Calibration and stabilization of photonic circuits

Abstract

A method for operating an optical Fourier transform (OFT) device: applying an input signal to at least one component of the OFT device; detecting a first value of a parameter of an optical signal generated by the input signal; and determining a first relationship between the detected first value and a first reference value of the parameter.

Description

Technical Field

[0001] This disclosure relates to an optical Fourier transform (OFT) device and a method for operating the OFT device. Background Technology

[0002] Compared to using electronic circuits to perform mathematical operations such as Fourier transforms, optical and photonic computing methods promise to perform similar operations at higher speeds and lower energy consumption. Traditionally, free-space optics has been used to perform OFTs, but more recently integrated photonics has been employed.

[0003] Free-space OFT systems are well-known and include a Fourier transform lens illuminated by a coherent beam arranged in one or two dimensions. The output from the lens is the OFT of the input beam.

[0004] Photonic integrated circuits (PICs) for OFT include integrated photonic devices capable of performing one-dimensional OFT. The input to an OFT device is an array of input waveguides or optical channels that carry light.

[0005] Photonic devices and PICs are curved devices and are susceptible to manufacturing deviations during production. Therefore, two devices with the same design features may exhibit slightly different behaviors. Furthermore, photonic devices are susceptible to thermal drift and environmental fluctuations, which can lead to signal instability during operation.

[0006] Inside an OFT device, light from the input waveguide or optical channel undergoes diffraction, the diffraction properties of which are determined by the OFT device's geometry. The OFT device's geometry—including the arrangement of the input and output waveguide arrays—is chosen such that the diffracted light forms an interference pattern within the OFT device. This diffracted light illuminates the output waveguide array. In other words, the output optical field at the output waveguide array is the optical Fourier transform of the optical field incident on the OFT device, which is introduced at the input waveguide array.

[0007] The demand for high-speed, high-precision OFT devices is growing. Signal integrity is crucial for accurate and reliable data throughput, especially at high speeds where errors are more likely to occur. If not handled properly, variations in OFT device manufacturing and operating conditions can lead to a loss of signal integrity. Summary of the Invention

[0008] The present invention is defined in the appended claims. Attached Figure Description

[0009] The following description of this disclosure will be better understood when read in conjunction with the accompanying drawings. However, it should be understood that this disclosure is not limited to the precise arrangement of the OFT device, its operation, or the precise logic or data format shown. In the accompanying drawings:

[0010] Figure 1a A schematic diagram of an optical Fourier transform device is shown;

[0011] Figure 1b A schematic diagram of an optical Fourier transform device is shown, illustrating the arrangement of the input and output ports;

[0012] Figure 2 An optical path diagram of a photonic integrated circuit is shown, which includes an optical Fourier transform device;

[0013] Figure 3 A pair of Algonquin diagrams are shown, which illustrate the infinitely continuous possible values ​​in real and imaginary spaces for the input and output states of an optical Fourier transform device.

[0014] Figure 4 A pair of Algonquin plots are shown, which illustrate a finite range of discrete or quantized values ​​in real and imaginary spaces for the input and output states of an optical Fourier transform device.

[0015] Figure 5 A pair of Algonquin plots are shown, which illustrate a finite range of larger quantization values ​​in real and imaginary spaces for the input and output states of an optical Fourier transform device.

[0016] Figure 6 A pair of Algonquin diagrams are shown, depicting the possible input and output states of a deterministic OFT system for a quaternary-binary optical Fourier transform device;

[0017] Figure 7 Four pairs of Algonquin diagrams are shown, each depicting the possible input or output states at a specific input or output port among the four input ports and four output ports of a deterministic OFT system;

[0018] Figure 8 Sixteen figures are shown, depicting the real part of the output state of an OFT system with four input ports and four output ports;

[0019] Figure 9 Sixteen figures are shown, depicting the imaginary part of the output state of an OFT system with four input ports and four output ports;

[0020] Figure 10 A truth table mapping input vectors to expected output vectors is shown for use in a quaternion-binary OFT system.

[0021] Figure 11a This is a block diagram of an example optical Fourier transform system.

[0022] Figure 11b This is a partial exploded view of the block diagram, depicting... Figure 11a The example optical Fourier transform system and its components are depicted in the figure.

[0023] Figure 12 This is a flowchart of an example method for processing the output signal from an optical Fourier transform device, which can be used in... Figure 11a It is implemented in an optical Fourier transform system.

[0024] Figure 13 This is a flowchart illustrating an example method for obtaining a complete complex optical Fourier transform result from an optical Fourier transform device. This method can be used... Figure 11a It is implemented in an optical Fourier transform system.

[0025] Figure 14 This is a flowchart of an example method for processing data using a Fourier transform device. This method can... Figure 11a It is implemented in an optical Fourier transform system.

[0026] Figure 15 This is a flowchart of an example method for processing data using a Fourier transform device. This method can... Figure 11a It is implemented in an optical Fourier transform system.

[0027] Figure 16 This is a flowchart of an example method for operating an optical Fourier transform device or system that includes N input channels and K output channels, for example... Figure 11a Optical Fourier transform system.

[0028] Figure 17 It is a truth table that shows the optical Fourier transform device (e.g. Figure 11a An example of an input-output combination of an optical Fourier transform device, wherein the original output data is modified by a carry-add method to form a PAM4 signal, and the original output data is subsequently recovered from the PAM4 signal.

[0029] Figure 18 It is a truth table that shows the optical Fourier transform device (e.g. Figure 11a An example of an input-output combination of an optical Fourier transform device, wherein the original output data is modified by a first saturation method to form a PAM4 signal, and the original output data is subsequently recovered from the PAM4 signal.

[0030] Figure 19 It is a truth table that shows the optical Fourier transform device (e.g. Figure 11aAn example of an input-output combination of an optical Fourier transform device, wherein the original output data is modified by a second saturation method to form a PAM4 signal, and the original output data is subsequently recovered from the PAM4 signal.

[0031] Figure 20 It is a truth table that shows the optical Fourier transform device (e.g. Figure 11a An example of an input-output combination of an optical Fourier transform device, wherein a first subset of the outputs is derived from a second subset of the outputs based on known relationships between them.

[0032] Figure 21 A flowchart depicting a method for operating an OFT device is shown.

[0033] Figure 22 A schematic diagram of the calibration procedure according to the first embodiment is shown.

[0034] Figure 23 A flowchart depicting a method for calibrating an OFT device is shown.

[0035] Figure 24 A flowchart depicting a method for calibrating an OFT device is shown.

[0036] Figure 25 A lookup table for the calibration procedure described herein is shown.

[0037] Overview

[0038] This disclosure seeks to address the aforementioned and related problems with OFT devices and methods for operating such devices. This overview is not intended to be limiting, and covers various modifications and variations of the described methods and devices in addition to those described.

[0039] calibration

[0040] This disclosure seeks to resolve the aforementioned and related problems of OFT devices and methods for operating such devices.

[0041] The embodiments address the discrepancy between the signal intended to be output from the input channel or optical modulator and the signal actually output from them, and / or the discrepancy between the signal read at the output channel and the actual signal at the output channel.

[0042] This disclosure includes a method for monitoring signal integrity in an OFT device (or apparatus), and an OFT device, controller, or processor configured to perform the method. The OFT device can be any OFT device described herein. After an input signal is applied to the OFT device, the detected optical signal is compared to a reference (e.g., expected, theoretical, or benchmark) value. If the detected signal differs from the reference value, action can be taken to control components of the OFT device in response to the comparison. This is implemented for different parts of the OFT device. For example, the method can be performed to ensure the integrity of input signals transmitted by each input channel, or to ensure the integrity of output signals received at any output channel. Phase control associated with the input and output channels can be adjusted in response to the degree of detected signal integrity to improve signal integrity. The OFT device can be partially or completely calibrated using any of the following implementations of the method for monitoring signal integrity.

[0043] In one embodiment, a method for initializing or calibrating a single input channel is provided. In this embodiment, light generated in the input channel is sampled using a monitoring channel branching off from the input channel. The signal in the monitoring channel is then mixed or combined with a signal from a perturbation channel. The signal in the perturbation channel has a phase that can be controlled by phase control in the perturbation channel. The mixed perturbation signal and the input channel signal are received at a monitoring receiver. The monitoring receiver detects the amplitude of the mixed signal. Assuming that the amplitudes of light in both the signal channel and the perturbation channel are constant, the amplitude of the signal at the receiver is determined by the phase difference between the perturbation signal and the input channel signal; therefore, the monitoring receiver detects the phase difference between the perturbation signal and the input channel signal. This amplitude can be compared to a reference or expected amplitude to determine the phase relationship between the perturbation signal and the input signal. The expected amplitude can be a maximum or minimum amplitude (or any other amplitude), which can be known / pre-recorded. If the amplitude is maximum, this indicates that the phase difference is zero or 2π radians, i.e., the input channel signal and the perturbation signal are in phase. If the amplitude is minimum, this indicates that the phase difference is π radians, i.e., the input channel signal and the perturbation signal are completely out of phase. Once the phase difference between the input channel and the perturbation channel is known, the amplitude of the input channel can be compared to the known value or adjusted to find the maximum value. Phase control in the input channel can then be set so that the phase difference matches (or is within tolerance) the expected or target phase difference. This process can also be repeated for each of the other input channels. This can include using a perturbation signal with the same phase as that used when initializing the first input channel and adjusting the phase control in the remaining input channels. In this way, the quadrature points of all modulators in the input channels can be matched.

[0044] Another implementation of the method for monitoring signal integrity is also provided, which can be performed independently of or after the method for initializing a single input channel. In this implementation (which can be described as a method for calibrating an output receiver), an input vector is applied via the input channel in such a form that only one input state has a value of 1, while the other input states have values ​​of 0 (this form may be referred to as a diagonal input vector, but this term is used for convenience only and its use in this disclosure does not necessarily mean that the input vector must be part of a diagonal matrix). If the method is performed after the method for initializing a single input channel, then 1 is applied to that single input channel. The amplitude of the received optical output signal is then detected at the receiver in the output channel and, given the applied diagonal input vector (e.g., a first diagonal input vector), it is compared with the maximum, minimum, or zero amplitude achievable at the receiver. Simultaneously with the application of the input vector, the maximum and minimum values ​​can be pre-recorded or obtained empirically by scanning the phase of phase control applied to a reference channel. Then, based on the relationship between the amplitude of the output signal detected at the corresponding receiver and the minimum or maximum (or zero) amplitude achievable at the receiver given the first diagonal input vector, the phase control in each reference channel can be adjusted individually. Given the diagonal input vector, using a truth table of expected amplitude values ​​at the receiver in each output channel, the phase control in each reference channel can be adjusted so that the receiver reads a signal that matches (or is within tolerance of) the values ​​in the truth table. In this way, the receivers in the output channels are calibrated by setting the phase control in the reference channels.

[0045] In another embodiment of the method for calibrating a signal applied by a transmitter in an input channel, this embodiment can be performed independently of, or after, the method for calibrating an output receiver, by monitoring the amplitude of the optical signal received at the receiver when another diagonal value (e.g., a second diagonal input vector different from the first diagonal input vector) is applied to the input channel. In this case, the phase control in the reference channel is not adjusted. Instead, the phase control in the input channel with a value of 1 (i.e., the input state) applied in the second diagonal vector is adjusted, and the amplitude of the output signal detected at each receiver in the output channel is compared with the corresponding reference (expected) value (i.e., the output state) in the truth table (which, depending on the applied diagonal input vector, is minimum, maximum, or zero). The phase control in the input channel can then be adjusted such that the receiver reads a signal that matches (or is within tolerance) the value in the truth table. This process can be performed in the same manner for one or more (e.g., all) of the remaining diagonal input vectors, such that the signal at each of the corresponding one or more remaining input channels is calibrated by setting the phase control in the corresponding input channel.

[0046] In addition to any or all of the above-described embodiments of the method for monitoring signal integrity, a final check can also be performed by monitoring the signal at the output channel for any one or a combination of all possible input vectors that can be applied to the input channel. For each applied input vector, the output signal received at one or more receivers in the output channel can be compared with a reference (expected) output signal (e.g., using a truth table). The input vectors can include those expected in the data stream applied to the OFT device.

[0047] In any of the methods described herein, a state machine is arranged to control the OFT device to perform the method. The state machine is connected to a transmitter, input channel phase control, reference channel phase control, and perturbation channel phase control, and sends signals to control the operation of these components to perform the described steps. The state machine is also connected to a monitoring receiver and an output receiver, and receives output signals from these receivers to perform the desired method steps.

[0048] The state machine can also control the host and data transmission channels to control the data flow. In the embodiments for monitoring signal integrity described so far, the state machine stops or pauses the data flow to be processed by controlling the host and / or data transmission channels, and the data passing through the input channels conforms to the embodiments described. Alternatively, the data passing through the input channels conforms to the embodiments described before the initial operation of the OFT device, such as after manufacturing, during OFT device commissioning, or before the start of a data streaming session.

[0049] In contrast, the following implementation of the method for monitoring signal strength can be performed during the normal flow of data to be processed. In this case, the host and the state machine simultaneously send and / or receive data from the components of the OFT device.

[0050] Another implementation of the method for monitoring signal integrity is also provided, which can be performed in combination with or independently of the implementation of the method described above. This implementation involves applying a slowly varying signal to a perturbation channel, such that the phase of the perturbation signal varies over time. This can be performed using phase control in the perturbation channel. Slow variation means a periodic variation much longer than the data frame period during normal use of the OFT device when processing multiple data frames (e.g., approximately 10, 100, 1000, 10,000, or 100,000 times, or any range therebetween). The data frame period can be based on the operating frequency of the OFT device, on the order of GHz. The slowly varying perturbation signal is compared to a time-averaged input channel signal obtained using a monitoring receiver (e.g., during normal use of the OFT device processing a data stream). The perturbation signal can be a linear phase-varying, sinusoidal, or any other slowly varying signal. The combined input channel and perturbation signal will have a desired relationship, thus, based on the perturbation signal, the desired signal will be present at the monitoring receiver for each input channel. If the signal detected at the monitoring receiver differs from the expected signal, or if the difference exceeds a threshold, this can indicate that drift has occurred in the input channel signal. If the difference is within the amplitude of the detected mixed signal, this can be used to determine that amplitude drift has occurred in the input channel. If the difference is within the derivative or gradient of the detected mixed signal, this can be used to determine that phase drift has occurred in the input channel.

[0051] Another implementation of the method for monitoring signal integrity is also provided, which can be performed in combination with or independently of the implementation of the method described above. In this case, the time-averaged signal from all output channels (e.g., during normal use of the OFT device processing the data stream) is compared with an expected time-averaged signal. The expected time-averaged signal is the average signal from all possible output vectors when all possible input vectors are represented in a given data stream. The expected average signal may differ at different output channels. If a difference exists between the detected time average and the expected value, this can indicate the presence of drift. If drift is detected, any or all of the above-described implementations of the method for monitoring signal integrity can be performed to reset the system, either during data stream operation in the case of small drifts within tolerance, or during data stream pauses in the case of drifts greater than the tolerance value. The method is most effective if all possible input vectors are equally represented in the data stream.

[0052] In a specific example of an OFT device, if the number of inputs and outputs is limited to only four ports (quaternary), and the input data is further limited to unsigned 1 bit (binary) (i.e., a quaternary binary OFT system), the receiver design is simplified to a five-level (PAM5) design. This allows for faster data processing, similar to that of photonic transceivers used in data and telecommunications systems. Quaternary binary OFT devices require communication with an external host to ensure the correctness of data processing. The methods and related implementations for ensuring signal integrity described herein are particularly important for such devices, as high throughput can lead to a loss of signal integrity if not properly addressed.

[0053] The advantages of the method described in this article may include:

[0054] • Improve signal integrity in quaternary binary OFT devices.

[0055] • Control the data processing pipeline based on the signal integrity status, where the data processing pipeline is a meaningful stream of digital data that needs to be processed by a quaternary binary OFT device.

[0056] • Control the stability of the modulator device in a quaternary binary OFT device.

[0057] • Control the stability of the OFT device in a quaternary binary OFT device.

[0058] • Control the stability of the input and output channels (or waveguides) connected to OFT devices.

[0059] Deterministic, quantitative OFT apparatus and methods

[0060] In this embodiment, instead of providing continuous input to the waveguide or input ports, the OFT input is quantized or discretized at each input channel, such that the input optical field has only a finite number of allowed values ​​and a finite number of combinations of values ​​over a finite set of input ports. The discretization (or quantization) of the input channels results in the discretization (or quantization) of the output optical field at the output channels. The output is also deterministic, and the output value can be predicted in advance if the input values ​​are known.

[0061] Predictable output values ​​existing only in discrete levels or bands allow for highly accurate and reliable determination of the output OFT without the need for high-precision receivers / decoders and associated ADCs. Similarly, discretization of the input allows for the creation of the input optical field without the need for high-precision transmitters / encoders and associated DACs. Therefore, receivers with lower effective bits per second (ENoB) can be used compared to analog OFT devices. Lower ENOB allows for faster throughput because the receiver can operate at higher frequencies, increasing bandwidth and data throughput. Furthermore, the predictability of the output for a given input allows for reliable integrated testing and calibration of the OFT device, providing a foundation for more reliable and robust data processing. Moreover, when the output is limited to a small number of output values ​​(or levels or states), data can be more easily packaged and transmitted over widely used telecommunications links.

[0062] In other words, in a simulated OFT device, the possible states or values ​​transmitted (or encoded) from each input channel at the input port are continuous and can have any value within the dynamic range of the transmitter in the input channel. Therefore, the possible states or values ​​received (or decoded) at the output channel are not discrete (also continuous) and can have any value determined by the OFT of the input optical field within the dynamic range of the OFT device.

[0063] In order to determine the OFT of this simulated optical function, the receiver (or detector or decoder) in the output channel must be able to determine the optical signal with high accuracy and precision. In other words, the receiver must have high resolution and produce repeatable readings so that the OFT device can process accurately across multiple frames.

[0064] The inventors have recognized that by controlling the optical signal at the input channel, a quantized output signal can be obtained at each output channel, such that each input channel produces a quantized signal with a finite number of states or levels. In this way, the output becomes (easier) to determine because the output can be mapped by predicting the signal state at each output channel based on a pattern of each possible input state. Specifically, real and imaginary values ​​at the output channels can be detected separately to determine the complete complex value at each output channel. In this way, the output OFT of the input function is determined for a single frame.

[0065] Furthermore, because the output is quantized, the receiver or decoder at each output channel can determine the correct output state faster or more accurately (or both), or a receiver or decoder with lower resolution or lower fidelity can be used. Therefore, more accurate, higher frequency, and / or cheaper OFT devices can be obtained.

[0066] In other words, traditionally, the possible input values ​​(or input states) of an OFT device or PIC are unrestricted and can have any value. However, in the OFT apparatus and method described herein, the input and output values ​​in the OFT are optimized such that the output of the OFT:

[0067] It is not a numerical continuum, but rather discrete or quantized. This means that when noise is introduced at the transmitter or receiver, as long as the interval between quantization levels is greater than the noise level, the result can still be classified into a quantized state and is deterministic.

[0068] • It can be represented by a finite number of output quantization states, and the quantization states have a uniform difference.

[0069] • It allows for the offset or adjustment of the quantized states, which is a multiple of the difference between the quantized states. This makes it possible to design reusable receivers with predetermined or known offsets.

[0070] • It can be constructed as a truth table and used as a function of the input state.

[0071] • It can have quantized state intervals that are proportional to the intervals between input states (inherited from the intervals between input states, scaled according to any loss in the OFT device).

[0072] • It can have a larger dynamic range than the input; the maximum dynamic range of any output channel scales with the number of inputs.

[0073] Advantages may include: the electronic driver of the OFT can be constructed using the same transmitter or encoder electronics for each input channel in the OFT, with offset or adjustment, and / or the same receiver or decoder electronics for each output channel in the OFT, with simple offset or adjustment, but this disclosure is not limited thereto.

[0074] Methods for controlling an OFT device include any one or a combination of the following:

[0075] • Limit the number of inputs to OFT while keeping the number of inputs the same as the number of outputs from the OFT device.

[0076] • Limit the dynamic range of the input and each input channel.

[0077] • Introduce dynamic range and discretization asymmetry between input and output.

[0078] • Offset or adjust the output of some or all output channels so that the dynamic range of all channels is a subset of the dynamic range of the entire OFT output.

[0079] OFT devices may include:

[0080] • The same transmitter or encoder electronic driver is used for each input channel.

[0081] • The same receiver or decoder electronic driver is used for each output channel.

