modulation shape modulator is configured to manipulate the propagation mode in at least one quantum, and the processor comprises at least one output register (15) located in the waveguide (10) at the output and configured to decode information corresponding to the manipulated mode. The present invention also relates to a four-state quedit containing quantum orbital angular momentum (OAM). The present invention also relates to an input register for a quantum processor and a method for modulating a quedit with a quantum processor.

Description

[Technical Field]

[0001] The present invention relates to a quantum processor, a quantum logic gate, a method for modulating a quantum information unit (qudit) in a quantum processor, a qudit, and a register of a quantum processor.

[0002] Quantum processors, methods, and units have been developed with a focus on realizing quantum devices such as quantum computers. However, other applications are not excluded, such as realizing classical computers that can consume less power and perform calculations faster than current bit-based computers.

[0003] "Quantum": In physics, a quantum (derived from the Latin word quantum, meaning quantity) is a distinct, indivisible elementary quantity of a specific scale. More specifically, in this specification, the term is also used synonymously with elementary particles associated with a force field (Wikipedia). A preferred example of a quantum in this invention is the photon.

[0004] A "quantum device" comprises at least one "quantum circuit," which is, for example, part of a quantum processor and is configured to perform quantum computation to generate output data.

[0005] The output data may be quantum (qubits), classical data (bits), extensions thereof such as qudits, or a combination thereof, and is generally in string form.

[0006] Quantum circuits also have input data, which preferably includes classical data such as bits, objects, events, symbols, or signals in a quantum regime, and are generally in string form.

[0007] Quantum computing is used to construct output data based on input data.

[0008] A quantum device may comprise a quantum computer, its network, a quantum data transmission device, a sensor, classical ones, and combinations thereof, or their networks.

[0009] The definition of a quantum device also includes a quantum encryption system, which may in some cases be combined with other "classical" devices and is capable of generating a sequence of encryption keys, or objects, events, or symbols.

[0010] The concept of "event" is understood as a general physical manifestation that occurs within the universe, or within any multiverse, or more abstractly in the metaverse.

[0011] Thus, for example, it can be understood how this device can be made into a quantum sensor, and in this case quantum computing is measurement.

[0012] Each such quantum device is realized on a related piece of hardware, which is referred to herein as a device supporting the hardware.

[0013] [[ID=2,0]]A quantum device may include a single quantum circuit or a collection thereof. A preferred example of the quantum device of the present invention is a photonic quantum device, i.e., one that operates by photons.

[0014] )]]A quantum processor is a processor that is suppressed or managed by laws related to quantum mechanics, which means any processor capable of performing at least one "quantum calculation". A quantum processor may comprise one or more quantum circuits. In the present invention, a preferred example of a quantum processor is a photonic quantum processor, i.e., one configured to perform calculations using photons.

[0015] Quantum computing is defined as "any process performed by a quantum circuit", such as any of the following operations based on the laws of quantum mechanics. That is, without further limitation, measuring, generating, and manipulating quantum states, sequences of symbols derived therefrom, and the like.

[0016] Quantum computing has always been based on a basic unit called a qubit, that is, a unit of quantum information represented by a superposition of two states.

[0017] Quantum computing can also be based on the extension of qubits. For example, a qudit is known. A qudit is a unit of quantum information represented by a superposition of multiple states, and the number of states is an integer greater than 2 (Wikipedia).

[0018] Furthermore, a qutrit is known. It is a unit of quantum information represented by a superposition of three states, and thus a three-state qudit.

[0019] For example, a qubit is the state of a subatomic particle such as a photon or an electron. Each particle can be in several different states simultaneously and with different probabilities due to the principle of superposition, so it can "surpass" the duality of the classical binary 0 / 1 code and carry more information, thereby enabling several operations to be performed simultaneously.

[0020] TE: For TE waves, it means an electromagnetic wave having only the electric field components (i.e., the components in the x-plane and y-plane) perpendicular to the propagation direction of the electromagnetic wave. The TE mode is often called TE mn and here m and n are mode indices representing the number of nodal points along the x-direction and y-direction respectively. A = TE 10 = the basic mode of electricity B = TE 01 = the secondary mode of electricity The positive orbital angular momentum is the value C = (OAM 1 = +1) = +1. The negative orbital angular momentum is the value D = (OAM 1 = -1) = -1. [Background technology]

[0021] Quantum computing systems are widely known to be developing at an unprecedented pace, offering new computing powers and services unimaginable before, which will likely have an impact on and influence the ever-increasing number of users and their activities in the future.

[0022] Despite the fact that current electronic technology enables the use of computers at unprecedented speeds exceeding petabits and can achieve extremely high computational efficiencies (expressed as bits per second), it is also widely known that the increasing demand for greater computing capacity has led to the exploration of classical computers, substantially based on the standard bit, in relation to the physical limitations of materials and circuits.

[0023] In this case, it is particularly important to develop new methods, along with new technological solutions, that make deeper use of knowledge of natural laws and enable the solving of increasingly complex computational problems, as is required today.

[0024] Furthermore, and particularly important in this regard, is the computational process performed by quantum logic, so-called quantum computation, which has recently been found to be more powerful than classical computation for solving problems of certain magnitudes. This advantage stems from qubits, a quantum capability similar to bits, which maintain stable coherence between different classical states, expressed as a coherent superposition state of 0s and 1s, representing an understanding related to classical bits. This property enables quantum computers to perform computations simultaneously on many classical input states, allowing quantum computers to increase computation speed exponentially, at least theoretically.

[0025] Widely known examples include devices for detecting quantum effects in quantum computing, ranging from astronomy and other experimental sciences to communications, quantum and classical cryptography, and combinations thereof.

[0026] To perform quantum computation, it is necessary to manipulate the quantum states associated with quanta in a controlled manner. That is, in practice, it is necessary to change the quantum state of a particle, for example, a photon, which has a relevant wavelength, polarization, or other quantum mechanical properties.

[0027] This has traditionally been achieved by waveguides such as optical fibers that establish a path for photons, and components that change the quantum state of photons, called "quantum modulators," are inserted along this path. Examples of such components include polarizers and phase adjusters that change the polarization of photons in a controlled manner.

[0028] However, these insertions interrupt the fiber, thus creating "disconnections" in the photon path. No matter how carefully the device is interrupted and inserted, such as with anti-reflective measures, they still result in "photon loss." Since there will be many of these interruptions and insertions along the fiber in the independent circuits used in the calculations, the introduced drawbacks will add up, resulting in a significant loss of photons, and more generally, a loss of information associated with the photons. All of this inevitably translates into significant errors in quantum computing.

[0029] Finally, it must be understood that in order to achieve high computational power at the quantum level, a very large number of qubits must be used to cover the entire state space that we wish to represent. In detail, to date, representing an N-dimensional system in terms of qubits requires at least n1 = log2N qubits. This affects the complexity of the circuit, which increases "exponentially".

