Optical analog-to-digital converter, PAM-n-signal receiver, method, computer program and computer-readable data carrier

Abstract

The invention relates to an optical analog-to-digital converter, A / D converter, for converting an analog optical input signal (Sin) into a digital output signal. The optical A / D converter comprises a first optical channel for generating from the analog optical input signal (Sin) a first portion of a to-be-digitized optical output signal (Soutl) providing a first bit (Bl), wherein the first optical channel comprises a first interferometer. The optical A / D converter also comprises a second optical channel for generating from the analog optical input signal Sin a second portion of the to-be-digitized optical output signal (Sout2) providing a second bit (B2). The second optical channel is configured as a comparator.

Description

[0001] Optical analog-to-digital converter, PAM-N signal receiver, method, computer pro- gram and computer-readable data carrier

[0002] Technical field

[0003] The invention relates to an optical analog-to-digital converter, a PAM-N signal receiver, a method for converting an analog optical input signal to a digital output signal, a com- puter program and a computer readable data carrier.

[0004] Background

[0005] To transmit data over an optical fiber, a predetermined code may be adopted to con- vert the data to a physical state of light. The code may be a dictionary that translates the original data (images, text, audio) into a form that can be transmitted into the opti- cal fiber (power, polarization, wavelength, etc.). The act of translating the data into the physical property used to transmit the data is called encoding. The opposite process is called decoding. The data can be represented in analog or digital form.

[0006] In the case of an analog signal, the data can be encoded into a continuously variable property of the physical state used for the transmission. A continuously variable prop- erty can assume any intermediate value in an accepted predetermined range. Analog encoding is often referred to as a type of modulation. As an example, for audio data, the sound pressure can be encoded in the power of the light transmitted through a fi- ber. This is a type of amplitude modulation (AM).

[0007] In the case of digital encoding, the data can be translated into a series of letters of a predefined alphabet. Each letter in not a continuous variable but can assume only spe- cific values, i.e. the possible (predetermined) letters of the alphabet. Writing a text is a form of digital encoding, but commonly an alphabet is also translated into a series of bits, wherein each bit can have two values: 0 and 1.

[0008] Once the data is in digital form, it can be encoded in the physical state of the system to be transmitted. These states are called symbols. A symbol corresponds to a waveform, a state or a significant condition of the communication channel that persists for a fixed period of time. A signal can be considered as a sequence of symbols transmitted sequentially over time. A clock cycle is the fixed time span during which a symbol is transmitted.

[0009] A basic form of encoding for optical telecommunication is ASK (Amplitude-shift keying) that represents a digital form of amplitude modulation (AM). The power of the light is adjusted between two different levels (one of which could be 0, i.e., no power) and each level corresponds to the symbol 1 or 0. If the modulation consists in turning on and off the light emitter (or blocking and unblocking the light from a constantly on light emitter) the modulation is called OOK (on-off keying). The modulation can last for the whole clock cycle, in which case the encoding is called NRZ (non-return zero) or can involve only the first part of the time span, in which case it is called RZ (return zero).

[0010] In order to increase the bandwidth of an optical fiber, more complicated schemes can be used. One example is PAM (pulse amplitude modulation). PAM is a form of ASK where the power of light is modulated between more than two levels, allowing to en- code more information in a single symbol.

[0011] As an example, PAM-4 is a modulation between 4 distinct power levels that allow to encode 2 bits in a single symbol (00, 01, 10, 11). PAM-8 is a modulation between 8 dis- tinct power levels that allow to encode 3 bits in a single symbol. PAM encoding can in- clude any number of levels in theory. However, PAM-3 and PAM-4 are the most com- monly used. Encodings up to PAM-32 exist, but are comparably rare. Using this no- menclature, a signal encoded using OOK can be defined as PAM-2. In general, PAM-N with N=2nlevels denotes pulse amplitude modulation with N=2npower or amplitude levels (power is merely the square of the amplitude) corresponding to a digital signal with n bits.

[0012] PAM-4 is beneficial to increasing the bandwidth in optical (and also electrical / elec- tronic) communication. However, most of the logic and data processing devices inter- nally use the OOK encoding (equivalent to PAM-2). At the input of such devices, it is necessary to convert the analog PAM-4 signal back to OOK encoding, i.e., PAM-2. This conversion typically requires expensive and high-quality photodiodes that must be able to measure a power level with a signal-to-noise ratio (SNR) low enough to differ- entiate between all levels. Furthermore, electronic devices that behave like threshold devices are needed to cor- rectly identify the symbols and finally, additional electronics is needed to split the sin- gle input signal line into multiple (power) lines encoded in OOK. All these requirements increase cost, power consumption and latency of the transceivers and a limited baud rate.

[0013] Thus, an object of the invention is to overcome such limitations and provide an optical analog-to-digital converter, a PAM-N signal receiver, a method for converting an analog optical input signal to a digital output signal, a computer program and a computer read- able data carrier that improve upon the available systems and methods for converting analog to digital signals.

[0014] This object of the invention is achieved by the optical analog-to-digital converter, the N- level pulse amplitude modulation signal receiver, the method for converting an analog optical input signal to a digital output signal, the computer program and the computer readable data carrier as described in the appended independent claims. Advantageous developments and embodiments are described in the dependent claims.

[0015] The invention relates to an optical analog-to-digital converter, A / D converter, for con- verting an analog optical input signal into a digital output signal.

[0016] The optical A / D converter comprises a first optical channel for generating from the an- alog optical input signal a first portion of a to-be-digitized optical output signal provid- ing a first bit, wherein the first optical channel comprises a first interferometer.

[0017] The first interferometer is configured to receive a first portion of the analog optical in- put signal in a first arm of the first interferometer. The first arm comprises a phase shifter. The phase shifter is configured to shift the phase of the first portion of the ana- log optical input signal.

[0018] The first interferometer is also configured to receive a second portion of the analog optical input signal in a second arm of the first interferometer. The second arm of the first interferometer comprises a first optical nonlinear threshold device. The first opti- cal nonlinear threshold device is configured to receive and modulate the amplitude of the second portion of the analog optical input signal. The first interferometer is further configured to generate a first interference signal by interfering the phase-shifted first portion of the analog optical input signal from the first arm of the first interferometer with the amplitude-modulated second portion of the analog optical input signal from the second arm of the first interferometer or with at least a first portion of the amplitude-modulated second portion of the analog opti- cal input signal from the second arm of the first interferometer.

[0019] The first optical channel is configured to generate the first portion of the to-be-digit- ized optical output signal from the first interference signal.

[0020] The optical A / D converter also comprises a second optical channel for generating from the analog optical input signal a second portion of the to-be-digitized optical output signal providing a second bit. The second optical channel is configured as a compara- tor.

[0021] The proposed optical A / D converter combines interference and optical thresholding behavior to enable and facilitate a particular accurate and efficient analog to digital conversion. Here, interference is mediated by the first interferometer. Optical thresh- olding behavior is mediated by the first optical nonlinear threshold device. Note, that in the context of this invention amplitude-modulation mediated by an optical nonlin- ear threshold device, e.g. the first or a second optical nonlinear threshold device, does not refer to the process of encoding (binary) information into a signal, but rather to the change of the amplitude / power of the optical signal propagating through the opti- cal nonlinear threshold device, wherein the amplitude / power of the optical signal changes from input power to output power in a nonlinear manner as described by a corresponding nonlinear input output power transfer characteristic.

[0022] For example, modulating, by the first optical nonlinear threshold device, the amplitude of the second portion of the analog optical input signal as its input signal to generate the amplitude-modulated second portion of the analog optical input signal as its out- put signal means that the (output) power of the amplitude-modulated second portion of the analog optical input signal as a function of the (input) power of the second por- tion of the analog optical input signal describes a nonlinear function with a threshold behavior as explained in more detail further below.

[0023] The encoding of (binary) information into the analog optical input signal, e.g., through pulse amplitude modulation (PAM), is assumed to have been carried out before the analog optical input signal is received by the optical A / D converter. Accordingly, the analog optical input signal may be a signal with PAM-N encoding with an arbitrary number of power / amplitude levels N. For example, N may be equal or higher than 3.

[0024] The first and second portion of the to-be-digitized optical output signal can each still be considered as an analog signal representing a single bit equivalent to OOK or PAM-2 encoding. The digital output signal or first and second bits may then be obtained di- rectly from the first and second portion of the to-be-digitized optical output signals. Therefore, the optical A / D converter may also be configured as a converter from a PAM-N signal with N=2nlevels to n PAM-2 signals, wherein each of the n PAM-2 signals carry information in two power levels corresponding to a single bit. For all practical purposes, conversion from PAM-N to PAM-2 can be considered as analog-to-digital conversion since a final step after the conversion of obtaining a bit value 0 or 1 from a PAM-2 signal can be realized based on a simple conversion / threshold rule, e.g., using a digital processing unit that may comprise an additional 1-bit comparator.