[0082] • Transmitter or encoder, and receiver or decoder electronic driver, which allows them to operate at the highest possible frequency, for example by controlling the receiver's ENOB in any of the ways described herein.

[0083] Therefore, a method for operating an optical Fourier transform (OFT) device is provided, the OFT device including N input channels and K output channels, the method comprising: quantizing light at the input channels to form a quantized output at each output channel.

[0084] An optical Fourier transform device is also provided, comprising: N input channels and K output channels. Each input channel includes a modulator, transmitter, or encoder, arranged to transmit a quantized input signal. Each output channel includes a receiver, arranged to detect the quantized state at the output channel.

[0085] An OFT device (or OFT apparatus) can be any optical device capable of performing an optical Fourier transform on an input light field transmitted via two or more input channels and received at a corresponding number of output channels. This can take the form of a waveguide array comprising N input channels, whose outputs are split into N mixing channels, and then the nth mixing channel from each input channel is combined to form K output channels. However, a more efficient arrangement is a free-space OFT device, in which the N input channels transmit light into a free-space region that allows the light from the input channels to diffract and produce interference patterns at the K output channels. N is an integer at least 2, optionally a multiple of 2 or 4, optionally 4. K is at least equal to N. See the example of an OFT device. Figure 1a , 1b The examples are described in 2, but this disclosure is not limited to these examples.

[0086] For an input vector, which includes frames of input states simultaneously present at the input channels (e.g., optical states defined by, for example, amplitude and / or phase), the complete discrete optical Fourier transform of the frames of input states is deriveable from the output states detected at all output channels. In this sense, the OFT device is arranged to perform the discrete Fourier transform. The discrete Fourier transform satisfies Equation 1:

[0087] (Equation 1)

[0088] Where k is the index of the output channel, X k It is the output of channel k, where n is the index of the input channel, x n It is the input of channel n, where N is the number of input channels.

[0089] Quantizing the light at the input channels is to limit the possible input states of the light at each input channel, such that each state is separated. In other words, quantizing the light at the input channels involves limiting the light emitted from each input channel to a finite number of possible input states. Each input state can be defined by a specific value of amplitude and / or phase. In an OFT system suffering from noise, the possible states are separated by an interval of appropriate size relative to the noise so that they can be reliably distinguished from each other by any detector or monitor connected to the input channels. That is, an appropriate signal-to-noise ratio is obtained by choosing an appropriate quantization interval. An example of how input state quantization leads to output state quantization is illustrated in the Algonquin diagram, see reference. Figure 4 and Figure 5 This disclosure is described, but is not limited thereto.

[0090] Restricting each input channel to two possible input states allows for binary input to each input channel, enabling data to be transmitted to each input channel at any given time in single-bit form. Alternatively or further, the input channels are controlled such that the possible input states of one input channel are identical to the possible input states of every other input channel. These measures, individually and in combination, provide a simplified interface with electronic systems controlling the input channel states and a simplified data flow through the OFT device. They also maximally limit the number of possible output states received at the output channel, thereby simplifying data extraction from the OFT device. An example of maximally limiting (i.e., binary) input states and their effects on output states is provided in [reference]. Figure 6 and Figure 7 An Algonquin diagram is used to describe this. However, the possible input states are not limited to binary. For example, three or even four possible input states are possible while still ensuring useful output quantization.

[0091] Limiting possible input states can be achieved by configuring one or more transmitters in each input channel and / or the electronics driving one or more transmitters such that they can only receive a limited number of input states. Alternatively or additionally, the data sent to one or more transmitters can be limited to restrict possible input states.

[0092] Possible input states are limited by the phase and / or amplitude of the light carried by the input channel. For example, possible input states may include or consist of a first input state having a first amplitude and a second input state having a second amplitude different from the first amplitude, wherein the first input state is in phase with the second input state. In other words, the input state is limited by amplitude shift keying (ASK). In another example, possible input states may include or consist of a first input state having a first phase and a second input state having a second phase out of phase with the first phase, wherein the first input state has the same amplitude as the second input state. In other words, the input state is limited by phase shift keying (PSK). In yet another example, possible input states may include or consist of a first input state having a first phase and a first amplitude, and a second input state having a second phase out of phase with the first phase and a second amplitude different from the first amplitude. In other words, the input state is limited by binary phase shift keying (BPSK).

[0093] Increasing the interval between possible input states can improve the signal-to-noise ratio (SNR) in a system. This can be achieved by configuring one or more transmitters or the electronics that control them, or by controlling the data sent to one or more transmitters. For example, in any of the ASK, PSK, or BPSK examples, the first and second states are separated by two or more distinguishable states, which are determined by the SNR and the resolution of one or more transmitters (and / or associated control electronics).

[0094] For maximum effect, the first input state is the lowest possible input state of the input channel, and the second input state is the highest possible input state of the input channel. Otherwise, the first input state and the second input state have the highest possible separation—for example, the first phase and the second phase are out of phase by 90 degrees or 180 degrees.

[0095] Possible input states can be restricted in any way described to ensure that the input state is an unsigned real number or an unsigned imaginary number.

[0096] Limiting the number of possible input states for each input channel restricts the number of possible input vectors. An input vector is defined as a representation of the input optical signal applied to the OFT device by the entire set of input channels. Possible input vectors are limited by the combinations of possible input states for each input channel. The number of possible input states can be limited such that the set of all input channels can only display M... NThere are N distinct input vectors, where each combination is further an N log2M bit combination, where each input vector is represented by a set of input states applied to the input channel in a single frame. For example, for a quaternary-binary system (four input ports and binary input states), the possible input vectors can be represented as

[0000] to

[1111] and fourteen vectors in between, as shown in reference Figure 10 M is the number of levels in the input and is a positive integer. A frame is a snapshot or sampling period during which the input vector remains constant. The input vector can vary frame by frame, causing the output vector received at the output channel set to vary frame by frame. The input vector is defined by frames of input states that are simultaneously present at the input channels (e.g., optical states defined by, for example, amplitude and / or phase). Similarly, the output vector is defined by frames of output states that are simultaneously present at the output channels, resulting from the aforementioned frames of input states (e.g., optical states defined by, for example, amplitude and / or phase). The channel of the input vector is defined as the input state (e.g., optical state defined by, for example, amplitude and / or phase) present at one of the input channels (e.g., at one of the input ports at the end of the input waveguide). Similarly, the channel of the output vector is defined as the output state (e.g., optical state defined by, for example, amplitude and / or phase) present at one of the output channels (e.g., at one of the output ports at the end of the output waveguide). The input channel and output channel can be represented by values ​​indicating the input and output states, respectively. Similarly, the input (or output) vector can be defined by a series of values ​​that define the frame of the input (or output) state that exists simultaneously at the input (or output) channel.

[0097] The output of an optical Fourier transform stage is deterministic, such that a finite number of possible input states produce a finite number of possible output states at each output channel. In this way, limiting the possible input vectors limits the number of possible output vectors. The output from each channel in an OFT device or PIC can be represented by a subset of quantized values ​​that represents the complete result set of the OFT device or PIC. The output can be offset or adjusted by any multiple of the difference between the quantized states to form the OFT output value. For an example of how the output vector is limited for a binary input system with four inputs and four output ports, see [reference needed]. Figure 8 and Figure 9 This disclosure is described, but is not limited thereto.

[0098] By limiting the number of possible output vectors (and by limiting the number of possible input vectors), the output receiver does not need to have high resolution. A high-resolution receiver can still be used, but it can be replaced with a lower-resolution receiver to provide improved bandwidth. A lower-resolution receiver may only receive discernible output states, which are determined by the signal-to-noise ratio and the resolution of one or more receivers (and / or associated control electronics). The number of discernible output states of the receiver can be described by the receiver's ENOB. By limiting the receiver's ENOB, the system's data throughput or the operating frequency of the receiver (and therefore the OFT device) can be increased.

[0099] In other words, the receiver ENOB does not need to be limited; however, doing so can be useful to increase the maximum operating frequency of the OFT device. To achieve the maximum operating frequency, ENOB can be limited such that the number of levels for each output channel receiver equals the maximum number of levels for the output signal, which can be generated from the quantized input. For this purpose, the effective number of bits "ENoB" for each output channel receiver is equal to log2(P+Q), where P is the total number of possible output states at the corresponding output channel for a given possible input state, and Q is the number of spare levels (equal to 0 for the minimum ENOB), which can be added in addition to the minimum level required to read the output signal. In other words, Q can be greater than zero, for example, 1, 2, 3, 4, 6, 8, or 10. This can reduce the need for offsetting the receivers to cover all levels of the output signal, or allow the OFT device to remain usable if the number of possible input states increases. Therefore, increasing Q can improve the ease of configuration or increase the flexibility of use of the OFT device.

[0100] ENoB does not necessarily need to be limited to this extent. Lower ENoB, which can be 12 or less, optionally 10 or less, optionally 8 or less, optionally 6 or less, optionally 4 or less, optionally 3 or less, optionally 2 or less, can increase the operating frequency compared to more continuous or pseudo-continuous receivers.

[0101] The level intervals in the receiver match the intervals of the possible output states. For ENOB values ​​greater than the minimum (i.e., when Q > 0), the receiver level intervals either match the intervals of the possible output states (with spare levels above and / or below), or the receiver level intervals are a fraction of the number of possible output states (with a rational denominator).

[0102] The possible input states are controlled such that a uniform difference exists between the first and second possible output states, and between the first and third possible output states. Alternatively or additionally, a uniform difference exists between the first and second possible output states; and between a third output state different from the first and second output states; and between a fourth output state different from the first, second, and third output states. The output receiver configuration can be simplified if a uniform difference exists between all adjacent possible output states; that is, if the output of OFT is represented by a finite number of output states, and the output states are evenly spaced. Adjacent output states are either real or imaginary output states.

[0103] The difference between adjacent quantized output states for each output channel is proportional to the difference between adjacent quantized input states for each input channel. This proportionality constant is proportional to the losses in the OFT device or associated systems or components connected to the OFT device (e.g., DAC or ADC systems, amplifiers, filters, or analog equalizers).

[0104] OFT data packaging

[0105] This disclosure includes a method for processing raw output data from an optical Fourier transform device, making the data easier and more efficient to process without losing any information related to the optical Fourier transform result. The method involves modifying the raw output data to make it compatible with PAM4. In other words, this disclosure includes a method for processing an output signal from an optical Fourier transform device, comprising: applying an input vector to the optical Fourier transform device to obtain an output vector representing the optical Fourier transform of the input vector; and modifying the channel (i.e., level or value) of the output vector to form a PAM4 signal.

[0106] Depending on the configuration of the OFT device, the raw output data may be inherently incompatible with PAM4. Due to the number of possible output states that can be received at the output channel of the OFT device, the raw output data may be in a format other than PAM4. For example, if the number of possible output states that can be received is 5, the raw output data may be referred to as (or include) a PAM5 signal (which has 5 possible states, levels, or values). That is, the number of levels in the channel can be greater than 4, optionally, where the channel carries a PAM5 signal (which has 5 possible levels). Modifying the channel involves monitoring a 5th (or higher) level to generate a PAM4 signal.

[0107] The modified original output data may have some channels that are PAM4 compatible without modification. Other channels of the original output data are not PAM4 compatible, and these channels can be modified according to the methods described herein. The channels of the output data (or output signal) can be understood as the data received at the physical output channel of the OFT device.

[0108] The original output signal comprises the complete complex result of the optical Fourier transform from the optical Fourier transform stage. This channel can represent the real part of the optical Fourier transform. In a quaternary-binary OFT device, the first or third output channel of the optical Fourier transform device represents the real part of the Fourier transform result. Due to the range of possible output vectors in such a system, the real part at the first and third channels includes a PAM5 signal. That is, for some input vectors, a fifth-level signal is received at the first or third channel. This is, for example, in... Figure 10 As shown in the truth table, for input vectors (1010) and (1111), their output vectors are (2020) and (0040) respectively (note that output channel Out1 in the table has five levels from -2 to 2, and output channel Out2 has five levels from 0 to 4).

[0109] The original output data is modified frame by frame. In each frame, the original output data of the OFT device includes an input vector and an output vector. The input vector defines the input state at each input channel as a function of the input channel position, and the output vector defines the output state at each output channel as a function of the channel position. Therefore, the vector is a matrix, where the positions in the matrix correspond to the physical channel positions in the OFT device (which may have cyclic offsets). As described throughout this disclosure, the output states are quantized. For a one-dimensional OFT device with N input ports, the vector can be in vector form (output state at channel 1...output state at channel N). If each row of output data is represented in vector form as described above (e.g., (1030)), Figure 10 The truth table provides examples of vectors.

[0110] The raw output data can be modified using any of a variety of methods. These methods include operating on incompatible channels of the raw output data to remove one or more levels, reducing the number of levels to four, or operating on compatible channels of the raw output data to add one or more levels. When more than four levels exist, modifying the channel involves detecting the highest level of the channel and removing it. When fewer than four levels exist, modifying the channel involves adding one or more levels. For example, if the channel of the raw output data is a PAM5 signal, the method involves removing the highest (fifth) level, or removing the lowest first level and offsetting the remaining levels by subtracting 1. Another example is if the channel of the raw output data is a PAM3 signal, the method involves adding a level to the channel to form a PAM4 signal. Adding a level to a PAM3 signal may involve adding more than one level if the value on the channel is at the first or second level.

[0111] Removing the highest level from a channel includes saturating or clipping the highest level to a fourth level, or subtracting one or more levels to form a fourth level. In the PAM5 example, the fifth level can be removed by simply subtracting one level from the fifth level of the signal, but it can also be removed, for example, by saturating the fifth level to make it the fourth level. More generally, higher levels (fifth and above) of incompatible channels are saturated. Saturating higher levels can be achieved on the receiver itself by reducing the dynamic range of the receiver in the output channel of the OFT device to clip the higher levels, or by post-processing the raw output data, such as by amplification and / or applying a limiter to achieve saturation.

[0112] Different channels of the original output signal may have different numbers of levels, some more than four and some less than four. In this case, carry-add logic can be applied to a pair of channels—the first channel has more than four levels and the second channel has less than four levels. One or more (or all) levels exceeding four can be removed from the first channel (by subtraction or saturation) and then added to the second channel, making both channels PAM4 compatible. Carry-add logic can be applied to create a unique output vector—where, given all possible input vectors, the value across the channels is a unique combination of all possible output vectors available through an OFT device.

[0113] Once the channel is modified to form a PAM4 signal, the PAM4 signal can then be transmitted via a PAM4 link. In this sense, the signal becomes PAM4 compatible. This makes it possible to use optical Fourier transform devices as part of the communication channel, allowing data to be processed inline within the communication channel.

[0114] For detailed examples of data packaging methods, please refer to [link / reference]. Figure 12 Describe it.

[0115] Data unpacking

[0116] The original output data can be recovered from a PAM4 signal by inverting the modification of the output signal, for example, by inverting the modification of the output signal channel. In other words, the method can include recovering the correct complete complex OFT data from a PAM4 signal by inverting the modification of the channel.

[0117] When one or more levels are initially removed from the original output data to form a PAM4 signal, the correct complete complex OFT data includes the one or more removed levels derived from the PAM4 signal. For example, when the fifth level is initially removed from the original output data to form a PAM4 signal, the correct complete complex OFT data includes the fifth level of the PAM5 signal derived from the PAM4 signal. This can be achieved by performing logical operations on the PAM4 signal. For example, when performing a carry-add operation on the channel to form a unique output vector that includes the PAM4 signal, the original output data can be recovered by understanding the relationship between the unique output vector and the applied carry-add logic. Inverting the carry-add logic provides the original output data.

[0118] By understanding the method for modifying output data to form a PAM4 signal described in this disclosure, the original data can be derived from the PAM4 signal by an entity receiving the data, such as via a PAM4 link, even if the entity is not the entity performing the optical Fourier transform. To this end, this disclosure also includes a method for processing an output signal from an optical Fourier transform device, the method comprising: receiving an output signal in the form of a PAM4 signal from the optical Fourier transform device; and recovering the complete complex output of the optical Fourier transform device from the output signal. This can be achieved by decoding the PAM4 signal, for example by deriving a signal with more than four layers from the PAM4 signal, optionally deriving a PAM5 signal from the PAM4 signal. Alternatively, decoding the PAM4 signal includes deriving a signal with fewer than four layers from the PAM4 signal, optionally deriving a PAM3 signal from the PAM4 signal. The complete complex output of the optical Fourier transform device recovered from the PAM4 signal is defined at least in part by the PAM5 and / or PAM3 signals. In other words, at least a portion of the complete complex output of the optical Fourier transform device recovered from the PAM4 signal can be defined by the PAM5 and / or PAM3 signals.

[0119] Detailed examples of data unpacking to recover the original output signal from a signal including a PAM4 signal are available for reference. Figure 15 Describe it.

[0120] Logical derivation of the output

[0121] This disclosure also includes using optical methods to determine the Fourier transforms of some input vectors in the data stream, and using other (non-optical) methods to determine the Fourier transforms of other input vectors in the same data stream; in other words, not performing optical Fourier transforms on other input vectors. The inventors have recognized that although optical Fourier transforms can generally process input vectors at much higher frequencies than electronic methods, the overall data throughput of OFT data can be further improved by selectively avoiding performing optical Fourier transforms on subsets of input vectors in the data stream and using other methods instead.

[0122] Therefore, a method for processing data using a Fourier transform device is provided, the method comprising applying an input data stream to the Fourier transform device, wherein the input data stream is applied in a series of sequential frames, each frame including an input vector. The method further comprises generating an output data stream from the input data stream by determining an output vector for each frame, wherein each output vector is a Fourier transform of the corresponding input vector in the input data stream. For a first input vector having a first value, a corresponding first output vector is determined by an optical Fourier transform of the first input vector. For a second input vector having a second value different from the first value, a corresponding second output vector is determined by means of an optical Fourier transform of the second input vector.

[0123] For a second input vector, the corresponding second output vector can be determined by applying logical operations to the second input vector. For example, if the second input vector has a total value of zero, the corresponding second output vector is determined either by copying the second input vector to produce the second output vector, or by determining that the second input vector also has a total value of zero. That is, a specific input vector for which a Fourier transform is performed, other than optically, can have a total value of zero. In this case, the resulting output will also have a total value of zero, where the input vector is simply copied to form the output vector, or the output vector is determined logically, digitally, or electronically from the input vector to have a total value of zero.

[0124] Alternatively, for the second input vector, the corresponding second output vector is determined by the optical Fourier transform of the first input vector. In this case, the second input vector is encoded in the input data stream as a repeated input of the first input vector, and the second output vector is logically determined from the optical Fourier transform of the first input vector. The second input vector can be one that has a closely matched corresponding output vector to the first input vector, while the output vector of the first input vector is, conversely, optically determined using an OFT device. In this case, the second output vector can contain the same corresponding vector value as the first output vector, except for one different vector value, wherein the second output vector is logically determined from the first vector by a known relationship between the different vector values. This known relationship can be sign inversion.

[0125] For a detailed example of the logical derivation of the output, please refer to Figure 14 Describe it.

[0126] Receiver optimization

[0127] The overall receiver configuration can be simplified by determining the form of all possible output states at each output channel. Given the number of input and output channels and all possible input states at each input channel, the form of all possible output states can be theoretically determined. This can be done mathematically based on the discrete Fourier transform equation (e.g., Equation 1) or electronically. The output vector for each input vector can be stored in a lookup table or other reference database. The lookup table can contain all possible output vectors. Knowing all possible output vectors, the inventors have realized that the receivers in the output channels can be simplified by identifying redundancy in the output, or by using only one receiver for each output channel, while still being able to determine the real and imaginary parts of the output state at each output channel. Furthermore, the inventors have realized that independently detecting imaginary and real values ​​at the output can also simplify the receiver configuration.

[0128] For example, each output channel has a receiver that is different from the receiver of at least one other output channel. Alternatively or additionally, at least one output channel includes a receiver for recovering only real data, while another output channel includes a receiver for recovering only imaginary data; for example, three output channels include receivers for recovering only real data, and another output channel includes a receiver for recovering only imaginary data. Alternatively or additionally, the number of levels that the receivers can detect may differ. For example, the ENOB of the receiver of one output channel may differ from the ENOB of the receiver of another output channel. Optionally, each output channel may have only a single receiver. The receiver being compared is the receiver used to detect the OFT result, and not a receiver used for any other purpose, such as monitoring.

[0129] Therefore, a processing method is provided to obtain a complete complex optical Fourier transform result from an optical Fourier transform device, which includes a first output channel through a fourth output channel. The method includes determining the real part of the channel output signal for each of the first, second, and third output channels, and detecting the imaginary part of the channel output signal in the fourth output channel. The fourth output channel is distinct from the first, second, and third output channels. The method further includes determining the complete complex result by obtaining the real and imaginary parts of all four output channels using only the detected real parts from the first, second, and third output channels and the detected imaginary part from the fourth output channel.