[0030] A known solution for obtaining a cudit using only two photons is described in D1, U.S. Patent Application Publication No. 2022171133, which utilizes multiple waveguides (202, 204) and couplers between them (200, 618, etc.) to "couple" the modes of two photons. The couplers in D1 are beam splitter 200 or optical coupler 618, etc.

[0031] However, this does not solve the problem of photon loss, because the photons are forced to "jump" from one waveguide to another. Nevertheless, prior art confirms a consistent direction toward a two-photon coupling solution, as described in the following scientific literature. · "QUANTUM INTERFERENCE BETWEEN TRANSVERSE SPATIAL WAVEGUIDE MODES-NATURE COMMUNICATIONS, VOL.8, 20 JANUARY 2017, PAGE 14010, XP55562943" by D2-ASEEMA MOHANTY et al., ·D3-RAKESH RANJAN KUMAR et al., “QUANTUM STATES OF HIGER-ORDER WHISPERING GALLERY MODES IN A SILICON MCRO DISK RESONATOR, ARXIV.ORG, CORNELL UNIVERSITY LIBRARY,201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA,NY14853,15 MARCH 2020,XP081621671”.

[0032] More specifically, these two articles consistently refer to the coupling of two photons, where the coupler is a unidirectional coupler, used to realize the Hong Hou Mandel interferometer in D2 and the disk resonator in D3. [Prior art documents] [Patent Documents]

[0033] [Patent Document 1] U.S. Patent Application Publication No. 2022171133 [Non-patent literature]

[0034] [Non-Patent Document 1] "QUANTUM INTERFERENCE BETWEEN TRANSVERSE SPATIAL WAVEGUIDE MODES-NATURE COMMUNICATIONS, VOL.8, 20 JANUARY 2017, PAGE 14010, XP55562943" by Aseema Mohanty et al. [Non-Patent Document 2] “QUANTUM STATES OF HIGER-ORDER WHISPERING GALLERY MODES IN A SILICON MCRO DISK RESONATOR,ARXIV.ORG,CORNELL UNIVERSITY LIBRARY,201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA,NY14853,15 MARCH” by RAKESH RANJAN KUMAR et al. 2020,XP081621671” [Overview of the Initiative] [Problems that the invention aims to solve]

[0035] The objective of this invention is to overcome, in whole or in part, the problems of the prior art.

[0036] Another object of the present invention is to obtain a cudit from a single photon, or generally from a single quantum.

[0037] Another object of the present invention is to manipulate a quedit corresponding to a single photon or, more generally, a single quantum.

[0038] Another objective of the present invention is to provide a circuit for quantum computing with minimal loss of quantum information.

[0039] Another objective of the present invention is to simplify quantum circuits by reducing the number of computational units required to cover the state space to be represented.

[0040] Another object of the present invention is to provide a quantum circuit that can be easily and inexpensively implemented.

[0041] A further scope of the present invention is to provide easy modulation of quantum states in quantum circuits. [Means for solving the problem]

[0042] According to a first overall aspect, the present invention relates to a quantum processor comprising at least one input register (5) of information units corresponding to at least one quantum placed as input in at least one waveguide (10) having a longitudinal expansion in a predetermined direction L, defining a path for at least one quantum. Each waveguide is considered to be configured to support the "d" propagation mode of the above at least one quantum. Each waveguide is considered to define an optical path between the input and output sections, comprising at least one modulation shape change section (24) having at least one change section in a cross section orthogonal to a continuous longitudinal direction L. Each waveguide is considered to be arranged in optical continuity with the rest of the waveguide. The modulation shape change section is configured to manipulate the above propagation mode in the above at least one quantum, and the processor comprises at least one output register (15) located in the waveguide (10) at the output section and configured to decode information corresponding to the manipulated mode.

[0043] Conveniently, the present invention makes it possible to generate a cuid and manipulate its state using only a single photon. Since the photon does not jump for manipulation, there is no loss. This is the opposite of the diversion in the prior art.

[0044] However, processors can conveniently handle multiple states associated with each individual photon, and can work with many indistinguishable or even entangled photons, thereby increasing quantum computing power.

[0045] Preferably, the modulation shape changing portion includes a reduction portion in the cross-section of the waveguide.

[0046] According to a particular preferred embodiment, the shape-modifying portion comprises a recess located in the depression of the cross-section, preferably comprising two branches projecting in different directions. These branches are such that, for example, the waveguide in the recess has a substantially V-shaped or L-shaped cross-section.

[0047] According to some preferred embodiments of the present invention, the waveguide is a single optical fiber.

[0048] In this case, preferably the optical fiber has a substantially rectangular cross-section or a core with a substantially rectangular cross-section, and the shape modification portion is a recess at the edge of this rectangular cross-section.

[0049] Preferably, the above operation involves a multimodal transformation, which is defined as switching from one to another among at least several "d" modes of at least one quantum.

[0050] According to some preferred embodiments, the input register is a qubit register with a "d" state, or a qubit register with a "d" level, corresponding to the above "d" propagation mode in at least one quantum. The waveguide (10) has shape support exclusively for the "d" propagation modes and their superpositions.

[0051] In a particularly preferred example, the modulation shape change unit (24) is configured to excite two linearly polarized LP modes in at least one quantum, hereafter referred to as LP modes, with different propagation constants (β1 and β2), starting from one of the input propagation modes. At the output of the shape change unit (24), the two LP modes interfere by generating at least one "d" quantum propagation mode.

[0052] According to some particularly preferred embodiments, d=4 and the waveguide (10) has a rectangular cross-section (12) configured to exclusively support the following four common orthogonal quantum propagation modes. A=TE 10 =Basic modes of electricity B=TE 01 = secondary mode of electricity The positive orbital angular momentum is the value C = (OAM 1 = +1) = +1. The negative orbital angular momentum is the value D = (OAM 1 = -1) = -1. And their superposition.

[0053] In this case, preferably the waveguide has main development dimensions X and Y that are orthogonal to each other and to the direction of quantum propagation Z=L. The cavity has dimensions "a" along X and "b" along Y such that a≧b and a≦2b, and preferably the modulation shape changing section (24) has waveguide regions with L or V cross-sections in a cross-section with respect to the propagation direction Z.

[0054] Preferably, the modulation shape changing section extends over a length L in the propagation direction Z, where L is

[0055]

number

[0056] According to some preferred embodiments, the input register comprises at least one four-state quid A, B, C, D along with at least two polarization states, preferably these polarization states being |H> along X and |V> along Y of the corresponding quantum, and preferably used as control bits (or qubits) for executing a desired circuit, or alternatively, used to double the computational power by creating a quid for each polarization. Alternatively, a pair of OAM states (or two OAM states) can be used as a control register, and the polarizations can be used as a pair of quids.