[0025] The second optical channel is configured as a comparator, e.g., a one-bit comparator. Preferably, the second optical channel can be configured as an optical comparator, most preferably as an all-optical comparator, or an optical nonlinear threshold device. For example, the second optical channel may be configured to generating low output power corresponding to a first bit value of the second bit at low input power and high output power corresponding to a second bit value of the second bit at high input power, wherein low and high input power are separated by a threshold input power or a range of threshold input powers. The difference between high and low output pow- ers should be sufficiently large to be able to clearly distinguish between the first and second bit value. Preferably, the second optical channel is configured as an all-optical comparator.

[0026] More specifically, the comparator may be configured to compare the input power of the analog optical input signal with a threshold power value or a range of threshold power values. Optionally, when the input power of the analog optical input signal is below the threshold power value or range of threshold power values, the output power of the second portion of the to-be-digitized optical output signal may corre- spond to a first power value associated with a first bit value of the second bit. When the input power of the analog optical input signal is above the threshold power value or range of threshold power values, the output power of the second portion of the to- be-digitized optical output signal may correspond to a second power value associated with a second bit value of the second bit. For example, when the input power of the analog optical input signal is below a (pre-determined) threshold power value, the out- put power of the second portion of the to-be-digitized optical output signal corre- sponds to a low power value associated with bit value 0 for the second bit. When the input power of the analog optical input signal is above a (pre-determined) threshold power value, the output power of the second portion of the to-be-digitized optical out- put signal corresponds to a high power value associated with bit value 1 for the second bit.

[0027] In particular, the proposed optical A / D converter may be an all-optical A / D converter. An all-optical device here refers to a device that transmits, conditions and controls the analog optical input signal and all signals / portions derived from it and controlled and acted on by the all-optical device in the optical domain, i.e. using optical means with- out converting the signals to other domains or influencing / controlling said signals by using directly another domain. Other domains / means may include thermal, electrical or electronic domain / means.

[0028] While other domains / means may still be used for the purpose of tuning or setting an operation point or operating parameters of the A / D converter or a component thereof, the other domain / means may not directly influence the signal at the bit rate and / or operating frequency of the all-optical device itself.

[0029] For example, using in a device an electro optical modulator (EOM) to transfer the sig- nal encoding on a constant power laser output, is not considered as an operation in the pure optical domain. Such a device would not be considered an all-optical device since it makes use of electro-optical modulation to modify the signal.

[0030] Using the same EOM to tune a component of the device to a specific transmission or phase setpoint or to activate and deactivate a certain optical functionality of the de- vice does not hinder its characterization as an all-optical device since such operations are not directly influencing the data / information flow at the data / information rate. Similarly, using electricity to power an input laser or an amplifier etc. as a component of the device does also not contradict its characterization as an all-optical device for the same reason.

[0031] It is also noted that the term "optical" in the spirit of this invention refers to electro- magnetic radiation in general and is not necessarily to be understood as limiting with regard to a frequency range of the electromagnetic radiation used. For example, sig- nals received, transmitted or output by the A / D converter or a component thereof may also be microwave radiation or electromagnetic radiation in the THz range. The inven- tion is therefore applicable and feasible in a wide range of the electromagnetic spec- trum.

[0032] Optionally, the second optical channel comprises the first optical nonlinear threshold device and is configured to generate the second portion of the to-be-digitized optical output signal from at least a second portion of the amplitude-modulated second por- tion of the analog optical input signal.

[0033] Alternatively, the second optical channel comprises a second optical nonlinear thresh- old device configured to receive and modulate the amplitude of a third portion of the analog optical input signal, and the second optical channel is configured to generate the second portion of the to-be-digitized optical output signal from at least a portion of the amplitude-modulated third portion of the analog optical input signal.

[0034] The output of the first optical channel can be spatially separated from the output of the second optical channel. The first and the second optical channel may overlap at least at the first arm of the first interferometer. The first interferometer and the sec- ond optical nonlinear threshold device and / or the second optical channel may also be spatially separated from each other. In this case, the optical A / D converter may com- prise the first optical nonlinear threshold device in the first optical channel different and spatially separated from the second optical nonlinear threshold device in the sec- ond optical channel.

[0035] Optionally, the first interferometer is configured to generate the first interference sig- nal at its output with a phase difference of π corresponding to destructive interfer- ence.

[0036] The phase of the phase-shifted first portion of the analog optical input signal from the first arm of the first interferometer may be shifted by a phase shift of π with respect to the phase of the amplitude-modulated second portion of the analog optical input sig- nal from the second arm of the first interferometer or with respect to the phase of the first portion of the amplitude-modulated second portion of the analog optical input signal from the second arm of the first interferometer when interfering to generate the first interference signal.

[0037] In this way, it may be ensured that the amplitude modulated second portion of the an- alog optical input signal from the second arm of the first interferometer or at least the first portion of the amplitude modulated second portion of the analog optical input signal from the second arm of the first interferometer interferes destructively with the phase-shifted first portion of the analog optical input signal from the first arm of the first interferometer. As a result, the first interference signal may exhibit destructive in- terference.

[0038] Similarly, the phase shifter in the first arm of the first interferometer may be config- ured to shift the phase of the first portion of the analog optical input signal received by the phase shifter by a phase shift of π .

[0039] The first optical channel can be configured to receive as an input the analog optical in- put signal or at least a portion thereof. The first optical channel may be configured to generate as an output the first portion of the to-be-digitized optical output signal.

[0040] The first interferometer can be configured to receive as an input the analog optical in- put signal or at least the first and second portion of the analog optical input signal. The first interferometer may be configured to generate as an output the first interference signal.

[0041] The first optical channel and / or the first interferometer can be configured with a non- monotonic input-output power transfer characteristic. The non-monotonic input out- put power transfer characteristic can exhibit partially a negative slope along which the power of the first portion of the to-be-digitized optical output signal and / or the power of the first interference signal decreases when the power of the analog optical input signal and / or the sum of the power of the first and second portion of the analog opti- cal input signal increases. The negative slope may arise due to the effect of destructive interference in the first interference signal.

[0042] "Partially" may here refer to a certain power range in which said non-monotonic input- output power transfer characteristic exhibits the negative slope. For power values lower or higher, i.e., below or above said certain power range, the non-monotonic in- put-output power transfer characteristic may exhibit a positive slope along which the power of the first portion of the to-be-digitized optical output signal and / or the power of the first interference signal increases when the power of the analog optical input signal and / or the sum of the power of the first and second portion of the analog opti- cal input signal increases. In other words, the non-monotonic input-output power transfer characteristic may ex- hibit a positive slope (increase), a negative slope (decrease) and then again a positive slope (increase) for the output power as a function of input power in the order of in- creasing input power. For example, the non-monotonic input-output power transfer characteristic may have a square-wave shape, e.g., with sine, square, triangle or saw- tooth waveforms.

[0043] The aforementioned non-monotonic behavior of the input-output power transfer char- acteristic with increasing, decreasing and again increasing output power as a function of input power may be exhibited for the input power range as defined by the power levels used to encode information into the analog optical input signal, i.e., for powers between the lowest and highest (information carrying) power level the analog optical input signal.

[0044] The second optical channel can be configured to receive as an input the analog optical input signal or at least a portion thereof. The second optical channel may be config- ured to generate as an output the second portion of the to-be-digitized optical output signal.

[0045] The first optical nonlinear threshold device may be configured to receive as an input the second portion of the analog optical input signal. The first optical nonlinear thresh- old device may be configured to generate as an output the amplitude-modulated sec- ond portion of the analog optical input signal.

[0046] The second optical nonlinear threshold device may be configured to receive as an in- put the third portion of the analog optical input signal. The second optical nonlinear threshold device may be configured to generate as an output the amplitude-modu- lated third portion of the analog optical input signal.

[0047] The second optical channel, the first optical nonlinear threshold device and / or the sec- ond optical nonlinear threshold device may be configured as an all-optical comparator or an all-optical non-linear threshold device.

[0048] Optionally, the second optical channel and / or the second optical nonlinear threshold device may be configured with a monotonic input-output power transfer characteristic that has a consistently positive slope. Along the positive slope the power of the second portion of the to-be-digitized optical output signal and / or the power of the amplitude-modulated second portion of the an- alog optical input signal can monotonously increase as a function of the power of the analog optical input signal and / or as a function of the power of the second portion of the analog optical input signal.