[0130] The operations applied to the fourth and second output channels can be interchanged, such that the method includes determining the real part of the channel output signal in each of the first, fourth, and third output channels, and detecting the imaginary part of the channel output signal in the second output channel. The second output channel is distinct from the first, fourth, and third output channels. The method also includes determining a complete complex result by obtaining the real and imaginary parts of all four output channels using only the detected real parts from the first, fourth, and third output channels and the detected imaginary part from the second output channel.

[0131] Optionally, obtaining the imaginary part of the channel output signal in the second output channel may include applying logical operations to the imaginary part detected in the fourth output channel. Alternatively, in an alternative arrangement, obtaining the imaginary part of the channel output signal in the fourth output channel may include applying logical operations to the imaginary part detected in the second output channel.

[0132] Alternatively or further, obtaining the real part of the channel output signal in the fourth output channel includes applying a logical operation to the real part of the channel output signal detected in the second output channel. Alternatively, in an alternative arrangement, obtaining the real part of the channel output signal in the second output channel includes applying a logical operation to the real part of the channel output signal detected in the fourth output channel.

[0133] Alternatively or additionally, obtaining the imaginary part of the channel output signal in the first and third output channels includes determining that the imaginary part is zero. Alternatively or additionally, obtaining the real part of the channel output signal of any one of the first to third output channels and / or obtaining the imaginary part of the channel output signal of any one of the second or fourth output channels includes converting the detected optical output signal level into a signed number or its digital equivalent.

[0134] More generally, a processing method is provided that obtains a complete complex optical Fourier transform result from an optical Fourier transform device including four output channels. The method includes: determining the real part of the channel output signal of each of three of the four output channels, and detecting the imaginary part of the channel output signal in the remaining output channels. The method further includes determining the complete complex result by using only the detected real parts from the three output channels and the detected imaginary parts from the remaining output channels to obtain the real and imaginary parts of all four output channels.

[0135] Obtaining the imaginary part of the channel output signal of one of the three channels may include applying logical operations to the imaginary parts detected in the remaining output channels. Alternatively or further, obtaining the real part of the channel output signal in the remaining output channels includes applying logical operations to the real part of the channel output signal detected in one of the three output channels. Alternatively or further, obtaining the imaginary parts of the channel output signals of two of the three channels includes determining that the imaginary part is zero. Alternatively or further, obtaining the real part of the channel output signal of any one of the three output channels and / or obtaining the imaginary part of the channel output signal of one of the three or any of the remaining channels includes converting the detected optical output signal level into a signed number or its digital equivalent.

[0136] More specific examples of receiver variations between different output channels include any of the following combinations:

[0137] • At least one output channel includes a receiver for recovering real data, the receiver having only three possible real output states, optionally, wherein the three possible real output states are zero, positive and negative, wherein positive and negative values ​​have the same amplitude.

[0138] • At least one output channel includes a receiver for recovering virtual data, the receiver having only three possible virtual output states, optionally, wherein the three possible virtual output states are zero, positive and negative, wherein positive and negative values ​​have the same amplitude.

[0139] • The two output channels include a receiver for recovering real data, which has only three possible real output states, optionally, wherein the three possible real output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0140] • The two output channels include a receiver for recovering virtual data, which has only three possible virtual output states, optionally, wherein the three possible virtual output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0141] • At least one output channel includes a receiver for recovering real data, which has only five possible real output states.

[0142] • The two output channels include a receiver for recovering real data, which has only five possible real output states.

[0143] For detailed examples of optimizing receiver configuration, please refer to [link / reference]. Figure 13 Describe it.

[0144] Limiting the receiver in any way described herein can lead to a simplification of OFT device construction, reduced manufacturing and component costs, and / or may result in a higher receiver operating frequency.

[0145] It should be understood that one or more methods of data packing, data unpacking, receiver optimization, and output logic derivation described in this overview can be combined to form a composite method. Attached Figure Description

[0146] This instruction is intended to be read in the context of the overview and is not intended to be limiting.

[0147] Deterministic, quantifiable OFT devices and methods

[0148] 1D OFT devices (such as integrated photonic OFT devices) can be implemented in a two-dimensional structure where light is confined in one dimension and diffracts freely in the other two. In an example of such an OFT device, the input and output are arrays of input and output ports, serving as the exit or entrance of a waveguide or optical channel. Light from each input port is diffracted into a wavefront broad enough to cover all output ports. At each output port, the contribution from each input port has the same intensity when the optical power at all input ports is the same. The angle of the wavefront at the output array is determined by the angle of the input ports and sets the phase delay at each output port.

[0149] One way to achieve this is to arrange the input and output arrays on the arcs of different circles, where the center of one circle lies on the first. The Fourier transform of the input array is formed on the line of the second circle. For a given circle radius and the number of input ports, the angular spacing between the input and output arrays can be calculated such that the output ports sample a single-order Fourier transform of the input data.

[0150] The device has a certain tolerance for the degree to which the input and output ports need to approximate an arc; small variations in radial or lateral distance (but not angle) can be compensated for. If the radius is large enough, the arc can be approximated as a straight line. "Large enough" means that the distance between the arc and its tangent at the zero-order Fourier transform edge is much smaller than the wavelength.

[0151] Figure 1aA known 1D OFT device (or Fourier transform waveguide) of this type is shown. In brief, Figure 1 illustrates a Fourier transform planar waveguide 100 in which light is constrained in one (Cartesian) dimension (z) and freely diffracted in the other two dimensions (x and y). The OFT device includes a first array 110 of input ports 111 and a second array 120 of output ports 121. The second array 120 of output ports is arranged along the Fourier plane of the Fourier transform waveguide. As those skilled in the art will understand, the Fourier plane can be a plane or a curved surface on which the Fourier transform of the light at the input ports is formed. An interference region 130 is defined in the space or volume between the first and second arrays and may include an input coupling region connecting the first input array and the diffraction region, and an output coupling region connecting the diffraction region and the second output array. The interference region 130 is surrounded by a boundary 141 in the x and y planes. Boundary 141 may be directly adjacent to interference region 130, or there may be a spatial margin between boundary 141 and at least some edges (perimeters) of interference region 130.

[0152] The interference region and any surrounding space within the boundary 141 can be formed by etching or machining blocks, plates, or wafers of material (e.g., as a relief within it, or as a freestanding block or imprinted structure) to create the interference region for light propagation from the input port. Therefore, the boundary 141 surrounding or adjacent to the interference region 130 includes side faces (such as facets of a three-dimensional plate). The interference region is constrained by faces (such as the main face of a plate) in the z-direction (such as the thickness direction of the plate). The faces surrounding or adjacent to the boundary of the interference region 130 are exposed to light emitted from the input port.

[0153] A planar waveguide is any photonic waveguide and can be implemented using any of the following techniques or combinations:

[0154] Planar waveguides in silicon-on-insulator (SOI)

[0155] • Photonic crystals (PhC) with or without periodic defects or subwavelength apertures

[0156] Planar waveguides in silicon nitride (SiN)

[0157] Waveguides in composite photonics technology (III-V or II-VI)

[0158] Any plasma waveguide

[0159] Metasurface

[0160] • or other photonic technologies

[0161] In the interference region, there may be voids, or there may be (e.g., homogeneous) media, and / or media with a uniform refractive index or designed to have a uniform refractive index.

[0162] The first array 110 is arranged on (or along) the first arc 110a of the first circle 110c, and the second array 120 is arranged on (or along) the second arc 120a of the second circle 120c offset from the first circle. Preferably, the first circle 110c has a center located on the second arc 120a, and the second circle 120c has a center located on the first arc 110a. Preferably, the center of the first circle 110c is located at or near the center of the second array 120, and the center of the second circle 120c is located at or near the center of the first array 110.

[0163] The first arc 110a and the second arc 120a respectively define the first portion 131a and the second portion 131b of the perimeter 131 of the interference region 130. The third portion 131c and the fourth portion 131d of the perimeter 131 connect the corresponding ends (or edges) of the first portion 131a and the second portion 131b. That is, the third portion 131c is a virtual surface extending in the plane between the first end of the first portion 131a and the first end of the second portion 131b, and the fourth portion 131d is a virtual surface extending in a straight line (or plane) between the second end of the first portion 131a and the second end of the second portion 131b. The first portion 131a to the fourth portion 131d constitute the entire perimeter 131 in the xy plane (input and output ports are arranged in this plane).

[0164] Input port 111 is the exit point (e.g., exit pupil or exit aperture) of input channel 101, which can also be described as a waveguide or coupling structure. That is, the input port can be connected to any or a combination of the following:

[0165] Waveguides fabricated using the same techniques, methods, and / or materials as those used in the free-space Fourier transform planar waveguide region.

[0166] • Coupling structures, such as grating couplers or edge couplers, couple light from an external source.

[0167] • Coupled structures, such as tapered couplers, couple light between vertically displaced waveguides (i.e., waveguides in the silicon layer couple light into the SiN Fourier transform planar waveguide region).

[0168] Impedance matching structure, which reduces reflections between the free-space Fourier transform planar waveguide and any of the above.

[0169] The embodiments can be used with existing photonic technologies. For example, the phase and amplitude of light in input channel 101 connected to input port 111 can be controlled or modulated using a transmitter in the input channel. The term "transmitter" as used throughout this disclosure can be any photonic modulator. It can include a light source or transmitter (provided it is coherent and each transmitter has a detectable or controllable phase relationship with other transmitters in the OFT device); a transmitter arranged to control the transmission of (coherent) light from a light source; a modulator arranged to modulate (coherent) light from a light source; or an encoder arranged to encode a signal onto a (coherent) optical signal. That is, the transmitter can be an optical transmitter, an optical encoder, an optical emitter, or an optical modulator. Including but not limited to the following or combinations thereof:

[0170] Mach-Zehnder modulator

[0171] PN modulator

[0172] Ring modulator

[0173] Thermal modulator

[0174] • Another photonic integrated circuit is arranged to modulate the optical input signal.

[0175] The term "transmitter" also includes reference in WO2023170405(A1) Figure 2 , 3 Any optical encoder described in a, 3b, 4a-g, 5a-f, 6a-e, 7a, and 7b, the disclosure of which is incorporated herein by reference. Protection is sought for OFT devices as described in this disclosure and including such optical encoders, but this disclosure is not limited to transmitters of these types.

[0176] Output port 121 can be positioned to sample one or more orders of the Fourier transform of the input data, whether as a Fast Fourier Transform or otherwise. Typically, the output port is positioned to capture or sample the zeroth-order Fourier transform; however, in some embodiments, the output port may alternatively or additionally be positioned to capture or sample higher-order Fourier transform modes, such as first-order or second-order.

[0177] The embodiment relates to sampling the Fast Fourier Transform. If sampling the Fast Fourier Transform, the number of output ports used to extract data must be at least the same as the number of input ports. The maximum number of data points is the same as the number of input ports, meaning additional ports can be used to extract other orders of the Fourier Transform, but these additional ports will contain copies of the zeroth-order data without any additional information. The number of input ports 111 can be two or more, up to a maximum of sixteen. (See reference...) Figure 1b In a particularly advantageous arrangement, the number of input ports is four.

[0178] During operation, the Fourier transform planar waveguide is illuminated with modulated or unmodulated light at one or more input ports 111.

[0179] More specifically, a light source, such as a solid-state semiconductor laser source (not shown), provides coherent light. The light source can be disposed off-chip, in which case the light is coupled into the OFT device via an optical fiber, coupled to the fiber via a grating coupler or an edge coupler using a ferrule or V-groove. Alternatively, the light source can be packaged in the same carrier substrate as the OFT device and coupled into the OFT device via an edge coupler and photonic wire bonding. Alternatively, the light source can be integrated with the OFT device and coupled into the OFT device via an edge coupler and a tapered structure (with or without photonic wire bonding).

[0180] Before being split into multiple input channels 101, the coherent light from the light source is amplified using an optical amplifier (not shown). This can be achieved by separating the coherent light from the light source off-chip and coupling the light into an optical fiber bundle that provides the input channels or acts as an input channel itself, or by using an optical fiber splitter, or by initially coupling the light into a single channel and then splitting the light into input streams using a cascaded MMI or Y-branch. If a cascaded MMI or Y-branch is used, the input streams come from the output of the last stacked or cascaded 1×2 splitter. Input channels 101 carry the input stream of light separated from the light source to input ports 111.

[0181] The inserted modulated or unmodulated light diffracts within the interference region 130. The output port 121 is placed on the Fourier plane, i.e., where the diffracted light forms an optical Fourier transform of the light pattern from the input port 111.

[0182] Input port 111 introduces light into interference region 130. The light at input port 111 carries information encoded into the phase and / or amplitude of the field. The magnitude and shape of the optical field contained in the free-space Fourier transform planar waveguide at each input port 111 will define the envelope function of the Fourier transform according to the convolution theorem.

[0183] Output port 121 is positioned in a Fourier plane (which extends along the second arc 120a). Light arrives at output port 121 as a set of waves, the angular range of which matches the relative positions of input ports 111. Waves from all input ports 111 are superimposed (e.g., coherently) to form an analog Fourier transform. The Fourier transform in the Fourier (or output) plane (along the second arc 120a) contains the complete Fourier transform of the light pattern in the input plane (along the first arc 110a), including the input pattern shape and the input envelope function.

[0184] The Fourier transform (FT) is sampled by output port 121 for detection, and the detection method then determines whether the solution of the FT is analog or digital. The light collected by output port 121 is typically detected using at least one receiver.

[0185] The term "receiver" as used throughout this disclosure includes any component arranged to detect parameters of an optical signal, particularly relative or absolute amplitude (intensity), relative or absolute phase, or both. That is, the term receiver refers to an optical receiver. The receiver can be analog (connected to an ADC) or digital. The receiver can include one or more (e.g., a pair) photodiodes, with the light detected by a single photodiode on each channel, or the signal can be mixed with a reference beam for zero-difference detection. The receiver can be a coherent detector with a 90° optical mixer for detecting the phase of the optical signal. PIN or avalanche photodiodes can be used with a transimpedance amplifier. The receiver used to extract data for detection can maintain the amplitude of the Fourier transform, as well as the phase (if the phase is detected). The term receiver also includes references to WO2023170405(A1). Figure 8 Any balance detector or combination thereof described in a or 8b, the disclosure of which is incorporated herein by reference. Protection is sought for OFT devices as described in and including such balance detector arrangements, and related methods of operating such OFT devices, but this disclosure is not limited to these types of receivers.

[0186] Apart from the first arc 110a and the second arc 120a Figure 1a The waveguide shown also includes the boundary outside the third portion 131c and the fourth portion 131d of the perimeter 131 adjacent to the interference region 130. Specifically, the waveguide of FIG1 includes a first waveguide boundary 141a and a second waveguide boundary 141b. The first waveguide boundary 141a is the surface of a plane extending between the first end of the first arc 110a and the first end of the second arc 120a. The second waveguide boundary 141a is the surface of a plane extending between the second end of the first arc 110a and the second end of the second arc 120a.

[0187] Interference region 130 is located between input port 111 (i.e., at the input plane of the OFT device) and output port 121 (i.e., at the output or Fourier plane of the OFT device). In interference region 130, light from each input port interferes to produce a diffraction pattern. Output port 121 is positioned to capture the diffraction pattern at the Fourier plane (although the diffraction pattern will be present throughout the interference region).

[0188] Each output port 121 is connected via output channel 102 to an optical amplifier (not shown), and then to a detector (which may also be described as a receiver or decoder). Each optical amplifier is arranged to amplify the corresponding output stream. The optical amplifier may be a semiconductor optical amplifier (SOA). However, the embodiment is not limited to this, and other types of optical amplifiers may be used. Alternatively, no optical amplifier may be used.

[0189] The value of the first complex element is encoded onto each input stream emitted from input port 111, which may differ between input streams, such that the value of the input function varies with the position of the input port in the first array due to the spatial variation of the input function.

[0190] Two variables can be used to define or approximate each of the input and output functions: (i) the relative positions of the ports within the array; and (ii) the complex values ​​encoded onto the optical flow passing through (e.g., entering or leaving) the ports. Therefore, it can be understood that the input and output functions are each sampled versions of continuous functions, where the sampling resolution is determined by the port aperture size and / or port spacing or center distance.

[0191] The temporal variation of the input stream can also be introduced by changing the value of the first complex element encoded into each input stream over time. The input stream can be continuous (always on, or for multiple clock cycles). Alternatively, the input stream can also be pulsed (intermittently on and off, optionally synchronized with a clock signal). The value of the first complex element encoded into the input stream can change with each clock cycle, allowing multiple optical Fourier transforms to be performed continuously frame by frame.

[0192] The detector (or receiver or decoder) is arranged to detect (or decode) the second complex element encoded into each output stream acquired by output port 121.

[0193] The decoder is configured to decode a second complex element from each output stream based on at least one feature of the respective output stream. The value of the at least one feature is detected by the decoder and converted into a form representing the second complex element. In an embodiment, the at least one feature is the phase and / or amplitude of the output stream, and the value of the phase and / or amplitude is equal to or associated with the value of the second complex element. The phase may be the phase relative to the phase of the input stream at input port 111.

[0194] Figure 1b References are shown Figure 1a The 1D OFT device is described as a special form.

[0195] exist Figure 1bIn this diagram, each input port is arranged on a first arc, and each output port is arranged on a second arc. The distance between the first and second arcs is equal to the radius of both arcs. The input ports are located at an angle θn on the second arc, measured from the zero point, and the output ports are located at an angle θn on the first arc, measured from the zero point. The zero points on the first and second arcs are directly opposite each other relative to a virtual line that bisects the common radius of the first and second arcs. That is, the zero points on the first and second arcs are located at opposite ends of their common radius. The angle θn is defined by Equation 2:

[0196] (Equation 2)

[0197] Where n is the port number, N is the total number of input or output ports, and R is the radius. It is the effective refractive index of the waveguide mode. Figure 1b In the equation, the port number n starts from the lowest port and increases to the highest port by an integer. If N is even, it is -(N / 2)+1; if N is odd, it is -((N-1))⁄2. Figure 10 The corresponding port numbers in the diagram are input 1, which corresponds to the topmost port in the diagram, and input 4, which corresponds to the bottommost port (output ports use the same convention). Effective refractive index of the waveguide mode. The characteristics of guided mode propagation in OFT devices are described and can be determined using methods known to those skilled in the art, based on the optical properties of the material used to construct the waveguide and the dimensions of the waveguide.

[0198] The OFT device in the embodiment can also adopt a design other than the reference. Figure 1b Other forms besides those described. Examples include any OFT device capable of optically performing the Discrete Fourier Transform. Such OFT devices satisfy the Discrete Fourier Transform equation (Equation 1).

[0199] Figure 2 The optical path diagram of the OFT device is shown, with reference to... Figure 1a Or as described in 1b. The OFT device includes an OFT stage 210 having an input channel 201 and an output channel 202, a grating coupler 203, a source channel 211, a splitter 220, a combiner 230, a transmitter 240, an input channel phase control 242, a monitoring channel receiver 260, a monitoring channel phase control 265, an output channel receiver 270, a reference channel phase control 280, a reference channel 281, a reference source channel 282, and a reference channel amplifier 290.

[0200] The grating coupler 203 provides an entry point for a light source (e.g., a laser), as shown in the reference. Figure 1aThe input channel 201 is branched off from a single source channel 211 and extended to the OFT device using a series of cascaded splitters 220. A transmitter (or encoder) 240 is provided in series in each input channel.

[0201] A splitter 220 is provided to separate a portion of the optical input signal from the transmitter 240 via a monitoring channel 261 before the transmitter signal reaches the OFT device 210. A combiner 230 combines the optical input signal from the monitoring channel 261 with an optical perturbation signal provided by a perturbation channel 262 and provides the combined signal to the monitoring channel receiver 260. The perturbation channel 262 includes one of the monitoring phase controls 265.

[0202] Each output channel 202 is connected between OFT stage 210 and combiner 230, which combines the optical output signal from the corresponding output channel 202 with the optical reference signal from the corresponding reference channel 281, and provides the combined optical signal to the corresponding output channel receiver 270. Each reference channel 281 includes reference channel phase control 280. Reference channels 281 are connected to reference source channels 282 via a series of cascaded splitters 220.

[0203] Each transmitter 240 includes a modulator 241 and an input channel phase control 242. Figure 2 In one example, modulator 241 is a modulator pair, and phase control 242 is a phase control pair on a separate parallel branch of input channel 201. In other examples, input channel phase control 242 is a single phase control provided in the input channel downstream of the transmitter.