[0057] Preferably, the input register has at least eight, preferably sixteen, of the four state quedits described above, which are encoded in single quanta or photons, enabling all operations to be performed even with only a single quantum (e.g., a photon). In the prior art, this requires at least two photons. However, the processor of the present invention conveniently has multiple states associated with each individual quantum (e.g., a photon), and can operate with many indistinguishable or even entangled photons, thereby increasing quantum computing power.

[0058] According to some preferred embodiments, the processor is configured to generate at least one C-NOT gate comprising at least one waveguide 10, each having at least one modulation shape changing section, where the C-NOT gate is defined by the following truth table.

[0059] [Table 1] Here, the modulation shape changing unit (24) is configured to change the phase of the cue ditto by periodically shifting the cue ditto state.

[0060] Preferably, the cudit has four states A, B, C, and D, and a value of 1 means a phase change of π / 2, i.e., a periodic evolution of one state to the next A, B, C, D.

[0061] Preferably, the processor is characterized by being configured to generate at least one Toffoli gate CCNOT, which is configured to generate the above-described C-NOT gate. The Toffoli gate is characterized for each photon by the following truth table.

[0062] [Table 2] Here, the C-NOT polarization imposes an additional phase shift on the "d" propagation mode, which constructs the state of the input quedit described above.

[0063] Preferably, the processor is configured to generate at least one Hadamard gate including the C-NOT described above. The Hadamard gate is characterized by having, for each "d" state of a cuedit, represented by the symbol |Sd(n)>, i.e., S = d state of the cuedit, a "+" input obtained by superimposing the starting state using the next |Sd(n+1)>, and a "-" input obtained by inversely executing the basic CNOT circuit, which is obtained by superimposing |Sd(n)> using the previous |Sd(n-1)>.

[0064] According to some preferred embodiments, the above input register is located at the input of a plurality of waveguides. Each waveguide comprises at least one of the above modulation shape changing units (24) and is configured to generate parallel paths, which are coupled by interacting to superimpose, modulate, and modify the cued state. The output register is common to the plurality of waveguides.

[0065] According to some preferred embodiments, at least some of the modulation shape changing sections (24) are different from one another. For example, they have at least different extensions in the propagation direction.

[0066] According to a second aspect, the present invention relates to a cuedit having four states A, B, C, and D, which corresponds to the following modes in quantum propagation. A=TE 10 B=TE 01 C=(OAM 1=+1) D=(OAM 1=-1)

[0067] According to a third aspect, the present invention relates to an input register for a quantum processor. It comprises a quidt of at least four states according to the type described above, and at least two polarization states, for example projecting onto two precise orthogonal bases, such as a horizontal polarization state |H> along X and a vertical polarization state |V> along Y of the corresponding quantum, and is preferably used as a control bit (or qubit) to execute a desired circuit, or alternatively, to double the computing power by creating a quidt for each polarization.

[0068] According to a fourth aspect, the present invention relates to a method for modulating a quid in a quantum processor, and is characterized by providing a quantum processor (1) of the type shown above, by placing at least one quantum characterized in at least one state corresponding to a propagation mode supported by the waveguide at the input of the waveguide (10), and by manipulating the propagation mode by at least one of the modulation shape changing units so that the mode and corresponding state at the output unit correspond to one of the "d" propagation modes, even if the mode and corresponding state at the output unit are different from the input mode.

[0069] Preferably, the method is characterized by using a cudit characterized by the propagation state "d" of the quantum and a control register that includes the polarization state of the quantum.

[0070] According to some preferred embodiments, there are four "d" states represented by A, B, C, D, corresponding to the following propagation modes supported by the waveguide (10): A = TE 10 B = TE 01 C = (OAM 1 = +1) D = (OAM 1 = -1) Here, the above operation involves superimposing two LP linearly polarized modes with different propagation constants (β1 and β2) from the above propagation modes or their superposition by hitting the modulation shape changing part (24) of the waveguide (10), and has a phase. At the output part of the shape changing part (24), these two LP linearly polarized modes prevent the generation of one of the quantum propagation modes "d" according to the phase.

[0071] Another feature and advantage of the present invention will be best confirmed from the following detailed description of the preferred implementation of the present invention, by referring to the accompanying drawings, which are illustrative but not limiting.

Brief Description of the Drawings

[0072] [Figure 1] It is a schematic diagram of a quantum processor in which a quantum circuit has structural continuity and is represented by a waveguide, such as an optical fiber. [Figure 2] It is a schematic diagram of a mode selection cross-section in the waveguide of FIG. 1 having a rectangular cross-section. [Figure 3] It is a schematic diagram of a mode modulation area in the waveguide of FIG. 1 having a modulation shape changing part for the selection area and having structural continuity with respect thereto. [Figure 4] It is a diagram showing a fiber state formed by combining the linearly polarized LP modes (also referred to as LP-like modes) in the modulation area of FIG. 1 according to the phase shift they have with respect to each other at the exit of this area. [Figure 5] It is a diagram showing the electric and magnetic fields in the waveguide of FIG. 1 in a cross-section with a rectangular cross-section cavity before the modulation shape changing part. [Figure 6]This figure shows the first 12 modes in a waveguide with a known common rectangular cavity. [Figure 7] This figure shows the equations demonstrating the selection effect, generated by the cross-sectional dimensions a and b of the four known waveguide modes, TE10, TE01, (OAM)1=+1, and (OAM)1=-1, denoted as A, B, C, and D. [Figure 8] This figure shows the equations demonstrating the selection effect, generated by the cross-sectional dimensions a and b of the four known waveguide modes, TE10, TE01, (OAM)1=+1, and (OAM)1=-1, denoted as A, B, C, and D. [Figure 9] Figure 1 shows the change from the TE10 state to the OAM state due to the modulation shape change section. [Figure 10] This figure shows a quantum C-NOT gate comprising at least one circuit according to the present invention. [Figure 11] This figure shows a quantum C-NOT Toffoli circuit comprising at least one circuit according to the present invention. [Figure 12] This figure shows a quantum Hadamard gate comprising at least one circuit according to the present invention. [Figure 13] This figure shows a quantum processor according to the present invention, in which input and output registers are common to multiple parallel quantum paths. [Modes for carrying out the invention]

[0073] The general principle underlying this invention is to generate a quantum circuit comprising at least one waveguide (e.g., optical fiber) containing at least one local shape modifier, called a modulation shape modifier, which acts on the quantum state of each quantum (e.g., a single photon) passing through the waveguide. The modulation shape modifier has the function of manipulating the quantum state (and its associated quantum information), and this operation (also called modulation) is part of the quantum computing process. This reduces the loss of quanta, such as photons, compared to the known case where an external device is inserted into the waveguide, resulting in a blockage of the quantum path, and consequently reducing computational errors.