[0049] Alternatively, along the positive slope the power of the second portion of the to-be- digitized optical output signal and / or the power of the amplitude-modulated third por- tion of the analog optical input signal can monotonously increase as a function of the power of the analog optical input signal and / or as a function of the power of the third portion of the analog optical input signal.

[0050] Thus, for the whole relevant power range the monotonic input-output power transfer characteristic may exhibit a monotonic behavior with a positive slope for the output power as a function of input power. The nonlinear threshold type behavior may com- prise a threshold power range in which said non-monotonic input-output power trans- fer characteristic first exhibits a positive slope that is increasing and after a threshold value a positive slope that is decreasing as a function of input power in the order of in- creasing input power.

[0051] For power values lower or higher, i.e., below or above said certain (threshold) power range, the monotonic input-output power transfer characteristic may exhibit a zero or very small positive slope. This regime below or above the threshold power range may be a saturation regime, wherein the output power is almost constant or largely insensi- tive with respect to the input power (saturated).

[0052] For example, the monotonic input-output power transfer characteristic may have the shape of a threshold function, e.g., a ramp, step, sigmoid, logistic or Heaviside func- tion.

[0053] "Consistently" may here refer to a situation where the aforementioned monotonic be- havior of the input-output power transfer characteristic, i.e., the positive slope, is ex- hibited for the whole input power range as defined by the power levels used to encode information into the analog optical input signal, i.e., for powers between the lowest and highest (information carrying) power level of the analog optical input signal.

[0054] The first and / or second optical nonlinear threshold device may comprise at least one pair of a saturable absorber (coupled) in series with a saturable amplifier. The first and / or second optical nonlinear threshold device may also comprise a plurality of pairs, wherein each pair of the plurality of pairs comprises a saturable absorber in se- ries with a saturable amplifier. All pairs of the plurality of pairs may be coupled in se- ries.

[0055] For example, the first and / or second optical nonlinear threshold device may comprise a plurality of saturable absorbers and a plurality of saturable amplifiers. The saturable absorbers and saturable amplifiers may alternate in the first and / or second optical nonlinear threshold device. For example, each saturable absorber may be followed by and coupled in series to a saturable amplifier.

[0056] A pair of a saturable absorber coupled in series with a saturable amplifier may exhibit a monotonic input-output power transfer characteristic with a threshold behavior as dis- cussed further above, e.g., behave as an all-optical comparator.

[0057] A saturable amplifier may comprise an electrically pumped active material as the am- plifying material. For example, the electrically pumped active material may be a semi- conductorjunction with a direct band gap.

[0058] A saturable amplifier may also comprise an optically pumped active material as the amplifying material. For example, the optically pumped active material may be Erbium or Ytterbium.

[0059] A saturable absorber may comprise an unbiased or reverse biased active material as absorbing material. The unbiased or reverse biased active material may be InP or GaAs.

[0060] A saturable absorber may comprise a nonlinear material that changes the refractive in- dex as a function of applied power. This behavior may be based on self-phase modula- tion (SPM), cross-phase modulation (XPM) or the Kerr effect.

[0061] Additionally, the first and / or second optical nonlinear threshold device may also com- prise one or a plurality of linear attenuators and / or linear amplifiers. The linear attenu- ators / absorbers and / or linear amplifiers may be used to scale the power of an incom- ing input signal and / or the outgoing output signal into a desirable power range such that a nonlinear optical component of the first and / or second optical nonlinear thresh- old device and / or the components / devices following the first and / or second optical nonlinear threshold device in the respective optical channel can operate properly and with the desired functionality.

[0062] The first and / or second optical nonlinear threshold device may comprise a two-dimen- sional material, e.g., graphene, to mediate an optical nonlinearity.

[0063] For example, the first and / or second optical nonlinear threshold device may comprise the two-dimensional material as a saturable absorber. Preferably, the first and / or sec- ond optical nonlinear threshold device comprise graphene as a saturable absorber.

[0064] The Fermi level of the graphene can be adjusted or selected, e.g. by doping, in such a way that the graphene has a particularly strong or effective saturable absorption. The relaxation time of the saturable absorption of the graphene can be less than a picosec- ond. Accordingly, a clock rate of the optical A / D converter and the transmitting signals can be greater than 1 THz.

[0065] The graphene may comprise at least one graphene layer. The length(s) of the graphene or the at least one graphene layer can be flexibly adjusted so that the respective threshold power value and / or the propagation loss of the saturable absorption corre- sponds to certain desirable and predetermined target values with regard to the power / intensity of the signals coupled out by the first and / or second optical nonlinear threshold element in the respective optical channel. In this way, different input-output power transfer characteristics of the first and / or second optical nonlinear threshold device and / or a / the respective optical channel can be realized. The length(s) can also be chosen to avoid critical points, e.g. bistable phases.

[0066] The length(s) of the graphene or the at least one graphene layer can be between 1 and 50 micrometers, preferably between 5 and 15 micrometers. Thus, the intensity or power of the analog optical input signal may be less than 100 mW.

[0067] The use of graphene and / or a graphene layer also enables an all-integrated and / or CMOS-based design of the optical A / D converter and / or the first and / or the second optical nonlinear threshold device.

[0068] The first interferometer may comprise a first node and a third node. The first and sec- ond arms of the first interferometer may each be configured to connect the first and third node of the first interferometer.

[0069] The first node may be configured to split the analog optical input signal or at least a portion thereof into the first portion of the analog optical input signal received by the first arm and the second potion of the analog optical input signal received by the sec- ond arm. The third node may be configured to interfere the phase-shifted (by the phase shifter) first portion of the analog optical input signal from the first arm with the amplitude-modulated (by the first nonlinear threshold device) second portion of the analog optical input signal from the second arm or at least with a portion thereof to generate the first interference signal.

[0070] The first node may comprise an uneven (power / amplitude) splitter configured to une- venly split the analog optical input signal or at least a portion of the analog optical in- put signal into the first and second portion of the analog optical input signal such that the power of the first portion of the analog optical input signal is higher than the power of the second portion of the analog optical input signal.

[0071] In this case, the stronger first portion of the analog optical input signal is received by the phase shifter and may be (merely) phase-shifted with almost no power loss, while only the weaker second portion of the analog optical input signal is received by the first optical nonlinear threshold device and gets amplitude-modulated with potential power loss, thereby decreasing the overall loss / dissipation in the optical A / D converter as compared to a case where the first node comprises an even power splitter.

[0072] Optionally, the second optical channel and / or the second optical threshold device may also comprise a second interferometer.

[0073] The first arm of the second interferometer may be configured to receive a first portion of the third portion of the analog optical input signal. The second arm of the second in- terferometer may be configured to receive a second portion of the third portion of the analog optical input signal.

[0074] The first arm of the second interferometer may comprise a phase shifter. The first arm of the second interferometer may also comprise a saturable amplifier coupled in series with the phase shifter. The saturable amplifier may be configured to modulate the am- plitude of the first portion of the third portion of the analog optical input signal. The phase shifter may be configured to shift the phase of the amplitude-modulated first portion of the third portion of the analog optical input signal. Thus, the first arm of the second interferometer can be configured to modulate the amplitude and shift the phase of the first portion of the third portion of the analog optical input signal. The second arm of the second interferometer may comprise a pair of a saturable ab- sorber and a saturable amplifier coupled in series. The second arm of the second inter- ferometer may be configured to modulate the amplitude of the second portion of the third portion of the analog optical input signal.

[0075] The second interferometer can be configured to generate a second interference signal by combining and interfering the amplitude-modulated and phase-shifted first portion of the third portion of the analog optical input signal from the first arm of the second interferometer and the amplitude-modulated second portion of the third portion of the analog optical input signal from the second arm of the second interferometer.

[0076] The second optical channel can then be further configured to generate the second por- tion of the to-be-digitized optical output signal from the second interference signal.

[0077] The phase of the amplitude-modulated and phase-shifted first portion of the third por- tion of the analog optical input signal from the first arm of the second interferometer may be shifted by a phase shift of n with respect to the phase of the amplitude-modu- lated second portion of the third portion of the analog optical input signal from the second arm of the second interferometer when interfering to generate the second in- terference signal. As a result, the second interference signal may exhibit destructive in- terference.