[0204] In operation, modulator 241 in each input channel 201 modulates light from the light source to provide an optical input signal with an input state to OFT stage 210. Phase control 242 adjusts the input state.

[0205] The phase of the light in the perturbation channel 262 is controlled by the monitoring phase control 265 to generate an optical perturbation signal. The monitoring channel receiver 260 detects the state of the combined optical input signal and the optical perturbation signal, or detects only the input state if no perturbation light is provided in the perturbation channel.

[0206] Each output channel carries an optical output signal from OFT 210 in the corresponding output state.

[0207] The phase of the light in reference channel 281 is controlled by reference channel phase control 280 to generate an optical reference signal. Reference channel amplifier 290 controls the amplitude of the optical reference signal. Output channel receiver 270 detects the state of the combined optical output signal and optical reference signal, or detects only the output state if the optical reference signal is turned off by reference channel amplifier 290.

[0208] As used in this disclosure, the term "state" (e.g., input state or output state) refers to one or more parameters of an optical signal, or a digital or numerical representation of the parameter. A state can be understood as being or representing a value or level. A parameter may include the amplitude or intensity of the signal, or it may be the phase or relative phase of the signal. The digital or numerical form may be a complex number, or a signed or unsigned real or imaginary number, or its (bit) representation.

[0209] Figure 3 The first Algonquin diagram 301 and the second Algonquin diagram 302 are shown, each including a real axis 311 and an imaginary axis 312 and an origin 313 located at the intersection of the real axis 311 and the imaginary axis 312, which intersect at right angles. Figure 3 This is a diagram illustrating the possible values ​​at the input and output ports in an ideal system. Each port or channel, whether input or output, can have any value. The first Algonquin diagram 301 depicts, for example... Figure 1a The diagram illustrates possible input states of the OFT system (e.g., from input port 111 or array 110 of input ports). A second Argonne diagram 302 depicts possible output states of the OFT system (e.g., at output port 121 or array 120 of output ports). It is noteworthy that... Figure 3 In this context, the set of possible input and / or output states includes every complex number. That is, such an OFT system has (potentially unbounded) continuous inputs and continuous outputs. In practice, there are upper limits to the number of possible input states and the number of possible output states from the OFT system. For example, the number of possible input states can be determined by the optical amplitude in the input channel 101 connected to input port 111, and the number of possible output states can be determined by the sensitivity (range) of the receiver (or detector or decoder).

[0210] The values ​​shown on the Algonquin diagram in the attached diagram represent the mathematical inputs of the OFT device, not the amplitude or phase of light at the input or output channels or ports. The mathematical values ​​at the inputs can be set using the transmitter. The mathematical values ​​at the outputs can be read using the receiver.

[0211] In an OFT device, the amplitude (or intensity) of the light at the input channel can be any value between 0 and a limit set by the combination of the light source and the transmitter. The amplitude (or intensity) of the light at the output can be any value between 0 and a limit set by the receiver sensitivity. The phase of the light at the input or output can be any value between 0 and 2π or a multiple thereof, and can be detected using a receiver (e.g., a coherent detector with a 90° optical mixer).

[0212] same, Figure 4Two Algonquin diagrams 401 and 402 are shown, each including a real axis 411 and an imaginary axis 412, and an origin 413 located at the intersection of the real axis 411 and the imaginary axis 412, which intersect at right angles. Figure 4 This is a diagram illustrating the possible values ​​at the input and output ports in a real system. Each channel or port, whether input or output, can have a discretized value, the quantization of which is set by the system that inserts or sets the input value (i.e., the transmitter and its associated systems) and the system that reads the output value (i.e., the receiver and its associated systems).

[0213] and Figure 3 compared to, Figure 4 The set of possible input and / or output states shown is discrete. That is, the space of the complex input and / or output states is discretized by creating discrete intervals 421, 422 (e.g., corresponding to a grid partitioning overlaid on the complex plane), the union of which spans the complex plane (or its possible finite subplanes). Each discrete interval 421, 422 is a finite subplane of the complex plane, its size determined by the quantization set by the system inserting or inputting the input state and / or reading the output state. The size of the discrete intervals 421, 422 can be smaller or larger depending on the quantization of the system. Larger quantization (i.e., discretization that produces larger discrete intervals in the complex plane) can be set by the electronics and / or noise in the system. Figure 4 (and the discussion below) Figure 5 In the case of ), there is a one-to-one mapping between the quantized input / output state and the intersection or division of the grid.

[0214] Figure 5 Two Algonquin diagrams, 501 and 502, are also shown. Figure 5 The quantization ratio of the input and / or output states described in the figure. Figure 4 The larger or more obvious depiction in the text. That is to say, Figure 5 The diagram illustrates the possible values ​​at the input and output ports of a system with large quantization (differences between different values ​​or state levels). This represents a real-world analog system where large quantization is determined by the sensitivity of the electronics and the noise in the system. In other words, Figure 5 Each discrete interval 521, 522 span ratio Figure 4 Each discrete interval 421, 422 represents a larger complex plane region. Therefore, for example, a larger discretization of a finite subplane of a given complex plane produces fewer discrete intervals than a smaller discretization of the same subplane.

[0215] Reference Figure 4 and Figure 5Although all discrete intervals are described as having the same size, this is not necessarily the case. For example, the quantization of the input states can differ from the quantization of the output values, such that the size of the discrete intervals in the space of possible input values ​​can differ from the size of the discrete intervals in the space of possible output states. Similarly, the span of possible input states (i.e., the width of each discrete interval resulting from the quantization of the input states) can differ from the span of possible output states (i.e., the width of each discrete interval resulting from the quantization of the output states). That is, the quantization, dynamic range, and span of the input and output channels can be the same or different. The quantization, dynamic range, and span of each input channel can differ from that of another input channel. The quantization, dynamic range, and span of each output channel can differ from that of another output channel.

[0216] Figure 6 Similarly, two Algonquin diagrams are shown, depicting the possible input and output states of a deterministic OFT system. Figure 6 The input and output states of a deterministic optical Fourier transform device are shown. This system has a limited number of input and output ports, such as a one-dimensional OFT device with four input and four output ports, like... Figure 1b or Figure 2 As shown. In this case, the set of possible input states is restricted to the real axis, and the quantization of the input makes the width of the discrete interval 621 δ. in In particular, there exists exactly one discrete interval 621 (because the size of the set of possible input states is exactly the same as the width δ of the discrete interval). in (consistent), and therefore there exist exactly two quantized input channel states 631 and 632, which can be represented in binary form as real values ​​0 and 1. Due to the restrictions on possible input states, the set of possible output states is also restricted. In particular, in Figure 6 In the OFT system, there are 27 possible output states: nine of them have an imaginary part +bi, nine have an imaginary part 0i for some constant b, and nine have an imaginary part −bi. The real part of each of the 27 possible output states takes one of nine values. Therefore, the output of the OFT system is discretized on the real axis at nine levels 641 and on the imaginary axis at three levels 642; thus, there are 27 possible quantized output states. The quantization of the output states results in a discretization interval 622 with a width of δ. out .

[0217] Figure 7 Four pairs of Algonquin diagrams, 701, 702, 703, and 704, are shown. Each diagram depicts a possible input or output state at a specific input or output port among the four input ports and four output ports in a deterministic OFT system. Figure 7 and Figure 6The difference lies in that it breaks down the possible input and output states of each input and output channel in the OFT device. The output states are restricted such that each output channel receives only... Figure 6 The outputs shown represent a smaller subset of the overall dynamic range of the OFT outputs; and all channels have the same quantization step size (∂out). This reduces the requirement for receiver resolution, and all ports use the same receiver circuitry with appropriate offsets or adjustments applied. These output modes are possible by limiting the input states to only a few levels (here, two) with quantization (∂in), and also by limiting the inputs to unsigned real or unsigned imaginary numbers.

[0218] The first pair of 701 Algonquin plots depicts the possible input and output states at the first input and first output ports, respectively; the second pair of 702 plots depicts the possible input and output states at the second input and second output ports, respectively; and the third and fourth pairs of 703 and 704 follow the same pattern. In each of the four cases, the set of possible input states is restricted to the real axis, and therefore the corresponding set of possible output values ​​is also restricted. In particular, in each of the four cases, the size of the set of possible input values ​​is exactly equal to the width δ of the discrete interval. in Consistent. Therefore, quantization produces exactly one discrete interval, and thus for each of the four input ports, there are exactly two quantized input channel states 731, 732, which can be represented in binary form as real values ​​0 and 1.

[0219] The outputs of the first and third output ports are discretized on five levels 741 and 742 on the real axis (where the five levels 741 of the first output port overlap with but are different from the five levels 742 of the third output port), and also discretized on the same level on the imaginary axis. The outputs of the second and fourth output ports are discretized on three identical levels 743 on the real axis and on three identical levels 744 on the imaginary axis. Therefore, the outputs of the four output ports are discretized on seven unique levels on the real axis and three unique levels 744 on the imaginary axis.

[0220] Figure 8 Sixteen figures are shown, depicting the real part of the output state of an OFT system with four input ports and four output ports. Figure 8The input and output states at each port are shown in an exemplary deterministic OFT system with four inputs and four outputs, where the inputs are restricted to unsigned binary real numbers. The input and output states are obtained from a simulation of a one-dimensional OFT device, where each of the four input ports is sequentially illuminated with an unsigned binary real number input. The results for the real part of the output are compared with the expected results of an FFT algorithm (with cyclic shift) for all 16 input combinations (4 bits).

[0221] exist Figure 8 In this context, the set of possible input states is restricted to the real axis, and specifically, the size of the possible input set is exactly the same as the width of the discrete interval (therefore there are exactly two quantized input channel states, represented in binary form as real values ​​0 or 1), as shown in the reference. Figure 7 The system described above. For example, Figure 801 shows that when the four input ports are input with 0, 0, 0, 0 respectively, the real parts of the output states at the four output ports are 0, 0, 0, 0 respectively. As another example, Figure 802 shows that when the four input ports are input with 1, 1, 1, 1 respectively, the real parts of the output states at the four output ports are 0, 0, 4, 0 respectively. Only port 3 needs to be offset.

[0222] Figure 9 Similarly, sixteen figures are shown depicting the imaginary part of the output value of an OFT system with four input ports and four output ports, where the set of possible input states is restricted to the real axis, and in particular, the size of the possible input set is exactly the same as the width of the discrete interval (therefore there are exactly two quantized input channel states, represented in binary form as real values ​​0 or 1), as shown in the figure. Figure 7 The system compares the imaginary part of the output with the expected result of an FFT algorithm (with cyclic shift) for all 16 input combinations (4 bits). For example, Figure 901 shows the imaginary parts of the output states at the four output ports as 0i, 0i, 0i, 0i when the four input ports are 0, 0, 0, 0 respectively. As another example, Figure 902 shows the imaginary parts of the output states at the four output ports as 0i, −1i, 0i, +1i when the four input ports are 1, 1, 0, 0 respectively.

[0223] exist Figure 8 and Figure 9 In the diagram, the quantization state is shown as a dashed line and is consistent for all input ports and all input combinations. The quantization used to recover the real and imaginary parts of the OFT output is the same for all input combinations and all output ports.

[0224] Figure 6 , 7Figures 8 and 9 illustrate an OFT device with four input channels and four output channels, where the input states are binary, and the OFT output is discretized at seven levels on the real axis and three levels on the imaginary axis. A receiver capable of discerning only five levels is sufficient to recover all OFT data (including both real and imaginary parts). Further performance improvements are possible by simplifying the receiver arrangement as follows:

[0225] • Use two three-level receivers to recover the real data;

[0226] • Use two three-level receivers to recover the imaginary part of the data;

[0227] • Recover the real data using two five-level receivers; and

[0228] • Remove two receivers that were originally used to recover the imaginary part data from output channels 1 and 3 (because the imaginary value is always 0 for all drive conditions and does not need to be measured).

[0229] Figure 10 A table is shown that describes all possible input-output combinations of an OFT system with four input ports and four output ports, where each input port has exactly two quantized input states (which can be represented as real values ​​0 or 1), as shown in the reference. Figure 7 , 8 The system described in 9. For example, the first line 1001 shows that when the four input ports are input with (real values) 0, 0, 0, 0, the real parts of the four output ports are 0, 0, 0, 0, and the imaginary parts are 0i, 0i, 0i, 0i, 0i, respectively. As another example, the thirteenth line 1002 shows that when the four input ports are input with (real values) 1, 1, 0, 0, the real parts of the four output ports are 0, -1, 2, -1, and the imaginary parts are 0i, -1i, 0i, +1i, respectively.

[0230] Figure 11a The components of an Optical Fourier Transform (OFT) device or system 1100 are shown. The OFT system 1100 includes a host 1101, a data and control interface 1111, a calibration and monitoring controller 1121, an input digital-to-analog converter 1131, an input analog signal conditioner 1141, a photonic system 1151, an output analog signal conditioner 1161, and an output analog-to-digital converter 1171.

[0231] The application logic 1101a of host 1101 generates data to be processed by OFT system 1100. The data is stored in transmitter (“TX”) data frame memory 1101b and then sent to transmitter data packing logic 1101c. Transmitter data packing logic 1101c can determine whether certain data is allowed to be further processed by OFT system 1100. For example, transmitter data packing logic 1101c can determine that an input data frame consisting of all zeros should not be further processed by OFT system 1100. As another example, transmitter data packing logic 1101c can determine that an input data frame consisting of bits 0101 or 1010 should not be further processed by OFT system 1100. Transmitter data packing logic 1101c can also serialize the input data before further transmission through OFT system 1100. That is, transmitter data packing logic 1101c can, for example, convert a 64-bit input into 64 individual 1-bit inputs. Once the data is processed by the transmitter data packaging logic 1101c, it is sent to the TX queue memory 1101d, and then leaves the host 1101 and enters the data and control interface 1111.

[0232] The control system 1111a of the data and control interface 1111 determines, based on the results obtained from the calibration and monitoring controller 1121, whether the input data should be further transmitted through the OFT system 1100, or whether the further transmission of the input data should be delayed. The input data selected for further transmission through the OFT system 1100 then leaves the data and control interface 1111 and enters the input digital-to-analog converter 1131.

[0233] The input digital-to-analog converter 1131 includes a digital equalizer 1131a, a data encoder 1131b, a pre-emphasis stage 1131c, and a digital-to-analog converter 1131d, as is common in the art. The digital equalizer 1131a corrects for any errors that may accumulate in the input data during transmission from the host 1101 to the input digital-to-analog converter 1131. The data encoder 1131b encodes the input data into a format recognizable by the OFT system 1100. The data is then converted from digital to analog by the digital-to-analog converter 1131d, and then leaves the input digital-to-analog converter 1131 and enters the input analog signal conditioner 1141.

[0234] The input analog signal conditioner 1141 includes a pre-driver 1141a, a driver 1141b, and an amplifier chain 1141c, as is common in the art. The input analog signal conditioner 1141 amplifies the analog signal received from the input digital-to-analog converter 1131, making the signal compatible with the photonic encoder 1151a of the photonic system 1151. The amplifier chain 1141c may include, for example, a gain module or an amplifier. The amplified or conditioned signal leaves the amplifier chain 1141c and enters the photonic encoder 1151a of the photonic system 1151.

[0235] The photonic encoder 1151a of the photonic system 1151 converts the received analog electrical signal into a photonic signal. An optical Fourier transform stage 1151b is applied to the photonic signal, and the resulting photonic signal is received by the photonic receiver 1151c. The photonic receiver 1151c transmits the corresponding photocurrent to the transimpedance amplifier 1161a of the output analog signal conditioner 1161.

[0236] The output analog signal conditioner 1161 includes a transimpedance amplifier 1161a, an amplifier chain 1161b, and a filter stage 1161c. The transimpedance amplifier 1161a amplifies the received photocurrent and converts it into a voltage. The amplifier chain 1161b and the filter stage 1161c condition the signal in preparation for digitization by the output analog-to-digital converter 1171. The output analog signal conditioner 1161 transmits the conditioned signal from the filter stage 1161c to the analog equalizer 1171a of the output analog-to-digital converter 1171.

[0237] The output analog-to-digital converter 1171 includes an analog equalizer 1171a, an aligner 1171b, a sampler 1171c, and an analog-to-digital converter 1171d, as is common in the art. The purpose of the output analog-to-digital converter 1171 is to digitize the raw input voltage from the output analog signal conditioner 1161. The analog equalizer 1171a may apply, for example, continuous-time linear equalization (CTLE). The analog equalizer 1171a modifies the raw input voltage from the output analog signal conditioner 1161 in such a way that it is compatible with other components of the output analog-to-digital converter 1171, such as the aligner 1171b or the analog-to-digital converter 1171d. The aligner 1171b may introduce a delay or other fine-grained time characteristic into the signal, aligning the signal to a specific point. This point is then sampled or extracted by the sampler 1171c and then transmitted to the analog-to-digital converter 1171d for digitization. Therefore, the output analog-to-digital converter does not necessarily digitize the entire original input voltage, but may only digitize points or some points selected by the aligner 1171b and / or sampler 1171c. The digitized signal is then transmitted from the analog-to-digital converter 1171d to the output data receiver 1111b of the data and control interface 1111.

[0238] The output data receiver 1111b of the data and control interface 1111 transmits the received data to the host 1101, whereby the received data enters the receiver queue memory 1101e. The receiver detection and recovery logic (or "deserializer") 1101f of the host 1101 can apply a depackaging method to the data received from the receiver queue memory 1101e to, for example, recover complete complex data from PAM4 data. The receiver detection and recovery logic 1101f is applied to (possibly multiple) data frames, and its output is then transmitted to the receiver data frame memory 1101g. Finally, the data from the receiver data frame memory 1101g is consumed by the application logic 1101a.

[0239] Figure 11b An enlarged view 1102 of the host 1101 of the OFT system 1100 is shown. (Compared to...) Figure 11a Similarly, host 1101 includes transmitter data frame memory 1101b, receiver data frame memory 1101g, transmitter data packing logic 1101c, receiver detection and recovery logic 1101f, transmitter queue memory 1101d, and receiver queue memory 1101e.

[0240] Data is stored in the transmitter data frame memory 1101b and then sent to the transmitter data packetization logic 1101c. The transmitter data packetization logic 1101c can determine whether certain data is allowed to be further processed by the OFT system 1100. For example, in... Figure 11bIn the case where the transmitter data packing logic 1101c checks in step S1103 whether the data received from the transmitter data frame memory 1101b consists entirely of zero bits; if so, the input data is copied and then directly transmitted to the receiver data frame memory 1101g. That is, when the input data consists entirely of zero bits, the data is simply copied to the receiver data frame memory 1101g without further processing by other components of the OFT system 1100, such as the input and output analog signal conditioners 1141, 1161 or the photonic system 1151. When the input data does not consist entirely of zero bits, then in step S1104, it checks whether the signal mode of the data is in the form of 1010; if so, in step S1105, the input data is inverted and copied, and then sent to the transmitter queue memory 1101d. Therefore, in step S1105, the input (1010) is converted to (

[0101]

[0101] ), i.e., two different input frames, each containing a 4-bit input vector. If step S1104 determines that the signal pattern of the input data is not in (1010) form, the data is directly transmitted to the transmitter queue memory 1101d. In either case, the (possibly transformed) data is then sent to the data and control interface 1111, from where it is processed by other components of the OFT system 1100, and then returned to the data and control interface 1111 and sent to the receiver queue memory 1101e. Specifically, data that does not satisfy any of the checks in steps S1103 and S1104—that is, data that is not all zero bits and whose signal pattern is not in (1010) form—enters the transmitter queue memory 1101d without modification.

[0241] When data is received from the data and control interface 1111 into the receiver queue memory 1101e, it is sent for processing by the receiver detection and recovery logic 1101f. Figure 11b In this case, each data frame from receiver queue memory 1101e includes four channels, each channel being at one of four possible levels or states. Steps S1106 and S1107 determine whether the output data frame corresponds to input data that has undergone inversion and copying by steps S1104 and S1105 of transmitter data packing logic 1101c. If so, the output of one of the two different inputs is determined using a known relationship between the outputs of OFT system 1100 generated by two different inputs (e.g., inputs (0101) and (1010)).