[0074] The quantum processor may comprise a plurality of the above-mentioned modulation shape changing units distributed across one or more waveguides.

[0075] Generally, waveguides can be solid or tubular, and in both cases, they are often referred to as cavities in the literature. Therefore, this term is understood as a general synonym for the material through which radiation propagates and related forms.

[0076] From this point forward, we will use devices based on photonic quantum computing as examples, but we will not exclude devices based on similar procedures that utilize the quantum mechanical wave properties of any other quantum, such as atoms, ions, elementary particles, and non-elementary particles, or combinations thereof. We will also not exclude hybrid circuits that utilize two or more quantum types.

[0077] Quantum computation is performed in a single or superposition of states, or more generally, entangled or correlated states, which are generated by a single quantum state or multiple quantum states, preferably through the injection (input) of one or more quanta (photons in this example).

[0078] Each of these quanta has a predetermined quantum state associated with information. Quantum information is represented by the term qubit, which is a superposition of two mutually orthogonal states, or by the term qudit, which is a superposition of two mutually orthogonal "d" states. A qudit is associated with a d-dimensional Hilbert space, and a qubit is associated with a two-dimensional Hilbert space. The projection technique of shapes in Hilbert space allows for mapping Hilbert spaces of different dimensions into subspaces or product spaces. That is, a qudit can be represented by a suitable set of qubits.

[0079] For quantum computing, quanta (in this example, photons) are placed in a circuit (in this example, an optical circuit), along which one or more local modifications are placed in the shape of the waveguide cavity to generate a modulation path for the quantum state of the quantum (photon). To determine the presence or absence of quanta (photons) at a particular moment, a single quantum (photon) (photograph) detector, or an equivalent sensor, is placed at the end of the circuit, thereby obtaining the results required by quantum computing.

[0080] The path of each quantum being computed is obtained by confining either a quantum of light or something else within a set of waveguides. The information carried by (in this example, also one or more cavities simultaneously) is preserved and coupled to its quantum state. That is, it is coupled to a field propagation mode such as the polarization or orbital angular momentum (OAM) of a photon, a cavity mode such as a Hermitian Gaussian mode, or a more general property that can be associated with a quantum state that forms a preferred set of orthogonal states.

[0081] The photons used as quanta in the computation process can, for example, be found simultaneously in all parallel paths within a quantum circuit, or they can have all polarizations or OAM states superimposed in quantum mixing. As the computation progresses, new paths, and therefore new quantum states superimposed with previous coexisting ones, can be generated. The final quantum computation is to manage all superpositions (for example, for 16^16 modules) and project them into a final bit string of, for example, 64 bits using the final decay of the wave function, which is the result of the computation. It is an actual physical process for all intents and purposes, not something simulated using quantum logic in classical computation. Natural language is used.

[0082] Overall, the present invention comprises a processor having at least one input register 5 of information units in the form of a qubit or preferably a quudit, where specific information is encoded and set as input to at least one waveguide 10 that defines the path of the corresponding quantum. The waveguide is configured to support a plurality of "d" propagation modes of the quantum in the input section, and preferably in the output section as well, and includes a local shape-changing section configured to manipulate them by converting at least one propagation mode in the qubit or quudit to one of the other modes.

[0083] From this point forward, the local shape modification section will also be referred to as the modulation shape modification section. At the waveguide's output section, the result is read by decoding the information in the output register 15.

[0084] The modulation shape change section breaks the original rotational symmetry and divides the mode degeneracy into two linearly polarized LP propagation modes (also called LP-like modes, and hereafter referred to simply as LP modes). As the signal exits the modulation shape change section, the two LP modes interfere and, depending on the phase shift, generate one of the supported "d" modes, thereby enabling the modulation shape change section to switch from one mode to another.

[0085] Thus, each qubit or qudit has a "d" state manipulated by the waveguide shape. In a particularly preferred example, the waveguide shape supports four propagation modes, so that each qudit used in the calculation has four states, preferably orthogonal to each other, preferably TE 01 , TE 10These are OAM+1 and OAM-1. The modulation shape changer of each waveguide is configured to convert at least one of the four propagation modes, and therefore the four states of the cudet, to the other one. This conversion occurs because the modulation shape changer is configured to support two orthogonal LP linear polarization modes of each mode, which have two different propagation velocities. In the case of a waveguide with a rectangular cross section transverse to the propagation direction Z and with sides parallel to directions X and Y, for example, the two LP modes have optical axes rotated by approximately 45° with respect to X and Y.

[0086] When two LP modes interfere, they, in phase-dependent manner, generate one of the modes supported by the rectangular waveguide at the output of the modulation shape changer. When the two LP modes are in phase, they form a TE 10 Forms a mode.

[0087] Preferably, the modulation shape changing section is inserted between the input section and the output section of a waveguide that share a common shape.

[0088] Figure 4 shows a conversion table according to the phase difference (referred to as shift or phase shift) of the LP linear polarization mode at the output section of modulation region 24.

[0089] Since the four modes are all orthogonal and independent, they can be used both for channel multiplexing and for setting up four independent modes for quantum computing.

[0090] Based on these fundamental principles, Figure 1 represents a quantum processor 1 comprising at least one quantum circuit 2. The quantum circuit 2 comprises: a waveguide 10, a quantum (e.g., a photon) source 5 which is part of the input register to the waveguide 10, and a quantum (e.g., a photon) detection device 15 which is part of the output register from the waveguide 10.

[0091] Waveguide 10 generally defines an electromagnetic quantum propagation path in the propagation direction Z, corresponding to its main deployment direction.

[0092] In this example, the waveguide 10 comprises a tubular cavity 12 that extends along its overall length z. The cavity 12 may be empty or solid, and is not excluded from being filled with, for example, a dielectric material.

[0093] For example, a waveguide is an optical fiber with a rectangular cross-section for transporting photons.

[0094] Waveguide 10 has at least one propagation mode selection section 19 having a first cross section 20 transverse to the propagation direction Z. In the example in Figure 1, and more clearly seen in Figure 2, this section is characterized by a rectangular cavity 12, i.e., having two main expansion directions X, Y. The dimensions of the cavity 12 in the above directions are "a" and "b", respectively. Preferably, the ratio between a and b is such that it defines the selection of the supported propagation mode. In this example, the cavity in the first cross section 20 is TE 10 Maintain this as the basic mode, and TE 01 It is configured to maintain only the secondary TE mode, suppressing other higher modes.

[0095] More precisely, in this example, TE 10 and TE 01 A rectangular fiber is used, with dimensions (x,y)=(a,b) and two orbital angular momentum modes (OAM)=1=+ / -1, which are specified to have transverse electrical modes (TE) limited to each of the fundamental and secondary modes. As mentioned, the Z axis is the direction of propagation.