[0078] In the case of a second interferometer as described above, the second optical channel and / or the second optical nonlinear threshold device can also be configured with an input-output power transfer characteristic that exhibits the threshold behavior as de- scribed further above, but exhibits also a power range with a negative slope preceding a threshold power range (for increasing input power). However, in such a case the maximum output power resulting from the negative slope below the threshold power range shall be (considerably) smaller than the output power in the saturated power range above the threshold power range in order of increasing input power such that the second optical channel still behaves as a comparator that distinguishes between small output power for small input power below the threshold power range and high output power for high input power above the threshold power range.

[0079] The optical / D converter may comprise waveguides configured to receive and propa- gate the various optical signals controlled and transmitted by the optical A / D con- verter. For example, the optical / D converter may comprise an input waveguide, a first wave- guide, a second waveguide, a third waveguide, a first intermediary waveguide, a sec- ond intermediary waveguide, a third intermediary waveguide, a first further wave- guide, a second further waveguide, a first output waveguide, and / or a second output waveguide.

[0080] The input waveguide may be configured for receiving the analog optical input signal. The first waveguide may be configured for receiving the first portion of the analog op- tical input signal. The second waveguide may be configured for receiving the second portion of the analog optical input signal. The third waveguide may be configured for receiving the third portion of the analog optical input signal.

[0081] The first intermediary waveguide may be configured for receiving the first portion of the amplitude-modulated second portion of the analog optical input signal. The second intermediary waveguide may be configured for receiving an intermediary portion of the analog optical input signal. The third intermediary waveguide may be configured for receiving the second interference signal and / or an amplitude-modulated second portion of the analog optical input signal.

[0082] The first output waveguide may be configured for receiving the first interference signal and / or the first portion of the to-be-digitized optical output signal. The second output waveguide may be configured for receiving the second portion of the to-be-digitized optical output signal. The second output waveguide may also be configured for receiv- ing the amplitude-modulated second or third portion of the analog optical input signal, for receiving the second portion of the amplitude-modulated second portion of the an- alog optical input signal, and / or for receiving the second portion of the second inter- ference signal.

[0083] The first further waveguide may be configured to receive a first portion of the second or third portion of the analog optical input signal. The second further waveguide may be configured to receive a second portion of the second or third portion of the analog optical input signal.

[0084] The optical / D converter may comprise nodes configured to split or combine various optical signals received, controlled and transmitted by the optical A / D converter by a predetermined power / amplitude splitting ratio. For example, the optical / D converter may comprise an input node, a first node, an in- termediary node, a second node, a third node and / or a fourth node.

[0085] The input node, the first node, the intermediary node and / or the second node may comprise a splitter. The third node and / or the fourth node may comprise a combiner.

[0086] The first interferometer may comprise the first node and the third node connected by the first and second arm of the first interferometer, wherein the first and second arms of the first interferometer may be spatially separated.

[0087] The second interferometer may comprise the intermediary node and the fourth node connected by the first and second arm of the second interferometer, wherein the first and second arms of the second interferometer may be spatially separated.

[0088] The input, first and / or intermediary node may also comprise an uneven splitter as dis- cussed further above. The uneven splitter of the input node, first node and / or interme- diary node may be configured to unevenly split an incoming signal into a first portion receivable by the first arm of the first or second interferometer and a second portion receivable by the second arm of the first or second interferometer such that the power of the first portion in the first arm of the first or second interferometer is higher than the power of the second portion in the second arm of the first or second interferome- ter. The incoming signal of the input and / or first node may be at least a portion of the analog optical input signal, the analog optical input signal and / or the intermediary por- tion of the analog optical input signal. The incoming signal of the intermediary node may be at least a portion of the analog optical input signal, the second portion of the analog optical input signal and / or the third portion of the analog optical input signal.

[0089] An uneven splitter with an uneven power ratio can be realized with directional cou- plers of suitable lengths. Alternatively, or additionally, the first and / or second interfer- ometer may also be an asymmetrical multi-mode interferometer (MMI).

[0090] In the following, different optional configurations for a network of nodes and wave- guides corresponding to an optical A / D converter are described. These optional config- urations or single features of these optional configurations can be combined and modi- fied with other optional features as described further above and further below.

[0091] According to a first configuration, the first optical channel may comprise the input waveguide, the input node, the first and second waveguide, the second node, the first intermediary waveguide, the third node and the first output waveguide. The second optical channel may comprise the input waveguide, the input node, the second wave- guide, the second node and the second output waveguide.

[0092] In the first configuration, the first interferometer may comprise the input node, the first and second waveguide, the second node, the first intermediary waveguide and the third node. The first arm of the first interferometer may comprise the first wave- guide. The second arm of the first interferometer may comprise the second waveguide and the first intermediary waveguide.

[0093] According to the first configuration, the input node may be configured to split the ana- log optical input signal into the first and second portion of the analog optical input sig- nal. The input node may split the input waveguide into the first and second waveguide. The first node may be the same as the input node.

[0094] In the first configuration, the second node may be configured to split the amplitude- modulated second portion of the analog optical input signal into the first and second portion of the amplitude-modulated second portion of the analog optical input signal. For example, the second node may split the second waveguide into the first intermedi- ary waveguide and the second output waveguide.

[0095] In the first configuration, the third node may be configured to combine the phase- shifted first portion of the analog optical input signal and the first portion of the ampli- tude-modulated second portion of the analog optical input signal into the first interfer- ence signal. The third node may be configured to combine the first waveguide and the first intermediary waveguide into the first output waveguide.

[0096] In a second configuration, the second optical channel may comprise the second optical nonlinear threshold device that is different from the first optical nonlinear threshold device.

[0097] According to the second configuration, the first optical channel may comprise the in- put waveguide, the input node, the second intermediary waveguide, the first node, the first and second waveguide, the third node and the first output waveguide. The second optical channel may comprise the input waveguide, the input node, the third wave- guide and the second output waveguide. In the second configuration, the first interferometer may comprise the first node, the first and second waveguide and the third node. The first arm of the first interferometer may comprise the first waveguide. The second arm of the first interferometer may comprise the second waveguide.

[0098] In the second configuration, the input node may be configured to split the analog opti- cal input signal into the third and the intermediary portion of the analog optical input signal. For example, the input node may be configured to split the input waveguide into the third waveguide and the first intermediary waveguide.

[0099] In the second configuration, the first node may be different from the input node. For example, the first node may be configured to split the intermediary portion of the ana- log optical input signal into the first and second portion of the analog optical input sig- nal (which may also be denoted as the first and second portion of the intermediary portion of the analog optical input signal). In this case, the first node may be config- ured to split the first intermediary waveguide into the first and second waveguide.

[0100] In the second configuration, the third node may be configured to combine the phase- shifted first portion of the analog optical input signal and the amplitude-modulated second portion of the analog optical input signal into the first interference signal. The third node may combine the first and second waveguide into the first output wave- guide.

[0101] In the second configuration, the second output waveguide may be part of and con- nected to the third waveguide.

[0102] In a third configuration, the second optical channel may comprise a second interferom- eter.

[0103] According to the third configuration, the first optical channel may comprise the input waveguide, the input node, the second intermediary waveguide, the first node, the first and second waveguide, the third node and the first output waveguide.

[0104] In the third configuration, the first interferometer may comprise the first node, the first and second waveguide and the third node. The first arm of the first interferometer may comprise the first waveguide. The second arm of the first interferometer may comprise the second waveguide. According to the third configuration, the second optical channel may comprise the in- put waveguide, the input node, the third waveguide, the intermediate node, the first further waveguide, the second further waveguide, the fourth node and the second output waveguide.

[0105] The second interferometer may comprise the intermediate node, the first and second further waveguide and the fourth node. The first arm of the second interferometer may comprise the first further waveguide. The second arm of the second interferome- ter may comprise the second further waveguide.

[0106] In the third configuration, the input node may be configured to split the analog optical input signal into the third and the intermediary portion of the analog optical input sig- nal. For example, the input node may be configured to split the input waveguide into the third waveguide and the first intermediary waveguide.

[0107] In the third configuration, the first node may be different from the input node. For ex- ample, the first node may be configured to split the intermediary portion of the analog optical input signal into the first and second portion of the analog optical input signal. In this case, the first node may be configured to split the first intermediary waveguide into the first and second waveguide.

[0108] In the third configuration, the intermediary node may be configured to split the third portion of the analog optical input signal into a first and a second portion of the third portion of the analog optical input signal. For example, the intermediary node may split the third waveguide into the first further and the second further waveguide.

[0109] In the third configuration, the third node may be configured to combine the phase- shifted first portion of the analog optical input signal and the amplitude-modulated second portion of the analog optical input signal into the first interference signal. The third node may combine the first and second waveguide into the first output wave- guide.