[0242] In step S1106, each received data frame is checked to determine whether the first channel of the data frame is at a first level, the second channel at a second level, the third channel at a third level, and the fourth channel at a second level. For a given data frame, if it is determined that this is not the case, the data frame is directly transmitted for processing by the receiver detection and recovery logic 1101f. Otherwise (i.e., if it is determined that the channel of the data frame is at the level specified above), the data frame can be modified according to the known relationship between the outputs of the OFT system 1100 given two different inputs. Specifically, in step S1107, the contents of the four channels stored in the temporary storage 1102a are checked to determine whether they are at the same level as described above (i.e., the first channel, second channel, third channel, and fourth channel are at the first level, second level, third level, and second level, respectively). If so, the data received from the receiver queue memory 1101e is modified as follows: the first channel is set to the fourth level, while the other three channels remain unchanged. If the check in step S1107 fails, the data received from the receiver queue memory 1101e is not modified. In either case, the transmitted (modified or unmodified) data is processed by the receiver detection and recovery logic 1101f and sent from there to the receiver data frame memory 1101g.

[0243] OFT data packaging

[0244] Figure 12 This is a flowchart of an example method 1200 for processing the output signal from an optical Fourier transform device. Method 1200 can be used in... Figure 11a The method 1200 may be implemented in an optical Fourier transform system 1100 or another suitable system or apparatus. For example, the method 1200 may be implemented as a set of computer-readable instructions stored on a transient or non-transient computer-readable medium and executed by one or more processors.

[0245] Method 1200 begins at step S1202, wherein an input vector is applied to the optical Fourier transform device (e.g., to...). Figure 11a The optical Fourier transform system 1100. In step S1204, an output vector representing the optical Fourier transform of the input vector is obtained. The output vector represents the optical Fourier transform of the input vector, for example, through... Figure 11a A photonic system 1151 is obtained from the optical Fourier transform system 1100, and method 1200 is implemented on the optical Fourier transform system.

[0246] At step S1206, the channel of the output vector is modified to form a PAM4 signal. The channel of the output vector may include more than four levels. For example, the channel may carry a PAM5 signal. When the channel includes more than four levels, at step S1206, the modification of the channel includes reducing the number of levels in the channel to form a PAM4 signal. For example, if the channel carries a PAM5 signal, the number of levels in the channel is reduced by one to form a PAM4 signal. The number of levels in the channel can be reduced by detecting the highest level of the channel and removing it from the channel. For example, the fifth level of the channel carrying the PAM5 signal can be removed to form a PAM4 signal. Removing one or more levels from the channel can be performed by saturating or clipping the highest level to the fourth level, or by subtracting one or more levels to form the fourth level.

[0247] Alternatively, the channel may include fewer than four levels. For example, the channel may carry a PAM3 signal. If the channel includes fewer than four levels, at step S1206, modifying the channel includes adding one or more levels to form a PAM4 signal. For example, one or two levels may be added to form a fourth level.

[0248] Once a PAM4 signal is formed in step S1206, optionally, in step S1208, the PAM4 signal can be transmitted via a PAM4 link. For example, the PAM4 signal can be transmitted from host 1101 to [other device / location] via a PAM4 link. Figure 11a The OFT system 1100 has a data and control interface 1111 on which method 1200 can be implemented. In step S1210, the correct complete complex OFT data is recovered from the PAM4 signal by modifying the inverted channel. For example, the correct complete complex OFT data can be recovered from the PAM4 signal by deriving the fifth level of the PAM5 signal from the PAM4 signal. Steps S1208 and S1210 are optional; either one can be performed, both can be performed, or neither can be performed. Therefore, for example, method 1200 may end immediately after executing step S1206 (in which case neither steps S1208 nor S1210 are executed); or method 1200 may execute step S1208 after step S1206 and then end (in which case step S1208 is executed but S1210 is not); or method 1200 may execute step S1210 after step S1206 and then end (in which case step S1210 is executed but S1208 is not); or method 1200 may execute step S1206, then execute step S1208, then execute step S1210, and then end.

[0249] Receiver optimization

[0250] Figure 13 This is a flowchart of an example method 1300 for obtaining a complete complex optical Fourier transform result from an optical Fourier transform device. Method 1300 can be used in... Figure 11a The method 1300 may be implemented in an optical Fourier transform system 1100 or another suitable system or apparatus. For example, the method 1300 may be implemented as a set of computer-readable instructions stored on a transient or non-transient computer-readable medium and executed by one or more processors.

[0251] Method 1300 begins at step S1302, wherein the real part of the channel output signal in each of the first, second, and third output channels of the optical Fourier transform device is determined. For example, method 1300 may determine at step S1302 that the first, second, and third output channels are at a first level, a second level, and a third level, respectively, in which case the corresponding real parts may be real numbers -2, 0, and 2, respectively. The real number range corresponding to the level of the first output channel can be from -2 to +2, the real number range corresponding to the level of the second output channel can be from -1 to +1, and the real number range corresponding to the level of the third output channel can be from 0 to 4.

[0252] At step S1304, method 1300 detects the imaginary part of the channel output signal in the fourth output channel of the optical Fourier transform device, wherein the fourth output channel is different from the first output channel, the second output channel, and the third output channel. For example, method 1300 may determine at step S1304 that the fourth output channel is at one of the first level, the second level, or the third level, in which case the corresponding imaginary part may be the imaginary number −1i, 0i, or +1i, respectively.

[0253] At step S1306, the method determines the complete complex number result by obtaining the real and imaginary parts of all four output channels using only the detected real parts from the first, second, and third output channels and the detected imaginary part from the fourth output channel. For example, obtaining the imaginary part of the channel output signal in the second output channel may include applying a logical operation to the imaginary part detected in the fourth output channel. Similarly, obtaining the real part of the channel output signal in the fourth output channel may include applying a logical operation to the real part of the channel output signal detected in the second output channel. As another example, obtaining the imaginary parts of the channel output signals in the first and third output channels may include determining that the imaginary part is zero. As yet another example, obtaining the real part of the channel output signal of any one of the first to third output channels and / or obtaining the imaginary part of the channel output signal of any one of the second or fourth channels may include converting the detected optical output signal level into a signed number or its digital equivalent. Specifically, for example, the level of the first detected optical output signal can be converted to a signed number 1, the level of the second detected optical output signal can be converted to a signed number 2, and so on.

[0254] Logical derivation of the output

[0255] Figure 14 This is a flowchart of an example method 1400 for processing data using a Fourier transform device. Method 1400 can be used in... Figure 11a The method 1400 may be implemented in an optical Fourier transform system 1100 or another suitable system or apparatus. For example, the method 1400 may be implemented as a set of computer-readable instructions stored on a transient or non-transient computer-readable medium and executed by one or more processors.

[0256] Method 1400 begins at step S1402, wherein an input data stream is applied to the Fourier transform device. The input data stream is applied in a series of sequential frames, each frame including an input vector. For example, the input vector of the first frame may be of the form 0001, and the input vector of the next frame in the series of sequential frames may be of the form 0010. The input data stream and / or the series of sequential frames of the input data stream may be stored, for example... Figure 11aThe optical Fourier transform system 1100 is stored in the transmitter data frame memory 1101b. In step S1404, an output data stream is generated from the input data stream by determining an output vector for each frame, where each output vector is a Fourier transform of the corresponding input vector in the input data stream; for example, for the input vector (0010), the output vector can be a Fourier transform of the input vector, i.e., (1111). In step S1406, for a first input vector having a first value, a corresponding first output vector is determined by performing an optical Fourier transform on the first input vector. In step S1408, for a second input vector having a second value different from the first value, a corresponding second output vector is determined by performing a Fourier transform on the second input vector, in addition to performing a Fourier transform on the second input vector. For example, the corresponding second output vector can be determined by applying a logical operation to the second input vector. If the second input vector has a total value of zero, the second output vector can be determined by copying the second input vector. The corresponding second output vector can also be determined alternatively or additionally by the optical Fourier transform of the first input vector, or by a known relationship between different vector values ​​(e.g., sign reversal) from the first output vector.

[0257] Data unpacking

[0258] Figure 15 This is a flowchart of an example method 1500 for processing data using a Fourier transform device. Method 1500 can be used in... Figure 11a The method 1500 can be implemented in an optical Fourier transform system 1100 or another suitable system or apparatus. For example, the method 1500 can be implemented as a set of computer-readable instructions stored on a transient or non-transient computer-readable medium and executed by one or more processors. The method 1500 can be viewed as... Figure 12 The inverse of method 1200. That is to say, Figure 12 Method 1200 is a data packing method in which the output of the Fourier transform device is modified to be compatible with the PAM4 data link, while Figure 15 Method 1500 includes unpacking the PAM4 data to recover the complete Fourier transform result.

[0259] Method 1500 begins at step S1502, wherein the output signal includes a PAM4 signal received from the optical Fourier transform device. At step S1504, the complete complex output of the optical Fourier transform device is recovered from the output signal. Recovery may include decoding the PAM4 signal, which may further include deriving a PAM5 signal or a PAM3 signal from the PAM4 signal.

[0260] Figure 16This is a flowchart of an example method 1600 for operating an optical Fourier transform device or system comprising N input channels and K output channels. Method 1600 can be used in... Figure 11a The method 1600 may be implemented in an optical Fourier transform system 1100 or another suitable system or apparatus. For example, the method 1600 may be implemented as a set of computer-readable instructions stored on a transient or non-transient computer-readable medium and executed by one or more processors.

[0261] Method 1600 begins at step S1602, where the light arriving at the input channels of the optical Fourier transform device or system is quantized to form a quantized output at each output channel. For example, the light emitted from each input channel can be limited to a finite number of possible input states (e.g., limited to two, three, or four possible input states). Then, at step S1604, the light is applied... Figure 12 , 13 Any combination of at least one of methods 1200, 1300, 1400, and 1500.

[0262] Truth table

[0263] Figure 17 It is truth table 1700, which shows the optical Fourier transform device (e.g. Figure 11a An example of the input-output combination of the optical Fourier transform device 1100. The truth table 1700 includes six groups of columns 1710, 1711a, 1711b, 1712, 1713a, and 1713b, each group including four columns.

[0264] The real parts of all possible 4-bit input combinations are displayed in the first four columns 1710 of the truth table 1700, where each bit of each possible input corresponds to the corresponding input channel of the optical Fourier transform device. The next four columns 1711a display the real parts of the raw or unprocessed output data of the optical Fourier transform device when the corresponding 4 bits from the first four columns 1710 are input to the system. For example, the truth table 1700 shows that when

[0000] is input, the (real part) of the raw or unprocessed output of the optical Fourier transform device is

[3212] . The output represented in the second set of columns 1711a can be obtained, for example, from the optical Fourier transform stage 1151b of the optical Fourier transform device. Similarly, the third set of four columns 1711b displays the imaginary parts of the raw or unprocessed output of the optical Fourier transform device when the corresponding 4 bits from the first four columns 1710 are input to the system. For example, truth table 1700 shows that, at input

[0000] , the imaginary part of the raw or unprocessed output of the optical Fourier transform device is at a second level at the second and fourth output channels. The output represented in the third set of columns 1711b can be obtained, for example, from the optical Fourier transform stage 1151b of the optical Fourier transform device.

[0265] Depending on the configuration of the optical Fourier transform device, the raw output data may be inherently incompatible with PAM4. The raw output data may be in a format other than PAM4, due to the number of possible output states that can be received at the output channel of the OFT device. For example, if the number of possible output states is 5, then the raw output data can be said to be (or includes) a PAM5 signal (with 5 possible states, levels, or values). Therefore, the optical Fourier transform device can modify the raw output data to make it compatible with PAM4. Column 1712 of the fourth group of truth table 1700 describes the result of such modification of the raw output data in columns 1711a and 1711b of the second and third groups. Typically, this modification can be performed as follows:

[0266] • Detect the fifth level of the PAM5 signal in the first and third output channels.

[0267] • Subtract the fifth level from the PAM5 signal to generate the PAM4 signal.

[0268] • Add two levels to the PAM3 signal of the second output channel.

[0269] As an example, line 11, 1720 of truth table 1700 shows that when input

[1010] is received, the optical Fourier transform device outputs real part levels 5, 2, 3, and 2 in the first output channel, the second output channel, the third output channel, and the fourth output channel, respectively, and outputs imaginary part level 2 in the second output channel and the fourth output channel. Therefore, the fifth level 1708 of the PAM5 signal in the first output channel is detected, and the fifth level 1708 is subtracted from the PAM5 signal to generate the PAM4 signal 1701, and two levels are added to the PAM3 signal 1702 in the second output channel.

[0270] An optical Fourier transform device can recover the original output data from a PAM4 signal by inverting the aforementioned modifications to the output signal. Therefore, specifically, an optical Fourier transform device can recover correct, complete complex OFT data from a PAM4 signal by inverting the modifications to the output signal. In the case of removing one or more levels from the original output data to form the PAM4 signal, the correct, complete complex OFT data includes deriving one or more removed levels from the PAM4 signal. For example, in the case of removing a fifth level from the original output data to form the PAM4 signal, the correct, complete complex OFT data includes deriving the fifth level of the PAM5 signal from the PAM4 signal.

[0271] Columns 1713a and 1713b in the fifth and sixth groups of truth table 1700 represent the results of recovering correct, complete complex OFT data from a PAM4 signal, which is represented in column 1712 of the fourth group. It is noteworthy that no information is lost during the formation of or recovery of the PAM4 signal; that is, the recovered data (the "processed output" in the figure) is identical to the corresponding original output data. Typically, the real part of the original output data (e.g., correct, complete complex OFT data) can be recovered from the PAM4 signal using the following method:

[0272] • Adjusting the real part of the first output channel of the PAM4 signal to achieve:

[0273] • If both the second and third output channels of the PAM4 signal are equal to the fourth level, then it is equal to the third level; or

[0274] • If the second output channel of the PAM4 signal is equal to

[0275] If the fourth level and the third output channel is not equal to the fourth level, then it is equal to its current level plus an additional level; or

[0276] • Equal to its current level.

[0277] • Adjust the real part of the second output channel of the PAM4 signal to achieve:

[0278] • If the second output channel of the PAM4 signal is equal to the fourth level, then it is equal to the second level; or

[0279] • Equal to its current level.

[0280] • Adjust the real part of the third output channel of the PAM4 signal to achieve:

[0281] • If both the second and first output channels of the PAM4 signal are equal to the fourth level, then it is equal to the third level; or

[0282] • If the second output channel of the PAM4 signal is equal to the fourth level, then it is equal to its current level plus an additional level; or

[0283] • Equal to its current level.

[0284] • The real part of the fourth output channel of the data recovered from the PAM4 signal is equal to the real part of the fourth output channel of the original output data.

[0285] The imaginary part of the original output data (e.g., correct complete complex OFT data) can be recovered from the PAM4 signal using the following method:

[0286] • The imaginary part of the first output channel of the data recovered from the PAM4 signal is equal to the second level.

[0287] • The imaginary part of the second output channel of the data recovered from the PAM4 signal:

[0288] • If the imaginary part of the fourth output channel of the PAM4 signal is equal to the first level, then it is equal to the third level; or

[0289] • If the imaginary part of the fourth output channel of the PAM4 signal is equal to the third level, then it is equal to the first level; otherwise...

[0290] • Equal to the second level.

[0291] • The imaginary part of the third output channel of the data recovered from the PAM4 signal is equal to the second level.

[0292] • The imaginary part of the fourth output channel of the data recovered from the PAM4 signal:

[0293] • If the imaginary part of the fourth output channel of the PAM4 signal is equal to the first level, then it is equal to the first level; or

[0294] • If the imaginary part of the fourth output channel of the PAM4 signal is equal to the third level, then it is equal to the third level; otherwise...

[0295] • Equal to the second level.

[0296] The imaginary parts of the first and third output channels are always zero (i.e., level 2 in level 1, 2, and 3).

[0297] For example, line 11 (1720) of truth table 1700 shows that when input

[1010] is received, the optical Fourier transform device outputs real part levels 5 1708, 2, 3, and 2 in the first output channel, the second output channel, the third output channel, and the fourth output channel, respectively, and outputs imaginary part level 2 in the second output channel and the fourth output channel. The corresponding PAM4 signal formed according to the above description includes the first 1701, the second 1702, the third 1703, and the fourth output channel being at the fourth level, the fourth level, the third level, and the second level, respectively.

[0298] The following is an example of recovering the original output data (e.g., correct complete complex OFT data) from a PAM4 signal, limited to line 11, 1720 of truth table 1700. Since the real part of the second output channel 1702 of the PAM4 signal is equal to the fourth level, and the real part of the third output channel 1703 of the PAM4 signal is not equal to the fourth level, the real part of the first output channel 1704 of the PAM4 signal is adjusted to be equal to its current level plus an additional level (according to the second sub-point of the first point above). Therefore, specifically, the real part of the first output channel 1704 of the PAM4 signal is adjusted to be equal to the fifth level. Since the real part of the second output channel 1702 of the PAM4 signal is equal to the fourth level, the real part of the second output channel 1702 of the PAM4 signal is adjusted to be equal to the second level (according to the first sub-point of the second point above). In this particular case, the real part of the third output channel 1703 of the PAM4 signal is not adjusted (according to the third sub-point of the third point above), and thus remains equal to the third level. The real part of the fourth output channel of the data recovered from the PAM4 signal is equal to the real part of the fourth output channel of the original output data (according to point four above), and therefore specifically equal to the second level. As for the imaginary part of the recovered original output data, the imaginary parts of the first and third output channels of the data recovered from the PAM4 signal are equal to the second level (according to points one and three above, respectively). Since the imaginary part of the fourth output channel of the PAM4 signal is neither equal to the first level nor the third level, the imaginary part of each of the second and fourth output channels of the data recovered from the PAM4 signal is equal to the second level (according to the third sub-point of point two and the third sub-point of point four above, respectively).

[0299] Figure 18 Similarly, truth table 1800 shows an optical Fourier transform device (e.g., Figure 11a An example of the input-output combination of the optical Fourier transform device 1100. The truth table 1800 includes six groups of columns 1810, 1811a, 1811b, 1812, 1813a, and 1813b, each group including four columns.

[0300] and Figure 17 The truth table is the same as 1700. Figure 18 The first four columns 1810 of the truth table 1800 show the real parts of all possible 4-bit input combinations, where each bit of each possible input corresponds to the corresponding input channel of the optical Fourier transform device. The next four columns 1811a show the real parts of the raw or unprocessed output data of the optical Fourier transform device when the corresponding 4 bits from the first four columns 1810 are input to the system. Similarly, the third set of four columns 1811b shows the imaginary parts of the raw or unprocessed output of the optical Fourier transform device when the corresponding 4 bits from the first four columns 1810 are input to the system.

[0301] For reference Figure 17 The optical Fourier transform device can modify the original output data to make it compatible with PAM4. The fourth column 1812 of truth table 1800 describes the result of this modification of the original output data in the second and third columns 1181a and 1811b. (Refer to reference...) Figure 17 Compared to the modification of the original output data, Figure 18 The original output data in the background is modified as follows:

[0302] • For both the first and third output channels, the highest signal level (e.g., the fifth level) of the original output data is saturated to the second high signal level (e.g., the fourth level).

[0303] Columns 1813a and 1813b in the fifth and sixth groups of truth table 1800 represent the results of recovering correct, complete complex OFT data from the PAM4 signal, which is represented in column 1812 of the fourth group. It is noteworthy that no information is lost during the formation of the PAM4 signal or the recovery of data from it; that is, the recovered data (the "processed output" in the figure) is identical to the corresponding original output data. Due to the aforementioned saturation, the real part of the original output data (e.g., correct, complete complex OFT data) can be recovered from the PAM4 signal, which is greater than... Figure 17 The description is simpler, as follows:

[0304] • Adjust the real part of the first output channel of the PAM4 signal to achieve:

[0305] • If the real part of the first output channel of the PAM4 signal is equal to the fourth level and the real part of the third output channel of the PAM4 signal is equal to the third level, then it is equal to the fifth level; otherwise…

[0306] • Equal to its current level.

[0307] • The real part of the second output channel of the data recovered from the PAM4 signal is equal to its current level.

[0308] • Adjust the real part of the third output channel of the PAM4 signal to achieve:

[0309] • If the real part of the first output channel of the PAM4 signal is equal to the third level and the real part of the third output channel of the PAM4 signal is equal to the fourth level, then it is equal to the fifth level; otherwise…

[0310] • Equal to its current level.

[0311] • The real part of the fourth output channel of the data recovered from the PAM4 signal is equal to the real part of the fourth output channel of the original output data.

[0312] With Figure 17In the same manner, the imaginary part of the original output data (e.g., the correct complete complex OFT data) can be recovered from the PAM4 signal.

[0313] Figure 19 Similarly, truth table 1900 shows an optical Fourier transform device (e.g., Figure 11a An example of the input-output combination of the optical Fourier transform device 1100. The truth table 1900 includes six groups of columns 1910, 1911a, 1911b, 1912, 1913a, and 1913b, each group including four columns.