[0096] These four are independent and mutually orthogonal propagation modes, indicated by A, B, C, and D, and define the four modes that define the preferred cueed state according to the present invention.

[0097] Note that, by convention, the propagation modes are defined according to the well-known wave field approach based on Maxwell's equations.

[0098] According to this approach, waveguide propagation modes are divided into transverse electric (TE) or transverse magnetic (TM) modes depending on whether the electric or magnetic field is purely transverse to the propagation direction.

[0099] The above cross-section 20 is obtained when a ≥ b. This is because TE 01 The low cutoff frequency for this mode is f c,01 This is because =1 / (2b√(με)) is the active case for a≦2b.

[0100] f c1 If we define it as having = 1 / (2a√(με)), then μ and ε are the inductance and permeability of the fiber, and c = 1 / √(με) is the propagation speed of light in the medium.

[0101] Figure 5 shows the basic mode TE when a ≥ b. 10 The following shows the electric and magnetic fields in the rectangular cross-section cavity 12 of the waveguide 10.

[0102] Figure 6 shows the first 12 modes in a waveguide with a rectangular cavity, as a whole. Arrows indicating the direction and polarization of the E-field are drawn.

[0103] The three "rules" for the E and H(M) fields in a waveguide, according to Maxwell's equations, are: Electromagnetic waves do not pass through or cross conductors; they are always reflected by conductors. • The electric field line in contact with the conductor must be perpendicular to the conductor. • Magnetic field lines adjacent to a conductor must be parallel to the conductor.

[0104] Based on these rules, it is possible to derive the equations in Figures 7 and 8 that show whether the selection of a and b can be carried out in the selection of the supported propagation modes described above.

[0105] Waveguide 10 includes at least one modulation section 24 having a second cross-section 25 at the end of the selected section 19 with respect to the propagation direction Z. The modulation section 24 defines the modulation shape changing portion according to the present invention.

[0106] Supported by the selection region 19, for example, the secondary mode (TE 01 Considering this, passage through the second region 24 excites two orthogonal LP modes with different propagation constants (β1 and β2). As they propagate through the modulation region 24, the two LP modes have a phase shift between them, induced by the modified cavity shape 12a. The modulation region 24 is configured to produce a phase shift of π / 2 or a multiple thereof. This results in the output of one of four independent modes that form a cudet according to the present invention. Figure 9 shows TE 10 This shows an example of modulation to convert the input mode to the OAM output mode.

[0107] In a preferred shape example of section 24, the second section 25 defines a local modification of the shape of the cavity 12, as best seen in Figure 3, and in particular this modified cavity is indicated as 12a and has two main unfolding directions orthogonal to the propagation direction Z. In this example, these two directions are the X and Y directions. To realize this in the example, section 25 has a recess 26 with respect to section 20 in such a manner that it preferably forms L. As can be seen in Figure 1, the outer profile of the waveguide is assumed to follow the cavity except for the wall thickness, and the recess 26 forms a groove along the second section 24, which cuts off one of the edges in the rectangular section of the selected section 19.

[0108] Generally, the shape of the modulation region 24 plays a role in generating LP modes (linearly polarized) and maintaining them in a way that allows them to be distinguished.

[0109] At the exit of the modulation section 24, there is an LP mode interference section 30, which, depending on the phase, rejoins and forms one of the four modes A, B, C, and D. In fact, the interference section 30 has a shape that supports the above four modes, and preferably has the same shape as the selection section 19.

[0110] The phase shift between output LP modes depends on the length of the modulation shape change section in the propagation direction Z.

[0111] This length is denoted by L in Figure 1, and has a phase shift of π / 2 (in this case, TE). 10 OAM = 1 = +1 or 1 + 1) is given by the following:

[0112]

number

[0113]

number

[0114] Generally, L is

[0115]

number

[0116] Regarding the dimensions w and h of the grooves 26 forming the L shape, they are not generally derived from formal calculations, but rather are optimized based on inverse calculation techniques that start from expected results. In particular, they are optimized in reverse with respect to modulation, depending on the selected material, the desired effect of the modified electromagnetic field, and minimizing losses.

[0117] This effect is represented by a perturbation of the refractive index in a multimode waveguide (as in the present invention). Energy can be coupled from one waveguide mode to other modes (A, B, C, D in this example), and the amplitude of each mode along the propagation direction Z can be determined for each mode p in the cavity in question by a set of differential equations of this type (in this example, mode TE). 01 and TE 10 And for the other two, there exist two differential equations and two superposition modes TE 01 and TE 10 This explains the + / -π / 2 phase shift, which corresponds to modes OAM+1 and OAM-1.

[0118]

number

[0119]

number

[0120] Based on equation (1), there are two requirements for multiple mode conversion (i.e., switching between modes A, B, C, and D). The first requirement is to obtain a large value for the mode coupling coefficient between the incident mode (e.g., initiation mode A) and the target mode (e.g., mode B or the next mode to be converted). This concept also applies to mode superposition and therefore to the general cudet created by modes A, B, C, and D.

[0121] The second requirement is that the phase matching condition along the propagation direction is satisfied, which is the propagation constant Δβ (i.e., β) of the oscillating exponential term in equation 1. p -β q This compensates for the discontinuity between the p-mode and q-mode. The phase matching conditions for mode coupling between the p-mode and q-mode can be estimated from the following relationship.

[0122]

number

[0123] Therefore, by using the LP mode to switch from one state to another, or to convert one cue ditch to another, one cavity mode can be converted to another mode with high precision and clarity using cue ditches of various lengths L with grooves 26.

[0124] The simplest example of realizing the quantum device according to the present invention is to employ a circuit configuration based on a set of rectangular waveguide fibers (for example, shown in Figure 13), which is easy to construct and uses other materials such as SiO2, or polyethylene or silicon nitride, which can be printed on a silicon-on-insulator (SOI) chip as support, for example. The grooves 26 can be generated, for example, by locally cutting the optical fibers or by growing the cavities 12a in a suitable manner.

[0125] The primary advantage of selecting these TE modes is that they are stable for propagation in the circuit (rectangular fiber) and do not suppress these TE modes. By selecting transverse fiber dimensions a and b, all other higher-order modes are suppressed, preventing unwanted mode conversion.

[0126] A second major advantage is that the modulation shape change section can be used to transform (or rather modulate) one state to another. For example, the modulation shape change section of groove 26 can transform one of the four states in the rectangular waveguide 12 described above into another state (quantum state shift). In this way, the four fundamental states of each cuudit can be influenced by the cross section of a single groove, which is easier and cheaper to construct than phase masks, plasmon devices, etc., and reduces reflection losses and the discontinuities that would still exist in the switches introduced into the processor to construct computational paths for photons within the processor. In this way, by manipulating the quantum state of photons, some or all of the results in quantum computation can be obtained.