[0110] In the third configuration, the fourth node may be configured to combine the first and second portion of the third portion of the analog optical input signal into the second interference signal. For example, the fourth node may be configured to combine the first further and second further waveguide into the second output waveguide. In a fourth configuration, the first interferometer may comprise the second interfer- ometer and each arm of the first and second interferometer may comprise a same unit cell of optical components. The different functionality of each arm may be ensured by setting different operating parameters in each unit cell.

[0111] According to the fourth configuration, the first optical channel may comprise the input waveguide, the input node, the first waveguide, the second waveguide, the intermedi- ary node, the first further waveguide, the second further waveguide, the fourth node, the third intermediary waveguide, the second node, the first intermediary waveguide, the third node and the first output waveguide.

[0112] In the fourth configuration, the first interferometer may comprise the input node, the first waveguide, the second waveguide, the intermediary node, the first further wave- guide, the second further waveguide, the fourth node, the third intermediary wave- guide, the second node, the first intermediary waveguide, the third node. The first arm of the first interferometer may comprise the first waveguide. The second arm of the first interferometer may comprise the second waveguide, the second interferometer, the third intermediary waveguide, the second node and the first intermediary wave- guide.

[0113] According to the fourth aspect, the second optical channel may comprise the input waveguide, the input node, the second waveguide, the second interferometer, the third intermediary waveguide, the second node and the second output waveguide.

[0114] In the fourth configuration, the first and the second optical channel may each com- prise the second interferometer. The second interferometer may comprise the inter- mediate node, the first and second further waveguide and the fourth node. The first arm of the second interferometer may comprise the first further waveguide. The sec- ond arm of the second interferometer may comprise the second further waveguide.

[0115] In the fourth configuration, the input node may be configured to split the analog opti- cal input signal into the first and second portion of the analog optical input signal. For example, the input node may split the input waveguide into the first and second wave- guide. The first node may be the same as the input node.

[0116] In the fourth configuration, the intermediary node may be configured to split the sec- ond portion of the analog optical input signal into a first and a second portion of the second portion of the analog optical input signal. For example, the intermediary node may split the second waveguide into the first further and the second further wave- guide.

[0117] In the fourth configuration, the second portion of the analog optical input signal may be amplitude-modulated and / or phase-shifted before it reaches the intermediary node. This may be accounted for by noting that the intermediary node may be config- ured to split the amplitude-modulated and / or phase-shifted second portion of the ana- log optical input signal into a first and a second portion of the second portion of the analog optical input signal.

[0118] In the fourth configuration, the fourth node may be configured to combine the (possi- bly amplitude-modulated and / or phase-shifted) first and second portion of the (possi- bly amplitude-modulated and / or phase-shifted) second portion of the analog optical input signal into a second interference signal. For example, the fourth node may be configured to combine the first and second further waveguide into the third intermedi- ary waveguide.

[0119] In the fourth configuration, the second node may be configured to split the second in- terference signal into the first and second portion of the amplitude-modulated (and possibly phase-shifted) second portion of the analog optical input signal and / or first and second portion of the second interference signal. For example, the second node may split the third intermediary waveguide and / or the second waveguide into the first intermediary and the second output waveguide. Here, in the fourth configuration, the second interferometer and / or the third intermediary waveguide may be considered as part of the second waveguide.

[0120] In the fourth configuration, the third node may be configured to combine the phase- shifted (and possibly also amplitude-modulated) first portion of the analog optical in- put signal and the first portion of the amplitude-modulated (and possibly also phase- shifted) second portion of the analog optical input signal and / or the first portion of the second interference signal into the first interference signal.

[0121] The invention also relates to an N-level pulse amplitude modulation signal receiver, PAM-N signal receiver, comprising one or a plurality of optical analog-to-digital con- verter as described further above, wherein N is an integer equal or greater than 4.

[0122] For example, a 4-level PAM signal receiver, PAM-4 signal receiver, may be identical with or may at least comprise a first optical A / D converter. An 8-level PAM signal receiver, PAM-8 signal receiver, may be identical with or may at least comprise a first and a second optical A / D converter.

[0123] The PAM-N signal receiver may be configured to receive an analog optical input signal (pulse) which carries information that has been encoded using amplitude modulation with N=2npower or (power) amplitude levels of the analog optical input signal. The PAM-N signal receiver may be configured to convert the analog optical input signal into a digital signal comprising n bit values from n bits. For example, a PAM-4 signal re- ceiver may be configured to convert PAM-4 encoding (N=4) to OOK encoding (n=2).

[0124] In other words: The PAM-N signal receiver may be configured to receive an analog op- tical input signal comprising pulse amplitude modulation with N=2nlevels (PAM-N sig- nal) and convert it to n output signals, wherein each output signal provides a bit and thus corresponds to a PAM-2 signal. Each PAM-2 signal can then be converted to the digital domain in a direct manner to give a digital output signal.

[0125] The integer N may be equal or greater than 8. In this case, the PAM-N signal receiver may comprise at least a first and a second optical A / D converter as described further above.

[0126] The first optical channel of the first and second optical A / D converter may be coupled in series such that the analog optical input signal of the first A / D converter is obtained from the first portion of the to-be-digitized optical output signal of the second optical A / D converter.

[0127] For example, a PAM-8 signal receiver may be configured to convert an analog optical input signal with 8 level PAM encoding to a digital signal with 3 bit values. The PAM-8 signal receiver may be identical with or at least comprise two optical A / D converter, e.g., the first and the second optical A / D converter. The first bit of the first optical A / D converter may correspond to a first bit of the PAM-8 signal receiver. The second bit of the first optical A / D converter may correspond to a second bit of the PAM-8 signal re- ceiver. The second bit of the second optical A / D converter may correspond to a third bit of the PAM-8 signal receiver. The first, second and third bit of the PAM-8 signal re- ceiver may yield the 3 bit values of the digital output signal.

[0128] In general, the PAM-N signal receiver with N=2nlevels may be configured to convert an analog optical input signal with N level PAM encoding to a digital signal with n bit val- ues, where n is also an integer. The PAM-N signal receiver may comprise (n-1) optical A / D converter providing the n bits of the PAM-N signal receiver. All n optical A / D con- verter may be coupled in series, wherein a pair of neighboring optical A / D converter in the series is coupled via the first optical channels of the respective neighboring optical A / D converter. In this case, the analog optical input signal of the i-th A / D converter with i=1,..,n-1, is obtained from the first portion of the to-be-digitized optical output signal of the (i+1)-th optical A / D converter.

[0129] The first bit of the first optical A / D converter may correspond to a first bit of the PAM- N signal receiver. The second bit of the first optical A / D converter may correspond to a second bit of the PAM-N signal receiver. The second bit of the second optical A / D con- verter may correspond to a third bit of the PAM-N signal receiver etc. The second bit of the i-th optical A / D converter may then correspond to the (i+1)-th bit of the PAM-N signal receiver.

[0130] In other words: The second bit of the (n-1)th optical A / D converter may then corre- spond to the n-th bit of the PAM-N signal receiver. The first to n-th bit of the PAM-N signal receiver may yield the n bit values of the digital output signal of the PAM-N sig- nal receiver.

[0131] Optionally, the first bit of the PAM-N signal receiver may be the least significant bit (LSB) of the digital output signal. The n-th bit of the PAM-N signal receiver may corre- spond to the most significant bit (MSB) of the of the digital output signal.

[0132] For example, the PAM-N signal receiver may comprise a digital processing unit. The digital processing unit may be configured to convert at least the first portion of the to- be-digitized optical output signal to a first bit of a digital signal and a second portion of the to-be-digitized optical output signal to a second bit of the digital signal.

[0133] For example, the digital processing unit may be configured to convert the first portion of the to-be-digitized optical output signal of the first optical analog-to-digital con- verter to a first bit of the PAM-N signal receiver and / or the digital signal, the second portion of the to-be-digitized optical output signal of the first optical analog to digital converter to a second bit of the PAM-N signal receiver and / or the digital signal and / or the second portion of the to-be-digitized optical output signal of the second optical an- alog to digital converter to a third bit of the PAM-N signal receiver and / or the digital signal. The digital signal may be the digital output signal. The digital signal and / or the digital output signal may at least partially be obtained from the first and second portion of the to-be-digitized optical output signal. Optionally, the digital signal and / or the digital output signal may comprise bit values of the first bit and the second bit.

[0134] The invention relates to a method for converting an analog optical input signal into a digital output signal.