[0314] and Figure 17 and Figure 18 The truth table is the same. Figure 19 The first four columns 1910 of the truth table 1900 show the real parts of the possible 4-bit input combinations, where each bit of each possible input corresponds to the corresponding input channel of the optical Fourier transform device. The next four columns 1911a show the real parts of the raw or unprocessed output data of the optical Fourier transform device when the corresponding 4 bits from the first four columns 1910 are input to the system. Similarly, the third set of four columns 1911b shows the imaginary parts of the raw or unprocessed output of the optical Fourier transform device when the corresponding 4 bits from the first four columns 1910 are input to the system.

[0315] like Figure 17 and Figure 18 The optical Fourier transform device can modify the original output data to make it compatible with PAM4. The fourth column 1912 of truth table 1900 describes the result of this modification to the original output data in the second and third columns 1191a and 1911b. Figure 17 and Figure 18 Compared to the modification of the original output data, the following procedure shall be followed. Figure 19 Modification of the original output data in the context of the background:

[0316] • For the first output channel, saturate the highest signal level (e.g., the fifth level) of the original output data to the second high signal level (e.g., the fourth level).

[0317] It is worth noting that, compared to Figure 18 In this context, the signal levels of both the first and third output channels are saturated. Figure 19 In this context, only the signal level of the first output channel is saturated.

[0318] Columns 1913a and 1913b in the fifth and sixth groups of truth table 1900 represent the results of recovering correct, complete complex OFT data from the PAM4 signal, which is represented in column 1912 of the fourth group. It is noteworthy that no information is lost during the formation of the PAM4 signal or the recovery of data from it; that is, the recovered data ("Processed Data" in the figure) is identical to the corresponding original output data. Due to the aforementioned saturation results, and... Figure 17 Compared to the method described in 18, it is easier to recover the real part of the original output data (e.g., the correct complete complex OFT data) from the PAM4 signal, as follows:

[0319] • Adjust the real part of the first output channel of the PAM4 signal as follows:

[0320] • If the real part of the first channel output of the PAM4 signal is equal to the fourth level and the real part of the third channel output of the PAM4 signal is equal to the third level, then it is equal to the fifth level; otherwise...

[0321] • Equal to its current level.

[0322] • The real parts of the second, third, and fourth output channels of the data recovered from the PAM4 signal are equal to the real parts of the second, third, and fourth output channels of the original output data, respectively.

[0323] With Figure 17 and Figure 18 In exactly the same way, the imaginary part of the original output data (e.g., the correct complete complex OFT data) can be recovered from the PAM4 signal.

[0324] Figure 20 Similarly, truth table 2000 shows the optical Fourier transform device (e.g., Figure 11a An example of the input-output combination of the optical Fourier transform device 1100. The truth table 2000 includes six groups of columns 2010, 2011a, 2011b, 2012, 2013a, and 2013b, each group including four columns.

[0325] and Figure 17 , 18 Similar to the truth table of 19, Figure 20The first four columns 2010 of the truth table 2000 show the real parts of the possible 4-bit input combinations, where each bit of each possible input corresponds to the corresponding input channel of the optical Fourier transform device. The next four columns 2011a show the real parts of the raw or unprocessed output data of the optical Fourier transform device when the corresponding 4 bits from the first four columns 2010 are input to the system. Similarly, the third set of four columns 2011b shows the imaginary parts of the raw or unprocessed output of the optical Fourier transform device when the corresponding 4 bits from the first four columns 2010 are input to the system.

[0326] Optical Fourier transform devices can be referenced Figure 17 , 18 The original output data can be modified in the manner described in 19 to make it compatible with PAM4. Similarly, it can be referred to Figure 17 , 18 The real and imaginary parts of the original output data (e.g., correct complete complex OFT data) are recovered from the PAM4 signal in any of the methods described in 19.

[0327] Figure 20 Two rows in the sixteenth row of the truth table 2000, namely the first row 2001 and the eleventh row 2003, are omitted or blacked out, describing the input truncation form that can be formed by the optical Fourier transform device. The first row 2001 corresponds to the input vector

[0000] , and the eleventh row 2003 corresponds to the input vector

[1010] . The optical Fourier transform device always outputs

[0000] when the input is

[0000] , therefore it is not necessary to perform an optical Fourier transform stage (e.g., ...). Figure 11a The optical Fourier transform stage 1151b of the optical Fourier transform device 1100 and any other related processes. Instead, the output vector

[0000] can be simply copied from the input vector

[0000] . For example, the input vector

[0000] can be directly copied to... Figure 11aThe receiver data frame memory 1101g of the optical Fourier transform device 1100. Line 11 2003 corresponds to the input vector

[1010] . There is a known relationship between the outputs of the optical Fourier transform device on inputs

[1010] and

[0101] . In particular, for the two inputs

[1010] and

[0101] , the output of each output channel except the first output channel is the same. Therefore, on input

[1010] , the output of each output channel except the first output channel can be copied from the corresponding output channel of the optical Fourier transform device on input

[0101] . Similarly, on input

[0101] , the output of each output channel except the first output channel can be copied from the corresponding output channel of the optical Fourier transform device on input

[1010] . On input

[1010] , the output of the first output channel can be derived by inverting the output of the first output channel of the optical Fourier transform device on input

[0101] . For example, if the output of the first output channel at input

[0101] is equal to the first level, then it can be determined that the output of the first output channel at input

[1010] is equal to the fifth level. Similarly, if the output of the first output channel at input

[0101] is equal to the second level, then it can be determined that the output of the first output channel at input

[1010] is equal to the fourth level. Likewise, the output of the first output channel at input

[0101] can be derived from the output of the first output channel of the optical Fourier transform device at input

[1010] .

[0328] OFT Monitoring and Calibration

[0329] Figure 21 A flowchart depicting a method for operating an OFT device is shown. The OFT device, for example, is any OFT device described throughout this disclosure, including references... Figure 1b and Figure 2 The OFT devices described.

[0330] The method includes: in a first step S2102, applying an input signal to at least one component of an OFT device; in a second step S2104, detecting a first value of a parameter of an optical signal generated by the input signal; and in a third step S2106, determining a first relationship between the detected first value and a first reference value of the parameter.

[0331] The input signal can be an electronic input signal applied to one or more input channels of the OFT device to produce the optical output of the input channel, or an optical input signal applied to the OFT device via one or more input channels (i.e., the optical output produced by the input channel is input to the OFT device via the input channel). At least one component can be one or more of the following: one or more input ports, input channels and (optical) modulators (or transmitters) in the input channels, free space regions of the OFT device, output channels of the OFT device, or receivers in the output channels. The optical signal can be an optical input signal detected from one or more input channels, such as an optical input signal detected in a monitoring channel connected to one or more input channels of the OFT device. The optical signal at one or more input channels is detected (or measured or sampled) using a monitoring receiver connected to the input channel (e.g., in the monitoring channel). The optical signal is an optical output signal detected at one or more output channels of the OFT device (i.e., detected at the output channel using one or more output channel receivers of the OFT device). When at least one component is a free space region of the OFT device, the optical signal is an optical output signal detected at the output channel of the OFT device (e.g., at the receiver). The monitoring channel is directly connected to one or more input channels, i.e., not connected via a free space region.

[0332] Control OFT device

[0333] The method may further include: in step S2108, controlling the OFT device based on a first relationship by using a state machine and state machine-based data processing control. This control may be directed to… Figure 2 Control of any component in the example, including input channel phase control and perturbation channel phase control (both used to influence the input channel state), or reference channel phase control (used to influence the output state detected at the receiver).

[0334] The methods for monitoring signal integrity and the state machine-based data processing control can be applied to any OFT device (apparatus) described herein. The described methods are particularly suitable for OFT devices with at least two input channels, optionally multiples of two or four as described herein, such as the quaternary binary OFT devices and systems described herein.

[0335] One or more procedures (or algorithms) or methods are applied to an OFT device to ensure the accuracy of OFT results, such as using calibration steps or fixed procedures, such as applying a known calibration input pattern (or frame encapsulation) when the data stream stops or pauses or before the data stream is about to start. The method also involves performing a time-averaged measurement of the background signal level during data processing. In both examples, the method includes using a state machine arranged to monitor and control the OFT device, wherein the signal level is compared to a known (or expected) reading for each calibration input pattern or a known (or expected) background signal level. The one or more procedures and methods may also include statistical analysis, such as eye diagram monitoring, histograms, and bit error rate.

[0336] state machine

[0337] Figure 11a The OFT system 1100 includes a calibration and monitoring controller 1121, which is responsible for any processes and methods related to OFT device calibration. The calibration and monitoring controller 1121 includes a state machine, which further includes signal conditioning logic, a data path, and a host controller. The state machine reads a set of inputs and changes to different states based on these inputs. The signal conditioning logic manipulates signals in a certain way to prepare them for the next processing stage. The data path is used in conjunction with the state machine to perform operations on the data passing through this set of inputs. The host controller controls the data flow from the host and communicates with the host via a host control unit having the calibration and monitoring controller 1121.

[0338] The input monitor of the photonic system applies an input signal to the state machine via a transimpedance amplifier, one or more filters, and an analog-to-digital converter (ADC). The input monitor can be a monitoring channel connected to one or more input channels of the OFT device, or it can receive signals from a monitoring receiver in such a monitoring channel. At the state machine, a first value of a parameter of the optical signal generated by the input signal is detected, and a first relationship is determined between the detected first value and a reference value of the parameter. A lookup table (e.g., in the state machine or accessible by the state machine) is used. Figure 25The reference value is determined either by a lookup table provided in the document for the calibration procedure described herein, or by pre-calculated firmware settings or pre-calculated parameters stored in local memory, and the first relationship is the difference between the first value and the first reference value exceeding a threshold difference. If the first relationship is determined, the input data path selector control determines, based on the first relationship, whether to pause, reapply, or (re)calibrate the OFT device. If the data stream is paused, the host control sends a message to host 1101 to pause the input data stream. The input signal at the state machine can then be reapplied via the digital data generator. To (re)calibrate the OFT device, the input signal settings are applied to the input controller via the phase values ​​of the DAC and driver. The input controller may be a reference... Figure 2 The input channel phase control or disturbance channel phase control is one of the aforementioned phase control (or can be arranged to control those components).

[0339] In a similar manner, the output monitor of the photonic system applies the output signal to the state machine via a transimpedance amplifier, one or more filters, and an ADC. The output monitor can be one or more output channels of the OFT device, or a receiver connected to one or more output channels. At the state machine, a first value of a parameter of the optical signal generated by the output signal is detected, and a first relationship is determined between the detected first value and a reference value of the parameter. This reference value is determined using a lookup table, or via pre-computed firmware settings of the state machine or pre-computed parameters stored in local memory, and the first relationship is the difference between the first value and the first reference value exceeding a threshold difference. If the first relationship is determined, the output data path selector control determines, based on the first relationship, whether to pause, reapply, or (re)calibrate the OFT device. If the data stream is paused, the host control sends a message to host 1101 to pause the input data stream. The input signal at the state machine can then be reapplied via the host. To (re)calibrate the OFT device, the output signal, based on the output data received at the digital data receiver, sets the phase value via a DAC and a driver to be applied to the output controller. The output controller can be a reference... Figure 2 The one or more reference channel phase controllers (or may be arranged to control those components).

[0340] The host control can be used in conjunction with input and output data path selector control.

[0341] Methods for controlling OFT devices

[0342] A first embodiment (or first calibration method) is described, wherein the first reference value introduced in step S2106 is theoretically derived from the input signal and can be determined based on the discrete Fourier transform of the input vector. This discrete Fourier transform forms a theoretical output vector based on the input vector. This discrete Fourier transform is theoretical in that it is not performed optically via an OFT device, but rather through other means, such as numerical, digital, or electronically only.

[0343] A second embodiment (or second calibration method) is described, wherein the first reference value introduced in step S2106 is determined using the expected background signal level (e.g., time-averaged signal level) of one or more channels when the data stream passes through the channel and / or the disturbance signal applied to the monitored channel. The first and second embodiments can be combined. The first and / or second embodiments can be performed using an OFT device, for example, referring to... Figure 1a , 1b Or OFT as described in 2.

[0344] As described above, in order to address the discrepancy between the signal intended to be output from the input channel or optical modulator and the signal actually output from them, and / or the discrepancy between the signal read at the output channel and the actual signal at the output channel, the control OFT device as described in the first and second embodiments may include one or more of the following: calibration, restart or reset and recalibrate, control data flow, calculate signal integrity, and / or change the parameters of the PIC.

[0345] When controlling the OFT device involves calibrating, restarting, resetting, and recalibrating the OFT device, this method (re)calibrates the OFT device, or more specifically, the OFT PIC, to provide known (or theoretical, predicted, or expected) signal performance at the output channel, which is determined based on the input signal pattern or the background signal level of one or more signals as the data stream passes through the channel. The method includes determining whether a first value of a parameter of the optical signal is within an acceptable range or threshold associated with a first reference value for that parameter, determined by the known or expected signal performance.

[0346] If a first value of a parameter exceeds a threshold associated with a first reference value, the state machine can take action to control components of the OFT device, for example, to prevent the first value from exceeding the threshold. That is, the state machine can change parameters in the PIC to correct for deviations and ensure signal integrity. The parameters of the optical signal can be amplitude (or intensity) or phase (or state). The first reference value of the parameter is the expected value of the parameter given an input signal (or data input) and can be determined using a lookup table or derived from the first value of the parameter at some output channel. In practice, controlling the OFT device based on this comparison includes modifying the operating parameters of at least one component based on the first relationship, such that the optical signal generated by the input signal is modified.

[0347] Modifying the operating parameters means controlling those parameters to alter the light signal detected by the PIC. (See reference...) Figure 2 An example of this is modifying phase control in a reference channel, perturbation channel, or input channel (e.g., controlling the phase of light by changing the temperature of the heaters therein) to control the corresponding signals in those channels.

[0348] First embodiment (first calibration type - input signal pattern)

[0349] Calibration Pattern

[0350] In this embodiment, the input signal pattern is a known calibration input pattern, which is a binary signal value, or one or more binary numbers applied to each of one or more input channels of the OFT device in calibration mode. In a quaternary binary OFT device (a device having four input channels and four output channels, wherein the input state at each input channel is binary), there are four or sixteen 4-bit input vectors. Each binary number (input vector) is represented by a set of binary signal values ​​(input states). This set is applied at each of one or more input channels in a single frame.

[0351] For example, the input signal pattern involves alternately turning on each of one or more input channels to form a diagonal matrix comprising four 4-bit binary numbers, i.e., the individual input channels [0 0 0 1], [0 0 1 0], [0 1 00], and [1 0 0 0] are turned on in turn (in any order). In this way, one or more input channels can be referred to as a diagonal-calibrated input pattern. Thus, the input signal pattern is a diagonal form that represents an input vector with only a single "1" in its binary representation. The calibration pattern also includes: cyclically cycling all 16 possible combinations of the four 4-bit binary numbers using a quaternary binary OFT. This means that the input channels are sequentially turned off and on to produce all 16 possible combinations. That is, each input combination [0 0 0 0], [0 0 0 1], [0 0 1 0], [0 0 1 1], [0 1 0 0], [0 1 01], [0 1 1 0], [0 1 1 1], [1 0 0 0], [1 0 0 1], [1 0 1 0], [1 0 1 1], [1 1 0 0], [1 10 1], [1 1 1 0], and [1 1 1 1] are turned on sequentially (in any order). However, this disclosure is not limited to this. For example, alternatively, a selected subset of all possible combinations may be applied. Thus, the input signal pattern in calibration mode represents an input vector having any combination of 1s and 0s in 4-bit binary form.

[0352] Reboot or reset and recalibrate

[0353] Figure 22 A schematic diagram of the calibration procedure according to the first embodiment is shown. Figure 22 In the process, after the initial step S2202 of setting the orthogonality of the modulator (transmitter), there is a second step S2204 of setting the amplitude of the LO line. From there, the method proceeds to a third step S2206 of setting the LO phase to calibrate the output receiver, and then to a fourth step S2208 of setting the input phase of the remaining input channels. There is also a fifth step S2210 of optimizing the input and output phases by iterating through all possible input vectors.

[0354] In the initial step S2202, any one of one or more input channels forming a diagonal matrix comprising four 4-bit binary numbers is turned on, i.e., one of input channels [0 0 0 1], [0 0 1 0], [0 1 0 0], and [1 0 00]. The phase of the turned-on input channel or the selected input channel is set as the reference phase for the rest of the system. That is, the input channel phase of one of the selected input channels is calibrated to a known phase in the system. This can be done using, for example... Figure 2The apparatus is set by applying a perturbation signal to the perturbation channel, wherein the phase of the perturbation signal is known. The perturbation signal and the input channel signal are combined and then detected at a monitoring receiver. The perturbation channel phase control is altered or scanned, and the amplitude at the monitoring receiver changes accordingly. Alternatively, the input channel phase control is scanned to achieve the same effect. In either case, the phase is scanned until a maximum amplitude is detected at the monitoring receiver, indicating that the input channel phase is set to have the same phase as the perturbation signal. The minimum amplitude at the monitoring receiver indicates that the input channel has been set to be completely out of phase with the perturbation signal. The determined phase of the input channel sets a reference phase for the remaining calibration procedures. This can be repeated for all other input channels, but such repetition is not required.

[0355] In the second step S2204, after phase shifting, the amplitude of the local oscillator of the OFT device (i.e., the light source or reference light signal of the OFT device used for the OFT operation itself) is set to the maximum value.

[0356] In the third step S2206, the selected input channel is set to 1, such that the input vector is one of the diagonal input vectors. Then, phase control is varied within a certain phase range, affecting the signal received at the receiver in each of one or more output channels of the OFT device; for example, the phase control is varied from 0 or changed to π radians. In other words, the phase is scanned within a certain phase range by phase control. Phase control is applied to a reference channel combined with the output channel to feed the receiver, such as a reference... Figure 2 As described. The one or more output channels can output an optical output signal. As the phase control changes, a first value of the parameter detected at the receiver also changes, and a first relationship is determined between the detected first value and a first reference value. More specifically, the detected first value is compared with a first reference value of the parameter, where the first reference value is the expected value of the parameter given an input signal (or data input) and can be determined using a lookup table or derived from the first value of the parameter at some output channel. In this case, given a particular input vector, the first reference value is the maximum, minimum, or zero amplitude received at the receiver for all possible phases controlled by the phase control.

[0357] Phase control is set or fixed to a phase that exists when a detected first value equals or matches a first reference value associated with any of the input signal patterns (i.e., associated with any one of the one or more input channels forming a diagonal matrix). For example, when the first reference value is determined using a lookup table, phase control is set such that the detected first value equals the first reference value of the lookup table, which is associated with any one of the four 4-bit input vectors of the diagonal calibration input pattern. In this way, all four output receivers are calibrated based on the diagonal input vectors.

[0358] More generally, a first reference value is determined based on a discrete Fourier transform of an input signal pattern including the input signal, wherein the discrete Fourier transform is determined by a lookup table that maps each form of the input signal pattern to a corresponding theoretical output signal pattern to obtain the first reference signal.

[0359] In the fourth step S2208, the phase of each of the other inputs of the OFT device (i.e., one or more inputs excluding one of the initially activated input channels) is calibrated. The phase at each input channel varies within a certain phase range; for example, the phase of one or more input channels varies or changes from 0 to 2π radians at each input channel. In other words, another diagonal input vector is applied to the input channels such that the second input channel has a value of 1, while the other input channels are set to zero (e.g., the second diagonal input vector represented by [0 0 1 0]), including the input channels that set the reference phase calibrated in the initial step S2202. The phase of the second input channel is then changed until the output receiver detects all expected values ​​in the truth table corresponding to the respective diagonal vector. The process is repeated for the third and fourth input channels in the same manner, where the diagonal vectors are represented by [0 1 0 0] and [1 0 0 0], respectively.

[0360] In the fifth step S2210, the input signal pattern applied to the OFT device also includes causing all 16 possible combinations of the four 4-bit input vectors to cycle through the quaternary binary OFT in sequence. This means that the input channels are turned off and on (in any order) to generate all 16 combinations as described above in sequence.

[0361] During this process, the phase control of each of one or more input and output channels of the OFT device can be adjusted to ensure that the expected output signal is received for some or all input vectors. That is, the phase of each of one or more input channels or reference channels of the OFT device is set or fixed in a phase control that exists when a detected first value equals or matches a first reference value associated with a corresponding input signal pattern (i.e., associated with one or more corresponding input channels forming 16 input combinations). For example... Figure 25 The lookup in the table represents the 4-bit input vector of the input channel and the corresponding output vector of the output channel. Step S2210 allows for further phase changes if necessary and can be implemented as a final calibration check rather than a calibration step. Compared to the phase changes described in step S2206, these phase changes can be small adjustments or relatively minor alterations. For example, these changes can improve the eye opening of PAM3 and PAM5 eye diagrams.