[0127] Another advantage is that circuits based on this rectangular waveguide shape have four natural fiber modes, namely two transverse electrical states and two OAM orbital angular momentum states (TE). 10 , TE 01 This means supporting the equations 1=-1 and 1=+1.

[0128] Formally, for each of these orthogonal modes, we associate each eigenvector and use them to construct a d=4 quedit. The direct advantage is that using a d-order quedit requires fewer quedits to cover the state space of a given problem, such as a quantum register of principal order N, compared to a qubit. This is because a qubit of n1=log2N can represent an N-dimensional system, while n2=log d This is because only N queudites are needed. In this case, n² = log4N.

[0129] The equivalent (d=2) binary in those configurations is O(n 2 1N 2 This requires several qubit gates on the scale of O(n). By analogy, using the same configuration, the scale of the required qubit gates is O(n). 2 2N 2 Therefore, the quedit method is (log2d) more efficient than the simple quedit case. 2 It has scaling advantages. The scaling advantages in a queued object are (log2d) 2 Therefore, in this case, the factor is 4.

[0130] A 16-qubit port with d=4 is equivalent to a 32-qubit port or a 64-qubit port when considering the repetition of each polarization.

[0131] The dimension of n cudits in Hilbert space is dn, where d is the dimension of the cudit. In this case, n=16, so 4 16 =2 32 To obtain.

[0132] This circuit can also operate using qubits by factorizing the size of the qudit relative to the qubit, thus gaining an advantage in terms of circuit complexity.

[0133] We have confirmed that each photon has two independent polarization states, thereby each photon obtains four independent states A, B, C, and D, constructing a cudid of four states, and potentially resulting in eight independent cudids.

[0134] Without ruling out this possibility, however, in order to simplify the circuit and reduce computational errors, it is preferable to configure the processor to be used for controlling the quantum computation, i.e., for the four states of the quudit, instead of using polarization states further for computation.

[0135] Therefore, a preferred configuration of the processor according to the present invention is one that uses the four state cuedits described above and quantum polarization as a control register for quantum computation.

[0136] Input / Output: The input register 5 is created by entangled or single photons distributed along different simultaneous paths, for example, using indistinguishable photons such as NI=N / 4. The input register is preferably created with 16 quids in four states A, B, C, and D, along with horizontal polarization states |H> along X and vertical polarization states |V> along Y, and is used as control bits (quids) to move along the desired circuit.

[0137] One example is an input photon generated using a photon gun machine and timing + mode coding (a time-delayed input to equalize the photon timing), as known from the literature.

[0138] The input into the waveguide can be conveniently generated, for example, via a phase mask or a fixed-phase device equipped with an OAM source, thereby directly imposing those states supported by the cavity 12 of the waveguide 10, for example, the input generating photons using orbital angular momentum (OAM) 1 = -1.

[0139] This simplifies the input circuit, which would otherwise require complex devices, when employing a multiple input configuration, such as 16 qubits, which is desirable in the present invention.

[0140] In the output register 15, the division of the quedit into bits and the reading of N events are performed by a demultiplexer, which can have various shapes and configurations. Preferably, photon detectors are integrated into the circuit at the end of the processor to reduce losses. They are preferably SPADs (Single Photon Avalanche Diodes).

[0141] Optionally, a read-wait state can be provided: to reduce costs, instead of dividing the 64 simultaneous states into 16 or 32 states divided by time slots, read-output quantum states can be considered, and these can be sent via 16 or 32 SPADs instead of 64 different SPADs.

[0142] To minimize errors in quantum computing procedures, it is preferable to stop at a waveguide configuration supporting four independent modes. In fact, if one attempts to expand the limitations of this configuration by employing higher fiber modes and more than two LP modes, rather than limiting oneself to four selected modes (A, B, C, D), errors can be introduced due to mode conversion. This is because, when studying the evolution of electromagnetic field profiles in LP modes with different propagation constants, it has been observed that a simple-mode demultiplexer cannot accurately distinguish LP modes at the output end when more than two LP modes and mode conversions are present, thus introducing additional errors.

[0143] The circuit 2 described so far is a basic circuit and can be used to construct any kind of complex logic quantum circuit, or more specifically, logic gates, also known as gates, within the quantum processor 1.

[0144] One of the basic logic gates that can be constructed using basic circuit 2 is the controlled NOT gate, known as the C-NOT gate. The Toffoli gate, constructed from a series of C-NOT-controlled-controlled NOT gates (CCNOT), and the Hadamard gate, also related to quantum state entanglement, are given as examples.

[0145] A quantum computing processor can be equipped with a series of interconnected C-NOT gates. Additionally, to connect to classical computation and algorithms, a series of C-NOTs can be used to generate Toffoli circuits or gates according to the corresponding principles of quantum mechanics, resulting in a controlled-controlled NOT, or CCNOT.

[0146] The present invention provides a polarization-controlled C-NOT gate for each of the four states of a queudite, with polarization as the control register (Figure 10).

[0147] A classical C-NOT gate operates on a quantum register consisting of two qubits: a control qubit and a target qubit. The C-NOT circuit flicks the second qubit (target qubit) only if the first qubit (control qubit) is |1>.

[0148] [Table 3]

[0149] A C-NOT circuit can be represented by the following matrix. (Inset 1)

[0150] JPEG2026512633000012.jpg2145

[0151] In this case, using circuit 2 and the four-state Qudit, the truth table is equivalent to a value of 1 representing a phase shift of π / 2, i.e., the periodic evolution of one state to the next four states A, B, C, and D.

[0152] Figure 10 shows an example where PSB stands for Polarized Beam Splitter. The waveguide groove 26 changes the phase of the cuedit by periodically shifting the cuedit state.

[0153] CCNOT (Controlled C-not) gate

[0154] Introduction: The Toffoli gate D(π / 2), also known as the CCNOT gate or Deutsch gate, is a 3-bit gate, common in classical computing but not in quantum computing. The Toffoli quantum gate is defined here as having 3 qubits. If it accepts only input qubits that are |0> and |1>, then if the first two bits are in the |1> state, then apply Pauli-X (or NOT) to the third bit; otherwise, apply nothing. This is an example of a CC-U (controlled-controlled unitary) gate. Since it is a quantum analogue of a classical gate, it is fully specified by its truth table:

[0155] [Table 4] Inset Figure 2

[0156] JPEG2026512633000014.jpg54106

[0157] The following table shows implementations of Toffoli gates using Hadamard, phase, controlled NOT, and π / 8 gates. (Inset 3)

[0158] JPEG2026512633000015.jpg23113

[0159] The present invention provides a polarization-controlled toffoli gate in each of the four states of a quedit, with polarization, as the control register (Figure 11).

[0160] Toffoli gates are common reversible logic gates. Classical reversible circuits can be broken down into a series of Toffoli gates.