[0135] The method comprises generating in a first optical channel a first portion of a to-be- digitized optical output signal from the analog optical input signal for providing a first bit.

[0136] The generating of the first portion of a to-be-digitized optical output signal comprises

[0137] - receiving a first portion of the analog optical input signal in a first arm of the first interferometer;

[0138] - shifting in the first arm of the first interferometer the phase of the first por- tion of the analog optical input signal by using a phase shifter; and

[0139] - receiving a second portion of the analog optical input signal in a second arm of the first interferometer, and

[0140] -modulating in the second arm of the first interferometer the amplitude of the second portion of the analog optical input signal by using a first optical nonlin- ear threshold device; and

[0141] - generating a first interference signal by interfering the phase-shifted first por- tion of the analog optical input signal from the first arm of the first interferom- eter with the amplitude-modulated second portion of the analog optical input signal or with at least a first portion of the amplitude-modulated second por- tion of the analog optical input signal from the second arm of the first interfer- ometer; and

[0142] - generating the first portion of the to-be-digitized optical output signal from the first interference signal;

[0143] The method also comprises generating in a second optical channel a second portion of the to-be-digitized optical output signal from the analog optical input signal providing a second bit, wherein the second optical channel is configured as a comparator.

[0144] The generating of the second portion of the to-be-digitized optical output signal may comprise:

[0145] - generating the second portion of the to-be-digitized optical output signal from at least a second portion of the amplitude-modulated second portion of the an- alog optical input signal from the second arm of the first interferometer; or

[0146] -receiving and modulating the amplitude of a third portion of the analog optical input signal by using a second optical nonlinear threshold device, and

[0147] -generating the second portion of the to-be-digitized optical output signal from the amplitude-modulated third portion of the analog optical input signal.

[0148] The method may further comprise converting the first portion of the to-be-digitized optical output signal to a first bit and the second portion of the to-be-digitized optical output signal to a second bit.

[0149] When receiving and modulating the amplitude of a third portion of the analog optical input signal by using a second optical nonlinear threshold device, the method may fur- ther comprise:

[0150] - splitting the analog optical input signal into the third portion of the analog op- tical input signal and an intermediary portion of the analog optical input signal and

[0151] -splitting the intermediary portion of the analog optical input signal into the first and second portion of the analog optical input signal.

[0152] The invention also relates to a computer program comprising instructions which, when the program is executed by a computer using an optical analog to digital converter or a PAM-N signal receiver, cause the computer to carry out the method for converting an analog optical input signal to a digital optical output signal or any combination of its features as described above.

[0153] The computer program (or a sequence of instructions) may use software means for per- forming the method for converting an analog optical input signal to a digital optical output signal when the computer program runs in a computing unit. The computer pro- gram can be stored directly in an internal memory, a memory unit, or the computer.

[0154] The invention also relates to a computer-readable data carrier having stored there on the computer program described above. The computer program can be stored in ma- chine-readable data carrier(s), preferably digital storage media.

[0155] The proposed invention facilitates all-optical conversion of analog signals into the digi- tal domain (bits) in a particular accurate and efficient manner. This is achieved by com- bining optical nonlinear threshold behaviour with interference, wherein the interfer- ence is generated with an interferometer that comprises at least one optical nonlinear component in the second arm used to generate a minimum in the respective input- output power transfer characteristic. This can be achieved by using a combination of saturable absorption, saturable amplification and interference.

[0156] In addition, it is noted that the term "optical" in the sense of the present invention re- fers to electromagnetic radiation in general and is not necessarily to be understood as limiting with respect to a frequency range of the electromagnetic radiation used. For example, input or output signals of the optical A / D converter or its components may also be microwave radiation or electromagnetic radiation in the THz range. For exam- ple, an optical non-linear threshold device may also be formed with a superconducting circuit comprising at least one Josephson contact (emitting microwave radiation) or a (non-linear) Tera-Hertz resonator or Tera-Hertz metamaterial. Accordingly, the inven- tion is applicable and practicable in a wide range of the electromagnetic spectrum.

[0157] The optical A / D converter (or PAM-N converter / signal receiver) can be applied in a wide range of technological areas. For example, the optical A / D converter can be used in datacenters to convert between the PAM-4 encoding in the optical interconnects and the computer blades that typically use on-off keying (OOK) encoding. The inven- tion significantly reduces the latency, making it negligible if compared to the delay in- troduced by the cable length. Furthermore, the conversion and decoding operation can be performed at a significant lower power consumption and cost.

[0158] The optical A / D converter could be implemented as a digitizer for optical sensors that generate analog signals for further processing by optical or electronic computers. In this context the invention provides several advantages:

[0159] - Converting images from an image sensor based on fiber arrays (Sensors that focus light on a 2D fiber array in which every fiber acts as a pixel of the more common CCD or CMOS image sensor) directly to the digital domain. This appli- cation can be used in self-driving cars, augmented / virtual / mixed reality gog- gles, OCT detectors;

[0160] - Digitizing the output of optics based medical sensors;

[0161] - Providing an interface between digital computation, both optical and elec- tronic, with analog based types of optical computers. Examples of such devices are quantum optical computers, optical neuromorphic neural networks, optical vector matrix multiplicators;

[0162] Exemplary embodiments of the invention are illustrated in the drawings and will now be described with reference to figures 1a to 10b.

[0163] In the figures:

[0164] Fig. 1a shows a truth table for first and second bit,

[0165] Fig. 1b shows an input-output power transfer characteristic of the second opti- cal channel for the second bit,

[0166] Fig. 1c shows an input-output power transfer characteristic of the first optical channel for the first bit,

[0167] Fig. 2a shows an embodiment of the optical A / D converter,

[0168] Fig. 2b shows an input-output power transfer characteristic of an optical non- linear threshold device,

[0169] Fig. 3a shows an embodiment of the optical A / D converter according to a first configuration,

[0170] Fig. 3b shows an embodiment of the first optical nonlinear threshold device,

[0171] Fig. 4a shows an embodiment of the optical A / D converter with uneven splitter.

[0172] Fig. 4b shows an embodiment of the first optical nonlinear threshold device,

[0173] Fig. 5a shows an embodiment of a PAM-N signal receiver. Fig. 5b shows an embodiment of a PAM-8 signal receiver,

[0174] Fig. 5c shows an embodiment of the first and second optical nonlinear thresh- old device.

[0175] Fig. 6a shows input-output power transfer characteristics of a linear amplifier and a linear absorber.

[0176] Fig. 6b shows input-output power transfer characteristics of a saturable ab- sorber,

[0177] Fig. 6c shows input-output power transfer characteristics of a saturable ampli- fier.

[0178] Fig. 7 shows coupling schemes between a waveguide and graphene,

[0179] Fig. 8a shows an embodiment of the first optical nonlinear threshold device.

[0180] Fig. 8b shows an embodiment of the optical A / D converter according to a sec- ond configuration.

[0181] Fig. 8c shows an input-output power transfer characteristic of the second opti- cal channel,

[0182] Fig. 8d shows an input-output power transfer characteristic of the first optical channel,

[0183] Fig. 9a shows an embodiment of the optical A / D converter according to a third configuration.

[0184] Fig. 9b shows an input-output power transfer characteristic of the second opti- cal channel,

[0185] Fig. 9c shows an input-output power transfer characteristic of the first optical channel,

[0186] Fig. 10a shows an embodiment of a unit cell,

[0187] Fig. 10b shows an embodiment of the optical A / D converter according to a fourth configuration. Figure 1

[0188] Figure 1a shows a truth table for the first and second bit according to an embodiment of an optical A / D converter or an optical PAM-4 signal receiver. Here, Pindenotes the (input) power of an analog optical input signal Sin with N=4 power levels correspond- ing to power values 0, 1, 2 and 3 (in normalized dimensionless units). Here, the input power of the analog optical input signal Sin is normalized such that the minimal input power is 0 and the maximum input power corresponds to power value 3. The PAM en- coding of the analog optical input signal is chosen such that each power level corre- sponds to one combination of bit values of the first and second bit, i.e., the four equi- distant power values correspond to bit values 00, 01, 10 and 11.

[0189] Figures 1b and 1c show exemplary corresponding (to-be-engineered) input-output power transfer characteristics for the first B1 and second B2 bit.

[0190] The optical A / D converter and optical PAM-4 signal receiver comprises a first optical channel configured to generate from the analog optical input signal Sin the first por- tion of a to-be-digitized optical output signal Sout1 providing the first bit B1 and a sec- ond optical channel configured to generate from the analog optical input signal Sin the second portion of the to-be-digitized optical output signal Sout2 providing the second bit B2. The first and second to-be-digitized optical output signals Sout1, Sout2 can thus be considered as PAM-2 signals.