[0362] Control data stream (including phase data stream)

[0363] To monitor and control the data flow, one or more, a subset, or all of the input vectors representing the discrete Fourier transform of the input signal pattern are introduced into the input data flow, which includes the input signal, as reference values ​​or markers. The input vectors of the input signal pattern can be in diagonal form or one of 16 possible input combinations. Using this method, the state machine can control the data flow based on the signal integrity level in the quaternary binary OFT PIC, pausing the data flow through the quaternary binary OFT PIC when the signal integrity exceeds a threshold set in the algorithm, and recalibrating / resetting the signal integrity level as described above. The data flow resumes after the reset and / or after the threshold is not exceeded or after the signal integrity is within the threshold.

[0364] Back Figure 23 It shows a depiction of a calibration OFT device (e.g. Figure 1a , 1bA flowchart of a method using the OFT device (as described in any one of steps 2) compares an output optical signal (or the value of a parameter thereof) with a reference (expected) value. Step S2302 includes determining whether the difference between a detected first value based on the output optical signal and the corresponding reference value exceeds a threshold difference. In this case, the OFT device is controlled based on a first relationship, and the data flow through the OFT device is also controlled based on the first relationship. At step S2304, if there is no difference between the detected first value based on the output optical signal and the corresponding reference value (or the difference does not exceed the threshold difference), an input signal (normal data flow) is continuously applied to at least one component (step S2304). In step S2306, any deviations in the input signal can then be corrected to minimize the difference between the detected first value based on the output optical signal and the corresponding reference value.

[0365] Step S2308 determines that when the threshold difference is exceeded, the input signal is reapplied to at least one component. In other words, the data that caused the threshold difference violation is reapplied to check again whether there is a persistent error in the system, or whether the error is only transient. The method includes further detecting a second value of the parameters of the output optical signal and determining a second relationship between the detected second value and a first reference value. When the second relationship is that the difference between the second value and the first reference value does not exceed the threshold difference, the method continues to step S2304. That is, the method continues to apply the input signal to at least one component (continue the data stream) without further repetition. At step S2312, when the second relationship is that the difference between the second value and the first reference value exceeds the threshold difference, the method also includes pausing the input signal (i.e., pausing the data stream), and referencing Figure 22 The described one, several, or all of the recalibration steps.

[0366] Second calibration type – background signal

[0367] This (second) embodiment can be used independently or in combination with the first embodiment, and will be referred to Figure 24 Describe it. Figure 24 A flowchart is shown in which the signal level is compared with the background signal level or the applied slowly changing disturbance or reference signal.

[0368] calibration

[0369] Return to reference Figure 21 The second embodiment is a method comprising: in a first step S2102, applying an input signal to at least one component of an OFT device; in a second step S2104, detecting a first value of a parameter of an optical signal generated by the input signal; and in a third step S2106, determining a first relationship between the detected first value and a first reference value of the parameter.

[0370] However, in this case, refer to Figure 2 The input signal is a slowly varying perturbation (or reference) signal applied via a perturbation channel. At least one component of the OFT device is a perturbation channel or monitoring receiver corresponding to at least one input channel of the OFT device, and the optical signal generated by the input signal is a monitoring signal, which includes a combination of the sampled optical input signal and the perturbation signal applied to the perturbation channel. That is, at the monitoring receiver, the perturbation signal is mixed with the sampled optical input signal. The perturbation signal includes a known time-varying phase change and introduces a known (or fixed or predetermined) time-varying change at the monitoring receiver. This change can be linear or sinusoidal in time, or any other known time-varying change.

[0371] Because the second embodiment samples the optical input signal in at least one input channel of the OFT device, the conditions or state of the input signal applied to at least one input channel of the OFT device can be checked. The optical input signal in any one or more input channels can be sampled by a monitoring receiver. The monitoring channel can be arranged to be shared by one or more (e.g., all of one or more input channels) such that the monitoring channel is arranged to sample the optical input signal of one or more input channels of the OFT device. The monitoring channel can also be arranged to be shared by two of the one or more input channels.

[0372] Since the perturbation signal has a known phase, if the sampled optical input signal is coherent with the perturbation signal (i.e., the sampled optical input signal is in phase with the perturbation signal), a maximum signal level (or intensity) is detected in the monitoring channel due to constructive interference. If the sampled optical input signal is incoherent with the perturbation signal (i.e., the sampled optical input signal is out of phase with the perturbation signal), a minimum signal level (or intensity) is detected in the monitoring channel due to destructive interference. The minimum signal can be any value equal to or greater than 0, and the maximum signal can be any value greater than the minimum signal. Therefore, the phase change of the perturbation signal over time will cause the amplitude detected at the monitoring receiver to change over time. This change should not deviate from the known or expected change from the perturbation signal, because the time average of the input state at the input channel is likely known and constant. If there is a deviation from the expected change, this can indicate a phase drift or amplitude drift in the signal from the input channel. If there is an amplitude drift, the difference between the perturbation signal and the input channel signal will increase or decrease accordingly, and this can be detected by the monitoring receiver. If this increase or decrease is above a threshold amount, this can indicate that the system needs to be recalibrated. If phase drift exists, the gradient (differential) of this change will vary relative to the expected value within any given sampling period. For example, if the perturbation signal changes slowly over time in the form of a sine wave, the mixed signal will follow the same shape of change. Within any given sampling period, the sine wave will have a gradient that depends on the timing of the sampling period. Since the timing of the perturbation signal's change is known, a difference in the gradient of the signal detected at the monitoring channel within a sampling period (e.g., a phase difference exceeding a threshold amount) indicates unacceptable drift, and subsequently, recalibration of the OFT device may be necessary.

[0373] Reboot or reset and recalibrate

[0374] Figure 24 The diagram depicts a calibration OFT device (e.g.) Figure 1a , 1b A flowchart of a method using an OFT device (as described in any one of steps S2 or S2). In step S2402, the monitoring channel is located in the OFT device (such as...). Figure 2 The monitoring signal is sampled in at least one input channel (as shown), the monitoring signal comprising a combination of the sampled optical input signal and a disturbance signal applied to the monitoring channel.

[0375] In step S2404, the phase of each input channel in the OFT device sharing the monitoring channel varies within a certain phase range. For example, the phase of each input channel in the OFT device sharing the monitoring channel varies or changes from 0 to π radians at each input channel. In other words, the phase is scanned on the input channels of the shared monitoring channel. For each input channel sharing the same monitoring channel, the input phase is fixed when the strength of the signal detected at the monitoring channel is at its maximum value.

[0376] In one example, such as Figure 2 As shown, the first and second outputs share the first monitoring channel, and the third and fourth outputs share the second monitoring channel. Therefore, the first and second inputs are scanned at [1 1 0 0], associated with the first monitoring channel, and the third and fourth outputs are scanned at [0 0 1 1], associated with the second monitoring channel. During phase scanning, the input phase is fixed when the signal corresponding to each input vector channel is at its maximum value.

[0377] As a third step S2406, steps S2204 and S2206 are applied to set the amplitude of the local oscillator of the OFT device (i.e., the OFT device itself or the OFT device's reference signal) to its maximum value. The phase of each of one or more output channels of the OFT device varies within a certain phase range; for example, the phase control of one or more output channels varies or changes from 0 to π radians at each of the one or more output channels. In other words, phase control is performed via output channel scanning.

[0378] As described above regarding step S2406, the one or more output channels can output an optical output signal. As the phase control of the output channel changes, a first value of the detected optical output signal parameter also changes, and a first relationship is determined between the detected first value and a first reference value. More specifically, the detected first value is compared with a first reference value of the parameter, where the first reference value is the expected value of the parameter given an input signal (or data input) and can be determined using a lookup table or derived from the first value of the parameter at some output channel. The first reference value may represent the value of the optical parameter to be compared with the optical parameter value defined by the first value. Alternatively, the first reference value may represent the value of a numerical value (e.g., a binary value) to be compared with a numerical value (e.g., a binary value) defined by the first value.

[0379] More specifically, the first reference value is determined based on the Discrete Fourier Transform (DFT) of the input signal pattern, which includes the input signal. The DFT is determined via a lookup table that maps each form of the input signal pattern to a corresponding theoretical output signal pattern to obtain the first reference signal. The first reference value is a maximum value, a minimum value, or 0, such as... Figure 25 As shown. Since the maximum and minimum signal levels are found using a perturbation signal, therefore, based on the data from... Figure 25 The first reference value corresponding to each output is set to the maximum, minimum, or 0 for each output.

[0380] In other words, the phase is set or fixed to a phase that exists when the detected first value is equal to or matches a first reference value associated with any input signal pattern (i.e., associated with any one of one or more input channels forming a diagonal matrix).

[0381] Returning to step S2408, a known intensity variation is then introduced into the input channels where the associated monitoring channels are not shared. For example, since the first output and the second output share the first monitoring channel, an intensity variation can be introduced into the third or fourth input channel, and for the example where the third output and the fourth output share the second monitoring channel, an intensity variation can be introduced into the first or second input channel. In practice, the first monitoring set associated with the first monitoring channel includes input / output channels corresponding to the binary numbers [1 1 1 0] or [1 1 0 1], and the second monitoring set associated with the second monitoring channel includes input / output channels corresponding to the binary numbers [1 0 1 1] or [0 1 1 1].

[0382] In step S2410, the phase of each of one or more input channels of the OFT device varies within a certain phase range; for example, the phase of each of the one or more input channels varies or changes from 0 to 2π at each of one or more output channels. In other words, the phase is scanned on one or more input channels. The one or more output channels can output an optical output signal. As the phase changes, a first value of the parameter of the detected optical signal also changes, and a first relationship between the detected first value and a first reference value is determined. More specifically, the detected first value is compared with a first reference value of the parameter, wherein the first reference value is the expected value of the parameter given an input signal (or data input) and can be determined using a lookup table or derived from the first value of the parameter at some output channel. The phase of each of the other inputs of the OFT device is set or fixed to a phase that exists when the detected first value is equal to or matches the first reference value, which is associated with the corresponding input signal pattern (i.e., the input vector associated with the corresponding one or more input channels forming a diagonal matrix). For example, when using a lookup table to determine a first reference value associated with a first input channel, the phase of the first input channel is set such that the detected first value is equal to the first reference value of the lookup table associated with the 4-bit binary number corresponding to the input vector.

[0383] Due to the (re)calibration process, when the phase of the scan through the input channel equals the phase of the perturbation signal, if the maximum signal level is detected, at least one component or input channel outputs the correct value. This can be used to calibrate an optical modulator or check if the optical modulator is working correctly.

[0384] Calculate signal integrity (compared to a known calibration input signal pattern).

[0385] At the output channel, if the time-averaged signal is determined based on the integral of the output channel over a period of time, an average output vector can be found. This output vector results in the third output having a maximum signal level, while all other channels (i.e., those associated with the first, second, and fourth outputs) have a signal level of 0. If the average output deviates from this theoretical output over time, the system is recalibrated as described herein.

[0386] The method may further include controlling a receiver in the output channel based on a first relation, the receiver being arranged to detect parameters of the optical signal, wherein controlling the receiver based on the first relation includes controlling the phase of the receiver based on the first relation.

[0387] It should be understood that any monitoring and calibration method can be performed on any OFT device or PIC described herein, including reference devices. Figure 1a , 1b And those described in 2, and using Figure 11a The system or other equivalent processor, controller or system is used for control.

[0388] The advantages of the OFT apparatus and method described herein may include:

[0389] • Improve signal integrity in quaternary binary OFT devices.

[0390] • Control the data processing pipeline based on the signal integrity status, where the data processing pipeline is a meaningful digital data stream that needs to be processed by a quaternary binary OFT device.

[0391] • Control the stability of the modulator device in a quaternary binary OFT device.

[0392] • Control the stability of OFT devices in quaternary binary OFT devices.

[0393] • Control the stability of the input and output channels (or waveguides) connected to OFT devices.

[0394] • Simplify the dynamic range of receiver or decoder electronics.

[0395] • Discretization of OFT results is achieved by allowing only a subset of all possible input and output combinations to be processed by the driving circuit.

[0396] • Allows the construction of truth tables that describe combinations of operations for OFT devices or PICs.

[0397] • Allows for the design and construction of effective calibration procedures to ensure accurate OFT calculations throughout the device's lifespan.

[0398] • Eliminate extensive startup procedures, simplify the system's automatic calibration process, and streamline stability circuitry.

[0399] • Simplify receiver hardware, configuration, or use, thereby reducing manufacturing costs and complexity, and / or reducing operational complexity, and / or increasing data throughput of optical Fourier transform devices.

[0400] • Using OFT devices (such as quaternary-binary OFT devices) with PAM4-compatible transceiver ports makes it easier or more efficient to inline OFT devices in, for example, data transmission systems, data center links, transceiver ports, communication links, or computing systems or any combination thereof.

[0401] • Compatibility of OFT devices (such as quaternary binary OFT devices) and systems with existing backplane infrastructure, communication backends, ports, and legacy technologies that use PAM4 signals.

[0402] • Increase the data throughput or bandwidth of the Fourier transform device or system.

[0403] Any method described in this disclosure may be implemented or performed on one or more controllers (e.g., one or more data controllers), one or more processors, computer systems, data transmission systems, or a combination thereof that are connected to and configured to control the OFT device. Figure 11a An example of a controller is provided, however this disclosure is not limited thereto.

[0404] A system (e.g., a computing system, a data transmission system, or a combination thereof) is provided, comprising: an optical Fourier transform device including N input channels and K output channels; and one or more processors, wherein the one or more processors are configured to perform any of the methods described in full herein, or wherein the system includes one or more transient or non-transient computer-readable media storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations including any of the methods described in full herein. Figure 11a An example of a controller is provided, however this disclosure is not limited thereto.

[0405] Also provided are one or more tangible or intangible, transient or non-transient computer-readable media storing computer-readable instructions that, when executed by one or more processors or controllers, cause one or more processors or controllers to perform operations, including any methods described in full in this disclosure. Figure 11a An example of a controller is provided, however this disclosure is not limited thereto.

[0406] Those skilled in the art will understand that various modifications and alterations can be made to the above disclosure without departing from its broad inventive concept. Some of these have been discussed above, while others will be obvious to those skilled in the art. Therefore, it should be understood that the invention is not limited to the specific embodiments disclosed, but is intended to cover modifications falling within the concept and scope of this disclosure, as set forth in the appended claims.

[0407] This article also describes the following numbered items:

[0408] Project 1. An optical Fourier transform device, comprising:

[0409] N input channels;

[0410] K output channels;

[0411] Each input channel includes a modulator, transmitter, or encoder, which is arranged to cause the input channel to transmit a quantized input signal.

[0412] Each output channel includes a receiver, which is arranged to detect the quantized state at the output channel.

[0413] Project 2. An optical Fourier transform device according to any of the preceding projects, wherein N is an integer multiple of 2, optionally an integer multiple of 4, optionally 4.

[0414] Project 3. An optical Fourier transform device according to any of the preceding projects, where K equals N.

[0415] Project 4. An optical Fourier transform device according to any of the preceding projects, wherein the receiver in each output channel has 12 or fewer effective bits (ENoB), optionally 10 or fewer, optionally 8 or fewer, optionally 6 or fewer, optionally 4 or fewer, optionally 3 or fewer, optionally 2 or fewer.

[0416] Project 5. An optical Fourier transform device according to any of the preceding projects, wherein the receiver in each output channel has an effective number of bits (ENoB) equal to log2(P+Q), where P is the total number of possible output states at the corresponding output channel given possible input states, and Q is 0, 1, 2, 3, 4, 6, 8 or 10.

[0417] Project 6. An optical Fourier transform device according to any of the preceding projects, wherein each output channel has a receiver different from the receiver of at least one other output channel.

[0418] Project 7. An optical Fourier transform device according to any of the preceding projects, wherein at least one output channel includes a receiver for recovering only real data, and another output channel includes a receiver for recovering only imaginary data.

[0419] Project 8. An optical Fourier transform device according to any of the preceding projects, wherein three output channels include receivers for recovering only real data, and another output channel includes receivers for recovering only imaginary data.

[0420] Project 9. An optical Fourier transform device according to any of the preceding projects, wherein the ENOB of the receiver of one output channel is different from the ENOB of the receiver of the other output channel.

[0421] Project 10. An optical Fourier transform device according to any of the preceding projects, wherein each output channel has only a single receiver.

[0422] Project 11. An optical Fourier transform apparatus according to any of the preceding projects, wherein at least one output channel includes a receiver for recovering real data, the receiver having only three possible real output states, optionally, wherein the three possible real output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0423] Project 12. An optical Fourier transform apparatus according to any of the preceding projects, wherein at least one output channel includes a receiver for recovering virtual data, the receiver having only three possible virtual output states, optionally, wherein the three possible virtual output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0424] Project 13. An optical Fourier transform device according to any of the preceding projects, wherein the two output channels include a receiver for recovering real data, the receiver having only three possible real output states, optionally, wherein the three possible real output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0425] Project 14. An optical Fourier transform device according to any of the preceding projects, wherein two output channels include a receiver for recovering virtual data, the receiver having only three possible virtual output states, optionally, wherein the three possible virtual output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0426] Project 15. An optical Fourier transform device according to any of the preceding projects, wherein at least one output channel includes a receiver for recovering real data, the receiver having only five possible real output states.

[0427] Project 16. An optical Fourier transform device according to any of the preceding projects, wherein the two output channels include a receiver for recovering real data, the receiver having only five possible real output states.

[0428] Project 17. An optical Fourier transform device according to any of the preceding projects, wherein the quantized input signal includes a finite number of possible input states.

[0429] Project 18. An optical Fourier transform device according to any of the preceding projects, wherein the number of possible input states for each input channel is 2, 3 or 4.

[0430] Project 19. An optical Fourier transform device according to any of the preceding projects, wherein the possible input states for one input channel are the same as the possible input states for each of the other input channels.

[0431] Project 20. An optical Fourier transform device according to any of the preceding projects, wherein the possible input states include or consist of the following:

[0432] A first input state having a first amplitude and a second input state having a second amplitude different from the first amplitude, wherein the first input state and the second input state are in phase, or

[0433] A first input state having a first phase and a second input state having a second phase out of phase with respect to the first phase, wherein the first input state has the same amplitude as the second input state, or

[0434] A first input state having a first phase and a first amplitude, and a second input state having a second phase out of phase with respect to the first phase and a second amplitude different from the first amplitude.

[0435] Optionally, the first input state is the lowest possible input state of the input channel, the second input state is the highest possible input state of the input channel, and / or the first phase is 180 degrees out of phase with the second phase.

[0436] Project 21. An optical Fourier transform device according to any of the preceding projects, wherein the number of possible input states is such that the set of all input channels can only display M. NThere are N distinct N-bit input vectors, where each combination is an Nlog2M-bit combination, where each input vector is represented by a set of input states applied to the input channel in a single frame, where M is the number of levels in the input and is a positive integer.

[0437] Project 22. An optical Fourier transform device according to any of the preceding projects, wherein the input port at the exit of each input channel is arranged on a first arc, and the output port at the entrance of each output channel is arranged on a second arc, wherein the distance between the first and second arcs is equal to the radius of the first arc and also equal to the radius of the second arc.

[0438] Project 23. An optical Fourier transform device according to Project 22, wherein the input port is arranged at an angle θn, measured from zero on the second circular arc.

[0439] Project 24. An optical Fourier transform device according to Project 22 or 23, wherein the output port is arranged at an angle θn, measured from zero on the first circular arc.

[0440] Project 25. An optical Fourier transform device according to Project 24, wherein the zero point on the first arc and the zero point on the second arc are directly opposite each other with respect to a virtual line that bisects the common radius of the first and second arcs.

[0441] Project 26. An optical Fourier transform device according to Project 24, wherein the zero point on the first arc and the zero point on the second arc are located at opposite ends of the common radius of the first and second arcs.

[0442] Project 27. An optical Fourier transform device according to any of the preceding projects, wherein the angle θn is defined by the following formula:

[0443]

[0444] Where n is the port number, N is the total number of input or output ports, and R is the radius. eff is the effective refractive index of the waveguide mode, where the port number n starts from -N / 2 if N is even, or from -((N-1)) / 2 if N is odd, and increments by integers.

[0445] This document also describes the following numbered clauses:

[0446] Data Packaging

[0447] Clause 1. A method for processing an output signal from an optical Fourier transform device, comprising:

[0448] Apply an input vector to the optical Fourier transform device.