[0161] By definition, a logic gate n sent into input L using input x, x∈I(L(x)) and output y, y∈O(L(x)) is reversible if the following condition is met: bijection

[0162]

number

[0163]

number

[0164] An example of a reversible gate is the NOT gate. Reversible gates can be described as logical operations that can be performed in quantum computing because they behave as one-to-one transformations and therefore evolve reversibly.

[0165] In fact, the evolution of quantum states can be described either by the Schrödinger equation in terms of one-piece transformations (i.e., preserving the inner product of the eigenstates, which remains the same before and after the transformation), or by the collapse of the wave function, which in this case corresponds to a projection onto the eigenstates.

[0166] The unified transformation U is isomorphic between two Hilbert spaces H1 and H2, which means that U is a bijective function.

[0167]

number

[0168] The Toffoli gate is a general-purpose reversible C-CNOT gate. By definition, following the pigeonhole principle, the number of input bits is equal to the number of output bits. This gate has a 3-bit input and a 3-bit output. The first two bits have an inversion effect on the third bit, and a match occurs for any input only when both are set to 1.

[0169] A Toffoli gate is a mapping of three input bits to an output. {a,b,c}→{a,b,cXOR(aANDb)}

[0170] The generality of the Toffoli gate means that any operation described by a given Boolean function f(x1, x2, ...xm) of variables x1, x2, ...xm can be represented by a circuit constructed using the Toffoli gate. The circuit maps the input variables x1, x2, ...xm, and some additional bits set to 0 or 1, along with the initial variables x1, x2, ...xm, and some additional bits, if possible, to the output created by f(x1, x2, ...xm). Using the Toffoli gate, a quantum computer can perform all conceivable classical computations, although it cannot be used for general quantum computations.

[0171] Referring to Figure 11, the cudit-based Toffoli gate according to the present invention is generated in the following way: A first bit is given the polarization state. A second bit is a phase shift of 0 or π / 2 for modes A, B, C, and D. A second bit is selected to represent another phase shift of 0 or π for modes A, B, C, and D, which is encoded on the cudit and thereby has a 1-1 mapping of polarizations and a fiber state which prevents superposition of states from the Toffoli truth table.

[0172] The quantum Toffoli gate according to the present invention comprises circuit 2 and is controlled by C-NOT polarization. The C-NOT polarization imposes an additional phase shift π / 2 on the A, B, C, and D modes that constitute the cudet of the present invention. For example, if vertical polarization V is selected to represent the first bit with a value of 1, then horizontal polarization H will represent state 0.

[0173] The control register of the present invention is a Hadamard gate with four states of polarized cuedits (Figure 12).

[0174] Generally, the Hadamard Gate acts on a single qubit or qubit state represented by the following qubits: |0>→(|0>+|1>) / √2 |1>→(|0>-|1>) / √2

[0175] Therefore, measuring the output state from an Hadamard gate will generate an entangled state, or rather a superposition of shared states, with various qubits having the same possibility of giving an output of 1 or 0. It is preferably used in an entangled path involving a single quantum.

[0176] In this case, the two state cudits overlap. Here, + indicates a positive state shift, and - indicates a negative state shift.

[0177] The advantage of using state shifts "+" and "-" is that this language effectively simplifies circuits. Starting from one of states A, B, C, or D, input + and pass the photon through a quarter-wave or three-quarter-wave foil, resulting in a polarization evenly distributed along states |H> and |V> (meaning along the X or Y axis, respectively), making any polarization a balanced superposition of polarization states.

[0178]

number

[0179] For each quantum mechanical state of a photon associated with one of the orthogonal propagation modes A, B, C, and D, represented by the symbol |S4(n)>, i.e., S = cudit, there is a "+" input obtained by superimposing the initial state onto the next |S4(n+1)>, and a "-" input obtained by proceeding in reverse through a basic CNOT circuit where |S4(n)> is superimposed onto the previous |S4(n-1)>.

[0180] For example, in the figure, without loss of generality, we have |S4(1)>=C, |S4(2)>D, |S4(3)>A, and |S4(4)>B (which can be renamed as desired), and which have a periodicity of either clockwise or counterclockwise rotation, as shown in Figure 12.

[0181] Finally, referring to Figure 13, a processor 101 according to the present invention is shown, which differs from the processor 1 in Figure 1 in that its input and output registers 5 and 15 are common to a plurality of waveguides 2 that define corresponding parallel quantum paths.

[0182] Waveguide 2 can have different structural shape variations 24a, 24b, and 24c.

[0183] Thus, for example, a quantum particle, such as a photon, can travel through all waveguides simultaneously, creating new paths in the processor and generating new superpositions of states relative to the starting state or previous calculations.

[0184] Generally, in the present invention, it is preferable to use photons as quanta.

[0185] For understanding the object of the present invention, the term “comprising” and its derivatives, when used herein, are intended to be open-ended terms specifying the presence of a declared characteristic, element, component, group, integer, and / or stage, but do not exclude the presence of other characteristics, elements, components, groups, integers, and / or stages that are not declared. The foregoing also applies to words with similar meanings, such as the terms “including,” “having,” and their derivatives. Furthermore, 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 embodiments 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 embodiments described in the operating position. Finally, terms of degree, such as “substantially,” “about,” and “approximately,” when used herein, mean modified terms of a reasonable amount of deviation that does not significantly alter the final result.

[0186] 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 departing from the purpose of the invention as defined in the appended claims. For example, the size, shape, position, or orientation of various components may be changed as needed and / or desired. Components shown to be directly connected to or in contact with each other may have intermediate structures inserted 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. It is not necessary for all advantages to be present simultaneously in a particular embodiment. Each feature that is novel compared to the prior 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 according to 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. It is a quantum processor, At least one input register (5) of an information unit, corresponding to at least one quantum, placed as an input in at least one waveguide (10) that defines a path for at least one quantum and has a longitudinal extension in a predetermined direction L, wherein each waveguide (10) is considered to be configured to support the "d" propagation mode of at least one quantum, and each waveguide (10) has at least one change between the input and output, in a cross section perpendicular to the longitudinal direction L. It is considered to define a continuous optical path comprising a modulation shape changing unit (24), the waveguides (10) are considered to be individually arranged in optical continuity with the rest of the waveguide, the modulation shape changing unit is configured to manipulate the propagation mode in at least one of the quantum particles, and the processor has at least one input register (5) having at least one output register (15) installed in the waveguide (10) at the output unit and configured to decode information corresponding to the manipulated mode. A quantum processor equipped with [a specific feature].

2. The processor according to claim 1, characterized in that the modulation shape changing section comprises a section that reduces the cross-section of the waveguide.

3. The processor according to claim 1 or 2, characterized in that the modulation shape changing section includes a recess provided for defining the depression in the cross-section.