[0191] The output power levels of the second optical channel, i.e., the power levels of the sec- ond portion of the to-be-digitized optical output signal Sout2 therefore correspond to the bit values 0, 0, 1, 1 of Bit 2 in the order of increasing input power. Bit 2 and the sec- ond portion of the to-be-digitized optical output signal Sout2 therefore behave like a comparator. The comparator turns to a high bit value 1 when the input power is be- tween 1 and 2 in the normalized dimensionless units.

[0192] The output power levels of the first optical channel, i.e., the power levels of the first portion of the to-be-digitized optical output signal Sout1 correspond to the bit values 0, 1, 0, 1 of Bit 1 in the order of increasing input power. The behavior of the bit B1 is thus more complicated since it turns high, low and high again as the input power in- creases. Such a behavior cannot be achieved using linear optics and even accounting for nonlinear effects, it is difficult to engineer such a non-monotonic behavior in a pre- cise and reliable manner. Figure 1b shows input-output power transfer characteristics of the second optical channel for the second bit. The input-output power transfer characteristic of the sec- ond optical channel describes the output power of the second portion of the to-be-dig- itized optical output signal Sout2 as a function of the input power of the analog optical input signal Sin. It has the shape of a threshold function, i.e., step or Heaviside function (dotted), a linear ramp function (solid) and a sigmoid (dashed) function is shown. A threshold is engineered around the input power 1,5 and in the range between input power levels 1 and 2.

[0193] Figure 1c shows input-output power transfer characteristics of the first optical channel for the first bit. The input-output power transfer characteristic of the first optical chan- nel describes the output power of the first portion of the to-be-digitized optical output signal Sout1 as a function of the input power of the analog optical input signal Sin. The output power initially increases, then decreases, then increases again in the order of increasing input power. The output power has the shape of a square-wave with sine (dashed line), square (dotted line) and triangle or sawtooth (solid line) waveforms. A first turning point where the output power changes from increasing to decreasing power is engineered at the input power of 1 and a second turning point where the out- put power changes from decreasing to increasing is engineered at power level 2.

[0194] An input-output power transfer characteristic may also be called a response function, an input-output function or a transfer function.

[0195] Recurring features are provided in the following figures with identical reference signs as in Figures 1a to 1c.

[0196] Figure 2

[0197] Figure 2a shows an embodiment of the optical analog to digital converter for convert- ing the analog optical input signal Sin into a digital output signal.

[0198] The optical A / D converter comprises the first optical channel for generating from the analog optical input signal Sin a first portion of the to-be-digitized optical output signal Sout1 providing the first bit B1.

[0199] The first optical channel comprises a first interferometer configured to receive a first portion of the analog optical input signal S1 in a first arm 1.1 of the first interferome- ter. The first arm 1.1 comprises a phase shifter 1 to shift the phase of the first portion of the analog optical input signal S1. The phase shifter 1 is configured to shift the phase of the first portion of the analog optical input signal S1 by π. In this embodiment the first arm 1.1 also comprises an optical nonlinear threshold device 2.0 configured to receive and modulate the amplitude of the (phase-shifted) first portion of the analog optical input signal.

[0200] The first optical channel is also configured to receive a second portion of the analog optical input signal S2 in a second arm 1.2 of the first interferometer. The second arm 1.2 comprises a first optical nonlinear threshold device 2.1. The first optical nonlinear threshold device 2.1 is configured to receive and modulate the amplitude of the sec- ond portion of the analog optical input signal S2.

[0201] In this embodiment, the first optical nonlinear threshold device 2.1 comprises a linear attenuator 3.0 configured to attenuate the power of the second portion of the analog optical input signal S2 by a factor of 1 / 3. The first optical nonlinear threshold device 2.1 also comprises an optical nonlinear threshold device 2.0 configured to receive and modulate the (attenuated) amplitude of the second portion of the analog optical input signal S2.

[0202] In this embodiment, the first optical channel comprises an additional arm configured to receive a fourth portion of the analog optical input signal S4. The additional arm comprises a linear attenuator 3.0 configured to attenuate the power of the fourth por- tion of the analog optical input signal S4 by a factor of 1 / 5. The fourth arm also com- prises an optical nonlinear threshold device 2.0 configured to receive and modulate the amplitude of the (attenuated) fourth portion of the analog optical input signal S4.

[0203] The first optical channel is also configured to generate a first interference signal by in- terfering the phase-shifted first portion of the analog optical input signal Sil from the first arm 1.1 with the amplitude-modulated second portion of the analog optical input signal S21 from the second arm 1.2 at a third node N3.

[0204] The first optical channel is also configured to generate another interference signal by interfering the first interference signal with the amplitude-modulated fourth portion of the analog optical input signal S41 from the additional arm at an auxiliary node N3.1 (see explanations further below).

[0205] The first optical channel is further configured to generate the first portion of the to-be- digitized optical output signal Sout1 from the first interference signal. More specifically, the first optical channel is configured to generate the first portion of the to-be-digitized optical output signal Sout1 from the another interference signal which itself has been generated from the first interference signal.

[0206] The optical A / D converter comprises a second optical channel for generating from the analog optical input signal Sin a second portion of the to-be-digitized optical output signal Sout2 providing a second bit B2.

[0207] In this embodiment, the second optical channel comprises the second optical nonlin- ear threshold device 2.2. The second optical nonlinear threshold device 2.2 is config- ured to receive and modulate the amplitude of a third portion of the analog optical in- put signal S3.

[0208] In this embodiment, the second optical nonlinear threshold device 2.2 comprises a lin- ear attenuator 3.0 configured to attenuate the power of the third portion of the ana- log optical input signal S3 by a factor of 1 / 3. The second optical nonlinear threshold de- vice 2.2 also comprises an optical nonlinear threshold device 2.0 configured to receive and modulate the amplitude of the (attenuated) third portion of the analog optical in- put signal S3.

[0209] The output of the second optical nonlinear threshold device 2.2 is the amplitude-mod- ulated third portion of the analog optical input signal S31. The second optical channel is then configured to generate the second portion of the to-be-digitized optical output signal Sout2 from the amplitude-modulated third portion of the analog optical input signal S31. In this embodiment, the second portion of the to-be-digitized optical out- put signal Sout2 is identical with the amplitude-modulated third portion of the analog optical input signal S31.

[0210] In Figure 2a, the various optical components of the optical A / D converter are con- nected via waveguides and nodes.

[0211] More specifically, the first optical channel additionally comprises an input waveguide, a first node Nl, a first, second and fourth waveguide, the third node N3, a first inter- mediary waveguide, the auxiliary node N3.1 and a first output waveguide. The first node Nl is here identical with the input node.

[0212] The first interferometer additionally comprises the first node Nl, the first and second waveguide and the third node N3. The first arm 1.1 of the first interferometer additionally comprises the first waveguide. The second arm 1.2 of the first interferom- eter additionally comprises the second waveguide.

[0213] The second optical channel additionally comprises the input waveguide, the first node Nl, a third waveguide and the second output waveguide.

[0214] In Figure 2a, the input waveguide is configured to receive the analog optical input sig- nal Sin. The first node Nl is configured to split the analog optical input signal Sin into the first S1, second S2, third S3 and fourth S4 portion of the analog optical input signal. The first node Nl also splits the input waveguide into the first, second, third and fourth waveguide.

[0215] The first waveguide is configured to receive the first portion of the analog optical input signal S1 and is part of the first arm 1.1 of the first interferometer, the second wave- guide is configured to receive the second portion of the analog optical input signal S2 and is part of the second arm 1.2 of the first interferometer, the third waveguide is configured to receive the third portion of the analog optical input signal S3, the fourth waveguide is configured to receive the fourth portion of the analog optical input signal S4.

[0216] The first waveguide connects the first node Nl and the third node N3 and couples to the phase shifter 1 and the optical nonlinear threshold device 2.0 of the first arm 1.1 of the first interferometer. The second waveguide connects the first node Nl and the third node N3 and couples to the first optical nonlinear threshold device 2.1 of the sec- ond arm 1.2 of the first interferometer. The third waveguide connects to the first node Nl and couples to the second optical nonlinear threshold device 2.2. The fourth wave- guide connects the first node Nl and the auxiliary node N3.1 and couples to the linear attenuator 3.0 and the optical nonlinear threshold device 2.0 of the additional arm of the first optical channel.