[0449] Obtain the output vector representing the optical Fourier transform of the input vector;

[0450] Modify the channel of the output vector to form a PAM4 signal.

[0451] Clause 2. The method according to Clause 1, wherein the number of levels in the channel is greater than four, and optionally, wherein the channel carries a PAM5 signal.

[0452] Clause 3. The method according to Clause 2, wherein the channel represents the real part of the optical Fourier transform, optionally wherein the channel corresponds to a first output channel or a third output channel of the optical Fourier transform device.

[0453] Clause 4. The method according to Clause 2 or 3, wherein modifying the channel includes reducing the number of levels in the channel to form a PAM4 signal, optionally wherein the channel carries a PAM5 signal and the number of levels is reduced to four.

[0454] Clause 5. The method according to Clause 2, 3 or 4, wherein modifying the channel includes detecting the highest level of the channel and removing the highest level from the channel, optionally wherein the channel carries a PAM5 signal, and modifying the channel includes removing the fifth level from the PAM5 signal.

[0455] Clause 6. The method according to Clause 5, wherein removing the highest level from the channel includes:

[0456] Saturate or clip the highest level to the fourth level, or

[0457] Subtract one or more levels to form a fourth level.

[0458] Clause 7. The method according to any of the preceding clauses, wherein the channel carries a signal having fewer than four levels, optionally a PAM3 signal.

[0459] Clause 8. The method according to Clause 7, wherein modifying the channel includes adding one or more levels to the channel to form a PAM4 signal.

[0460] Clause 9. The method according to Clause 7 or 8, wherein the channel carries a real part or an imaginary part, and wherein the channel corresponds to a second output channel or a fourth output channel of an optical Fourier transform stage.

[0461] Clause 10. The method according to any one of Clauses 7-9, wherein modifying the channel includes forming a PAM4 signal by adding one or two levels to create a fourth level.

[0462] Clause 11. The method according to any of the preceding clauses, wherein the output vector comprises the complete complex result from the optical Fourier transform level.

[0463] Clause 12. The method according to any of the preceding clauses also includes transmitting PAM4 signals via a PAM4 link.

[0464] Clause 13. The method according to any of the preceding clauses also includes recovering the correct complete complex OFT data from the PAM4 signal by modifying the inverted channel.

[0465] Clause 14. The method according to Clause 13, wherein recovering the correct complete complex OFT data includes deriving one or more levels removed from the original output data from the PAM4 signal.

[0466] Receiver optimization

[0467] Clause 15. A method for processing a complete complex optical Fourier transform result obtained from an optical Fourier transform device, comprising:

[0468] Determine the real part of the channel output signal for each of three of the four output channels, and

[0469] Detect the imaginary part of the channel output signal in the remaining output channels.

[0470] The complete complex number result is determined by obtaining the real and imaginary parts of all four output channels using only the detected real parts from the three output channels and the detected imaginary parts from the remaining output channels.

[0471] Clause 16. The method according to Clause 15, wherein obtaining the imaginary part of the channel output signal of one of the three channels includes applying logical operations to the imaginary parts detected in the remaining output channels.

[0472] Clause 17. The method according to Clause 16, wherein obtaining the real part of the channel output signal in the remaining output channels comprises applying a logical operation to the real part of the channel output signal detected in one of the three output channels.

[0473] Clause 18. The method according to any one of Clauses 15-17, wherein obtaining the imaginary part of the channel output signal of two of the three channels includes determining that the imaginary part is zero.

[0474] Clause 19. The method according to any one of Clauses 15-18, wherein obtaining the real part of the channel output signal of any one of the three output channels and / or obtaining the imaginary part of the channel output signal of one of the three or any of the remaining channels comprises converting the detected optical output signal level into a signed number or its digital equivalent.

[0475] Logical derivation of the output

[0476] Clause 20. A method for processing data using a Fourier transform device, the method comprising:

[0477] An input data stream is applied to the Fourier transform device, wherein the input data stream is applied in a series of sequential frames, each frame including an input vector, and

[0478] An output data stream is generated from the input data stream by determining an output vector for each frame, where each output vector is the Fourier transform of the corresponding input vector in the input data stream.

[0479] Specifically, for a first input vector with a first value, the corresponding first output vector is determined through optical Fourier transform.

[0480] For a second input vector having a second value different from the first value, the corresponding second output vector is determined by means other than the optical Fourier transform of the second input vector.

[0481] Clause 21. The method according to Clause 20, wherein, for a second input vector, the corresponding second output vector is determined by applying logical operations to the second input vector.

[0482] Optionally, the second input vector has a total value of zero, and the second output vector is determined by either copying the second input vector to produce the second output vector or by determining that the second input vector also has a total value of zero.

[0483] Clause 22. The method according to Clause 20 or 21, wherein, for the second input vector, the corresponding second output vector is determined by the optical Fourier transform of the first input vector.

[0484] Clause 23. The method according to Clause 22, wherein the second input vector is encoded as a repeated input of the first input vector in the input data stream, optionally wherein the second output vector is logically determined from the optical Fourier transform of the first input vector.

[0485] Clause 24. The method according to any one of Clauses 20-23, wherein the second output vector contains the same corresponding vector values ​​as the first output vector, except for one different vector value, wherein the second output vector is logically determined from the first vector by a known relationship between the different vector values.

[0486] Clause 25. According to the method of Clause 24, where the known relation is sign reversal.

[0487] Data unpacking

[0488] Clause 26. A method for processing an output signal from an optical Fourier transform device, the method comprising:

[0489] Receive output signals in the form of PAM4 signals from the optical Fourier transform device;

[0490] The complete complex output of the optical Fourier transform device is recovered from the output signal.

[0491] Clause 27. The method according to Clause 26, wherein recovering the complete complex output of the optical Fourier transform device from the output signal includes decoding the PAM4 signal.

[0492] Clause 28. The method according to Clause 27, wherein decoding the PAM4 signal includes deriving a signal having more than four layers from the PAM4 signal, and optionally deriving a PAM5 signal from the PAM4 signal.

[0493] Clause 29. The method according to Clause 27 or 28, wherein decoding the PAM4 signal includes deriving a signal having fewer than four layers from the PAM4 signal, and optionally deriving a PAM3 signal from the PAM4 signal.

[0494] Clause 30. The method according to Clause 28 or Clause 29 (where subject to Clause 28), wherein the complete complex output of the optical Fourier transform device recovered from the PAM4 signal is defined at least in part by the PAM3 and / or PAM5 signals.

[0495] Clause 31. A method for operating an optical Fourier transform (OFT) device, the OFT device including N input channels and K output channels, the method comprising:

[0496] Quantize the light at the input channel, and

[0497] According to any of the methods in Clauses 1-14; and / or

[0498] According to any of the methods in Clauses 15-19; and / or

[0499] According to any of the methods in Clauses 20-25; and / or

[0500] According to any of the methods in Clauses 26-30.

[0501] Clause 32. The method according to Clause 31, wherein quantizing the light at the input channel includes limiting the light emitted by each input channel to a finite number of possible input states, and / or

[0502] The quantization of light at the input channel includes quantizing the light at the input channel to form a quantized output at each output channel.

[0503] Clause 33. The method according to Clause 32, wherein the number of possible input states for each input channel is 2, 3 or 4.

[0504] Clause 34. The method according to Clause 32 or 33, wherein the possible input states of one input channel are the same as the possible input states of each other input channel.

[0505] Clause 35. The method according to any one of Clauses 32-34, wherein the possible input states include or consist of the following:

[0506] A first input state having a first amplitude and a second input state having a second amplitude different from the first amplitude, wherein the first input state and the second input state are in phase, or

[0507] A first input state having a first phase and a second input state having a second phase out of phase with respect to the first phase, wherein the first input state has the same amplitude as the second input state, or

[0508] A first input state having a first phase and a first amplitude, and a second input state having a second phase out of phase with respect to the first phase and a second amplitude different from the first amplitude.

[0509] Optionally, the first input state is the lowest possible input state of the input channel, the second input state is the highest possible input state of the input channel, and / or the first phase is 180 degrees out of phase with the second phase.

[0510] Clause 36. The method according to any one of Clauses 32-35, wherein the number of possible input states is such that the set of all input channels can only display M. N There are N distinct N-bit input vectors, where each combination is an N log2M-bit combination, where each input vector is represented by a set of input states applied to the input channel in a single frame, where M is a positive integer.

[0511] Clause 37. The method according to any one of Clauses 32-36, wherein the output of the optical Fourier transform device is deterministic, such that a finite number of possible input states produce a finite number of possible output states at each output channel.

[0512] Clause 38. The method according to Clause 37, wherein the effective number of bits “ENoB” of the receiver in each output channel is equal to log2(P+Q), where P is the total number of possible output states at the corresponding output channel given possible input states, and Q is 0, 1, 2, 3, 4, 6, 8 or 10.

[0513] Clause 39. The method according to any one of Clauses 32-38, wherein a uniform difference exists between:

[0514] The first and second output states of possible output states, and

[0515] The possible output states are between the first and third output states; or where a uniform difference exists between them:

[0516] The first and second output states are possible output states; and

[0517] Between a third output state that is different from the first and second output states and a fourth output state that is different from the first, second, and third output states; or

[0518] Wherein, there is a uniform difference between all adjacent possible output states in the possible output states; or where, the output of OFT is represented by a finite number of output states, and there is a uniform interval between the output states.

[0519] Clause 40. The method according to any of the preceding clauses, wherein the receiver in each output channel has 12 or fewer effective bits “ENoB”, optionally 10 or fewer, optionally 8 or fewer, optionally 6 or fewer, optionally 4 or fewer, optionally 3 or fewer, optionally 2 or fewer.

[0520] Clause 41. The method according to any of the preceding clauses, wherein N is an integer multiple of 2, optionally an integer multiple of 4, optionally 4.

[0521] Clause 42. The method according to any of the preceding clauses, where K equals N.

[0522] Clause 43. The method according to any of the preceding clauses, wherein each output channel has a receiver different from the receiver of at least one other output channel.

[0523] Clause 44. The method according to any of the preceding clauses, wherein at least one output channel includes a receiver for recovering only real data, and another output channel includes a receiver for recovering only virtual data.

[0524] Clause 45. The method according to any of the preceding clauses, wherein three output channels include a receiver for recovering only real data, and another output channel includes a receiver for recovering only virtual data.

[0525] Clause 46. The method according to any of the preceding clauses, wherein the ENOB of the receiver of one output channel is different from the ENOB of the receiver of the other output channel.

[0526] Clause 47. The method according to any of the preceding clauses, wherein each output channel has only a single receiver.

[0527] Clause 48. The method according to any of the preceding clauses, wherein at least one output channel includes a receiver for recovering real data, the receiver having only three possible real output states, optionally wherein the three possible real output states are zero, a positive value, and a negative value, wherein the positive value and the negative value have the same magnitude.

[0528] Clause 49. The method according to any of the preceding clauses, wherein at least one output channel includes a receiver for recovering virtual data, the receiver having only three possible virtual output states, optionally wherein the three possible virtual output states are zero, positive, and negative, wherein the positive and negative values ​​have the same magnitude.

[0529] Clause 50. The method according to any of the preceding clauses, wherein the two output channels include a receiver for recovering real data, the receiver having only three possible real output states, optionally wherein the three possible real output states are zero, positive and negative, wherein the positive and negative values ​​have the same amplitude.

[0530] Clause 51. The method according to any of the preceding clauses, wherein the two output channels include a receiver for recovering virtual data, the receiver having only three possible virtual output states, optionally wherein the three possible virtual output states are zero, a positive value, and a negative value, wherein the positive value and the negative value have the same magnitude.

[0531] Clause 52. The method according to any of the preceding clauses, wherein at least one output channel includes a receiver for recovering real data, the receiver having only five possible real output states, or

[0532] Two of the output channels include a receiver for recovering real data, which has only five possible real output states.

[0533] Clause 53. A method comprising:

[0534] According to any of the methods in Clauses 1-14; and / or

[0535] According to any of the methods in Clauses 15-19; and / or

[0536] According to any of the methods in Clauses 20-25; and / or

[0537] According to any of the methods in Clauses 26-30; and / or

[0538] According to any of the methods in Clauses 31-52.

[0539] Clause 54. A computing system comprising:

[0540] Optical Fourier transform device;

[0541] One or more processors,

[0542] One or more processors are configured to perform any of the methods in Clauses 1-53, or

[0543] The computing system includes one or more transient or non-transitory computer-readable media storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations including any of the methods in clauses 1-53.

[0544] Clause 55. One or more tangible or intangible, transient or non-transitory computer-readable media storing computer-readable instructions that, when executed by one or more processors, cause one or more processors to perform operations including any of the methods in Clauses 1-53.

[0545] This document also describes the following numbered embodiments.

[0546] Example 1. An OFT device, comprising: an input channel, each input channel including a transmitter; an output channel; a monitoring channel connected to at least one input channel for sampling an optical input signal from the transmitter; and a monitoring receiver arranged to detect a monitoring signal in the monitoring channel, wherein the monitoring signal includes the optical input signal.

[0547] Example 2. The OFT device according to Example 1 further includes a perturbation channel connected to the monitoring channel to allow the perturbation signal to be mixed with the optical input signal to form a monitoring signal.

[0548] Example 3. The OFT device according to Example 1 or 2, wherein a monitoring receiver is arranged to detect the phase difference between the optical input signal and the disturbance signal.

[0549] Example 4. In some examples of the OFT apparatus according to Example 1, 2 or 3, the monitoring channel receiver is arranged to detect the strength of the monitoring signal.

[0550] Example 5. An OFT device according to any one of Examples 1 to 4, wherein the OFT device further includes at least one controller arranged to control an optical modulator to apply a first modulation to an optical input signal; detect a first value of a parameter of a monitoring signal from a monitoring receiver; and compare the detected first value with a reference value of the parameter.

[0551] Example 6. The OFT device according to Example 5, wherein at least one controller is further arranged to control the optical modulator based on the comparison to apply a second modulation to the optical input signal.

[0552] Example 7. The OFT device according to Example 6, wherein the second modulation is different from the first modulation.

[0553] Example 8. The OFT device according to Examples 2 to 7, wherein the OFT device further includes at least one controller arranged to control the disturbance signal to have a time-varying phase or a time-varying amplitude, and to record the monitoring signal at a monitoring receiver.

Claims

1. A method for operating an optical Fourier transform (OFT) device, comprising: An input signal is applied to at least one component of the OFT device; Detect a first value of a parameter of the optical signal generated by the input signal; as well as Determine a first relationship between the detected first value and a first reference value of the parameter.

2. The method according to claim 1, further comprising controlling the OFT device based on the first relationship.

3. The method according to claim 2, wherein, Controlling the OFT device based on the first relationship includes controlling the data flow through the OFT device based on the first relationship.

4. The method according to claim 3, wherein, Controlling the data flow through the OFT device based on the first relationship includes starting or pausing the data flow based on the first relationship.

5. The method according to claim 3 or 4, wherein, Controlling the OFT device based on the first relationship includes modifying the operating parameters of the at least one component based on the first relationship, such that the optical signal generated by the input signal is modified.

6. The method according to any one of claims 2 to 5, wherein, Controlling the OFT device based on the first relationship includes: The input signal is controlled based on the first relationship, or The receiver in the output channel is controlled based on the first relationship, and the receiver is arranged to detect parameters of the optical signal.

7. The method according to claim 6, wherein, Controlling the input signal based on the first relationship includes modifying the input signal.

8. The method according to claim 6 or 7, wherein, Controlling the receiver based on the first relationship includes controlling the phase of the signal applied to the receiver based on the first relationship.

9. The method according to any one of the preceding claims, wherein, The first relationship is that the difference between the detected first value and the first reference value exceeds a threshold difference.

10. The method according to claim 9, wherein, When the threshold difference is exceeded, the method further includes reapplying the input signal to the at least one component, detecting a second value of a parameter of the optical signal, and determining a second relationship between the detected second value and the first reference value.

11. The method of claim 10, wherein the second relationship is that the difference between the second value and the first reference value does not exceed the threshold difference, and the method further comprises: Continue to apply the input signal to the at least one component; or Pause the input signal.

12. The method according to any one of the preceding claims, wherein the optical signal is detected in the following ways: One or more output channels of the OFT device, or A monitoring channel connected to one or more input channels of the OFT device.

13. The method according to claim 12, wherein, The optical signal is an optical output signal detected in one or more output channels of the OFT device, and the at least one component is a free space region of the OFT device.

14. The method according to any one of the preceding claims, wherein, The input signal is an optical input signal.

15. The method according to claim 14, wherein, The at least one component of the OFT device is a monitoring channel arranged to sample the optical input signal in at least one input channel of the OFT device, and the optical signal is a monitoring signal that includes a combination of the sampled optical input signal and a disturbance signal applied to the monitoring channel.

16. The method according to claim 15, wherein, The parameter may optionally be the strength of the monitoring signal.

17. The method according to claim 15 or 16, wherein, The monitoring channel is configured to sample the optical input signal in one of the two input channels of the OFT device.

18. The method according to any one of claims 14 to 17, wherein, The at least one component is a modulator arranged to transmit the optical input signal.

19. The method of claim 18, wherein, The optical signal is: The optical input signal detected at the input channel or monitoring channel of the OFT device; or The optical output signal detected at the output channel of the OFT device.

20. The method according to any one of the preceding claims, wherein, The first reference value is theoretically derived from the input signal.

21. The method according to claim 20, wherein, The first reference value is determined based on the discrete Fourier transform of the input vector or the input signal pattern, which includes the input signal.

22. The method according to claim 21, wherein, The input vector includes a 4-bit input vector.

23. The method according to claim 21 or 22, wherein, The input vector comprises binary signal values ​​applied to each of one or more input channels of the OFT device in the form of a set of binary signal values.

24. The method according to any one of claims 21 to 23, wherein, The input signal pattern only adopts M N Any of the different forms of the input vectors can only display M N There are N different N-bit input vectors, where each combination is further an N log2M-bit combination, where each input vector is represented by a set of input states applied to the input channel in a single frame, where M is the number of levels in the input and is a positive integer, optionally where M is 2, 3 or 4, and optionally where N is a positive integer multiple of 2, 3, 4 or 2.

25. The method according to claim 21 or 24, wherein, Applying the input signal to the at least one component includes applying a first form of the input vector to the one or more input channels of the OFT device.

26. The method of claim 25, wherein, The first form of the input vector is represented in its binary representation as a binary number with only a single "1" at a first position in the vector. Furthermore, the first value of the parameter is detected in the output channel of the OFT device.

27. The method of claim 26, further comprising adjusting parameter control of the output channel until the criteria of the determined first relationship are met, optionally wherein, The parameter control is phase control.

28. The method according to claim 27, wherein, The standard is that the difference between the first value detected at the output channel and the first reference value is less than a threshold difference.

29. The method according to claim 28, wherein, The parameter is intensity, and the threshold difference is less than half the difference between the 0 and 1 binary signal states of the input signal.

30. The method according to any one of claims 25 to 29, further comprising: A second form of the input vector is applied to the set of input channels, wherein the second form is diagonal, and 1 is shown at a second position in the input vector that is different from the first position; Detect a second value of the parameter in the output channel of the OFT device; and Determine a second relationship between the detected second value and a second reference value of the parameter.

31. The method of claim 30, further comprising adjusting parameter control of the second input channel until the criterion of the determined second relationship is met, optionally wherein, The parameter control is phase control.

32. The method according to claim 31, wherein, The standard is that the difference between the second value detected at the output channel and the second reference value is less than a threshold difference.

33. The method according to claim 32, wherein, The parameter is intensity, and the threshold difference is less than half the difference between the 0 and 1 binary signal states of the input signal.

34. The method according to any one of claims 21 to 33, wherein, The discrete Fourier transform is determined via a lookup table that maps each form of the input signal pattern to a corresponding theoretical output signal pattern form, wherein each output signal pattern form includes the theoretically quantized signal state at each output channel of the OFT device.

35. The method of claim 34, further comprising: Applying a third form of the input signal pattern, other than the diagonal form, to the input channel, and wherein the method further includes detecting a third value of the parameter in the output channel of the OFT device; and A third relationship is determined between the detected third value and the third reference value of the parameter.

36. The method of claim 35, further comprising adjusting parameter control of the output channel until the criterion of the determined third relationship is met, optionally wherein, The parameter control is phase control.

37. The method of claim 36, wherein, The standard is that the difference between the second value detected at the output channel and the second reference value is less than a threshold difference.

38. The method according to claim 37, wherein, The parameter is intensity, and the threshold difference is less than half the difference between the 0 and 1 binary signal states of the input signal.

Images