4. The processor according to claim 3, characterized in that the waveguide in the recess has a cross-section with two branches protruding in different directions.

5. The processor according to claim 3, characterized in that the waveguide in the recess has a V-shaped or L-shaped cross-section.

6. The processor according to any one of claims 1 to 5, characterized in that the waveguide is a single waveguide or an optical fiber.

7. The processor according to claim 6, wherein the optical fiber has a substantially rectangular cross-section or a core with a substantially rectangular cross-section, and the modulation shape changing portion is a recess at the edge of the rectangular cross-section.

8. The processor according to any one of claims 1 to 7, characterized in that the operation includes a multimodal transformation, defined as switching from one of at least several "d" modes of at least one quantum to another.

9. The processor according to any one of claims 1 to 8, wherein the input register is a qubit register with a "d" state or a qubit register with a "d" level, corresponding to the "d" propagation mode in at least one quantum, and the waveguide (10) exclusively has shape support for the "d" propagation mode and its superpositions.

10. The processor according to claim 9, wherein the modulation shape changing unit (24) is configured to excite two linearly polarized LP modes in at least one quantum, which have different propagation constants (β1 and β2) and are hereafter referred to as LP modes, starting from one of the input propagation modes, and in the output unit of the modulation shape changing unit (24), the two LP modes interfere by generating at least one of the "d" propagation modes of at least one quantum.

11. d=4 and the waveguide (10) are the following four, namely A = TE １０ = Basic mode of electricity B = TE ０１ = secondary mode of electricity The positive orbital angular momentum where C = (OAM 1 = +1) = +1 D = (OAM 1 = -1) = -1, which is the negative orbital angular momentum. The processor according to claim 10, characterized by having a rectangular cross-section (12) configured to exclusively support a common orthogonal quantum propagation mode of superposition thereof.

12. The waveguide has main unfolding dimensions X and Y that are perpendicular to each other and to the direction of quantum propagation Z = L, the cavity has dimensions "a" along X and dimensions "b" along Y such that a ≥ b and a ≤ 2b, and preferably the modulation shape changing section (24) has waveguide regions with L or V cross-sections in a cross-section with respect to the propagation direction Z, as described in claim 11.

13. The modulation shape changing section extends over a length L in the propagation direction Z, where L is [Math 1] It is equal to, or a multiple of, The processor according to claim 12, characterized in that β1 and β2 are propagation constants of two LP modes.

14. The processor according to any one of claims 11 to 13, wherein the input register comprises four quids A, B, C, D of at least one of the same quantum, together with at least two polarization states, wherein the polarization states are a horizontal state |H> along X and a vertical state |V> along Y of the corresponding quantum, and are preferably used as control bits (or qubits) for executing a desired circuit, or alternatively, used to double the computing power by creating a quid for each polarization, wherein, as an alternative to polarization, a pair of OAM states (or two OAM states) can be used as a control register, and polarization can be used as a pair of quids.

15. The processor according to claim 14, characterized in that the input register has at least eight, preferably at least sixteen, of the queudits in the four states encoded with a single quantum.

16. It is configured to generate at least one C-NOT gate comprising at least the waveguide (10) with at least one modulation shape changing section, the C-NOT gate being defined by the following truth table, Table 1 The processor according to any one of claims 1 to 15, wherein the modulation shape changing unit (24) is configured to change the phase of the quedit by periodically shifting the quedit state.

17. The processor according to claim 16, characterized in that the quedit has four states A, B, C, and D, and a value of 1 means a phase change of π / 2, i.e., a periodic evolution of one state to the next A, B, C, D.

18. A Toffoli gate is configured to generate at least one C-NOT gate, wherein the Toffoli gate is characterized by the following truth table for each photon: Table 2 The processor according to claim 16 or 17, characterized in that the C-NOT polarization forces an additional phase shift into the "d" propagation mode that constructs the state of the input quedit described above.

19. The processor according to claim 16 or 17, configured to generate at least one Hadamard gate including the C-NOT, wherein the Hadamard gate has a "+" input obtained by superimposing the starting state using the next |Sd(n+1)> for each "d" state of a cuedit, where the symbol |Sd(n)> represents S = d state of the cuedit, and a "-" input obtained by inversely executing a basic CNOT circuit obtained by superimposing |Sd(n)> using the previous |Sd(n-1)>.

20. The input register is installed at the input of a plurality of waveguides, each of which is equipped with one of the modulation shape changing units (24). The waveguide is configured to generate parallel paths that interact to couple the states of the cudet by superimposing and / or modulating and / or changing them, and The output register is common to the plurality of waveguides. A processor according to any one of claims 1 to 19, characterized by the above.

21. The processor according to claim 20, characterized in that at least some of the modulation shape changing units (24) are different from each other.

22. A four-state cudet, wherein the four states A, B, C, and D are single quantum propagation modes as follows: A = TE １０ = Basic mode of electricity B = TE ０１ = secondary mode of electricity The positive orbital angular momentum where C = (OAM 1 = +1) = +1 D = (OAM 1 = -1) = -1, which is the negative orbital angular momentum. A four-state cue ditch corresponding to this.

23. An input register for a quantum processor, comprising a quidt of at least four states as described in claim 22, and at least two polarization states, for example projecting onto two precise orthogonal bases, such as a horizontal polarization state |H> along X and a vertical polarization state |V> along Y of a corresponding quantum, preferably used as a control bit (or qubit) to execute a desired circuit, or alternatively, used to double the computing power by creating a quidt for each polarization.

24. A method for modulating a quid with a quantum processor, comprising: providing a quantum processor (1) according to any one of claims 1 to 21; placing at least one quantum, characterized in at least one state corresponding to a propagation mode supported by the waveguide, at the input of the waveguide (10); and manipulating the propagation mode by at least one of the modulation shape changing units such that the mode and corresponding state at the output unit correspond to one of the "d" propagation modes, even if they differ from the input mode.

25. The method according to claim 24, characterized by using a cudit characterized by the "d" propagation state in the quantum and a control register having the polarization state of the quantum.

26. There are four "d" states, indicated by A, B, C, and D, which correspond to the following propagation modes of a single quantum supported by the waveguide (10): A = TE １０ = Basic mode of electricity B = TE ０１ = secondary mode of electricity The positive orbital angular momentum where C = (OAM 1 = +1) = +1 D = (OAM 1 = -1) = -1, which is the negative orbital angular momentum. The method according to claim 24 or 25, wherein the operation comprises generating from at least one of the propagation modes or by striking the modulation shape changing section (24) of the waveguide (10) to generate two LP linear polarization modes with two different propagation constants (β1 and β2) by superimposing them, and the output section of the shape changing section (24) prevents the two LP linear polarization modes from generating one of the phase-dependent "d" quantum propagation modes.

Images