[0217] The third node N3 is configured to combine and interfere the phase-shifted (and am- plitude-modulated) first portion of the analog optical input signal Sil and the ampli- tude-modulated second portion of the analog optical input signal S21 into the first in- terference signal. The third node N3 is also configured to combine the first waveguide and the second waveguide into the first intermediary waveguide. Here, the first inter- mediary waveguide is configured to receive the first interference signal and connects the third node N3 and the auxiliary node N3.1. The auxiliary node N3.1 is configured to combine and interfere the first interference signal and the amplitude-modulated fourth portion of the analog optical input signal S41 into the second interference signal identical to the first portion of the to-be-digit- ized optical output signal Sout1. The auxiliary node N3.1 is configured to combine the first intermediary waveguide and the fourth waveguide into the first output wave- guide.

[0218] Figure 2b shows a monotonic input-output power transfer characteristic of the optical nonlinear threshold device 2.0 as part of the first optical nonlinear threshold device 2.1 and the second optical nonlinear threshold device 2.2 as shown in Figure 2a. An optical nonlinear threshold device 2.0 with a normalized (monotonic) input-output power transfer characteristic as shown in Figure 2b can also be called normalized or cali- brated optical nonlinear threshold device 2.0.

[0219] The output power Poutshows a monotonic increase as a function of the input power Pinaround a threshold input power of 0.5 (dimensionless units) from almost zero input power to a saturation output power value X=1 (here chosen to be 1 for convenience). In Figure 2b, the input-output power transfer characteristic has the shape of a logistic curve or sigmoid curve similar to the dashed line in Figure 1b.

[0220] Note, that the shape of the input-output power transfer characteristic shown in Figure 2b is similar to the input-output power transfer characteristic of the second optical channel in Figure 1b. The linear attenuator 3.0 in the second optical channel of Figure 2a merely attenuates the (initial) power of the third portion of the analog optical input signal S3 and thus merely rescales / normalizes the input power to bring it into the threshold range of the optical nonlinear threshold device 2.0. So the x-axis for the in- put power Pin in Figure 2b has been rescaled with respect to Figure 1b.

[0221] The threshold region of input-output power transfer characteristic in Figure 2b can be even steeper such that any input power below 0.5 corresponds to an output power 0 and any input power above 0.5 corresponds to an output power 1.

[0222] The working principle of the optical A / D converter shown in Figure 2a can now be ex- plained in more detail as follows:

[0223] Assuming without loss of generality that the analog optical input signal Sin features N=4 different power levels with input powers 0, 4, 8 and 12 (in dimensionless units), respectively. The first node N1 comprises an equal power splitter such that each one of the first S1, second S2, third S3 and fourth S4 portion of the analog optical input signal then features N=4 power levels 0, 1, 2 and 3 (in dimensionless units). Note, that in comparison to the assumptions for Figure 1 each power level of the analog optical in- put signal Sin is here four times as high for illustrative convenience, which would corre- spond to a simple rescaling / normalization of the x-axis in Figures 1b and 1c by a factor of 4.

[0224] The first portion S1 passes through the phase shifter 1 such that the amplitude (note that power is the absolute square of the amplitude) in the first portion of the analog optical input signal becomes multiplied with the phase factor eiπ= — 1 (note, that a real amplitude multiplied with a phase factor may be denoted as complex amplitude). The phase-shifted first portion S1 then becomes the input to the optical nonlinear threshold device 2.0 of the first arm 1.1 with the input-output transfer characteristic shown in Figure 2b. The input power values are thus still 0, 1, 2, 3 (the phase shift only shows up in a multiplicative factor of the amplitude but not in the power). Correspond- ingly, the output, i.e., the phase-shifted (and here also amplitude-modulated) first por- tion of the analog optical input signal Sil has approximated power values 0, 1, 1, 1 since according to Figure 2b the output power is 0 for input power 0 and the output power is (approximately) 1 for the input power 1 and higher (saturated regime). This corresponds to amplitude values A=0, 1, 1, 1 with phase factor -1.

[0225] The second portion S2 first passes through the linear attenuator 3.0 of the second arm 1.2 and thus its power values 0, 1, 2, 3 get first transformed to 0, 1 / 3, 2 / 3, 1 which then correspond to the input power values for the optical nonlinear threshold device 2.0 of the second arm 1.2 with input-output transfer characteristic again shown in Fig- ure 2b. When the attenuated second portion S2 has passed the optical nonlinear threshold device 2.0, the power levels of the amplitude-modulated second portion S21 correspond approximately to 0, 0, 1, 1. Here, we have approximately set output power Pout=0 for input power Pin=0 and 1 / 3 and output power Pout=1 for input power Pin=2 / 3 and 1 in approximate correspondence with Figure 2b. In terms of amplitudes, the am- plitude levels of the amplitude-modulated second portion S21 are B=0, 0, 1, 1.

[0226] At the third node N3, the phase-shifted (and amplitude-modulated) first portion S11 with amplitude levels A=0, 1, 1, 1 and phase factor -1 destructively interferes with the amplitude-modulated second portion S21 with amplitude levels B=0, 0, 1, 1 such that the resulting first interference signal has amplitude values -A+B = 0, -1, 0, 0 corre- sponding to power levels 0, 1, 0, 0. Note, that when two electromagnetic signals are superposed and interfere amplitudes (with corresponding phase factors) and not power values have to be added.

[0227] Consequently, the first interferometer is configured with a non-monotonic input-out- put power transfer characteristic that has a negative slope along which the power of the output power of the first interference signal decreases (from 1 to 0) when the power of the analog optical input signal increases (from 4 to 8).

[0228] In the embodiment shown in Figure 2a, the first interference signal is then additionally brought into interference with the amplitude-modulated fourth portion of the analog optical input signal S41 at the auxiliary node N3.1.

[0229] In particular, the fourth portion S4 of the analog optical input signal with power levels 0, 1, 2 and 3 first passes through the linear attenuator 3.0 of the additional arm of the first optical channel to produce an output with power levels 0, 1 / 5, 2 / 5, 3 / 5. This at- tenuated fourth portion then passes through the optical nonlinear threshold device 2.0 of the additional arm to generate the output power levels 0, 0, 0, 1, where we have ap- proximately set output power Pout=0 for input power Pin=0, 1 / 5 and 2 / 5 (lower than in- put power threshold 0.5) and output power Pout=1 for input power Pin=3 / 5 (higher than input power threshold 0.5) in approximate correspondence with Figure 2b. The corre- sponding amplitude levels also read C= 0, 0, 0, 1.

[0230] The interference of the first interference signal with the amplitude-modulated fourth portion of the analog optical input signal S41 at node N3.1 then generates the another interference signal with amplitude values -A+B+C=0, -1, 0, 1 corresponding to power levels 0, 1, 0, 1 (i.e., absolute square of the amplitude -A+B+C).

[0231] In Figure 2a, the first to-be-digitized optical output signal Sout1 is identical with the an- other interference signal and thus also has the power value levels 0, 1, 0, 1. Therefore, the first optical channel as a whole has a non-monotonic input-output power transfer characteristic, where the output power has a positive slope (from 0 to 1), a negative slope (from 1 to 0) and again a positive slope (from 0 to 1) in order of increasing input power of the analog optical input signal (from 0 to 12) similar to the behaviour shown in Figure 1c (but with rescaled input power).

[0232] The third portion of the analog optical input signal S3 in the second optical channel with power levels 0, 1, 2 and 3 first gets attenuated by the linear attenuator 3.0 of the second optical nonlinear threshold device 2.2 to power levels 0, 1 / 3, 2 / 3 and 1 which are then put into the optical nonlinear threshold device 2.0 of second optical nonlinear threshold device 2.2 generating as output the amplitude-modulated third portion of the analog optical input signal S31 with power levels 0, 0, 1, 1.

[0233] In Figure 2a, the amplitude-modulated third portion of the analog optical input signal S31 is identical with the second to-be-digitized optical output signals Sout2. Therefore, the second optical channel as a whole has a monotonic input-output power transfer characteristic, where the output power has a positive slope with a monotonic increase of output power (from 0 to 1) as a function of increasing input power of the analog op- tical input signal Sin for all power levels from 0 to 12 of the analog optical input signal Sin in correspondence with Figure 1b (but with rescaled input power).

[0234] The bit value of the first bit B1 then simply corresponds to the power level of the first to-be-digitized optical output signal Sout1, i.e., 0, 1, 0, 1. The bit value of the second bit B2 then simply corresponds to the power level of the second to-be-digitized optical output signals Sout2, i.e., 0, 0, 1, 1.

[0235] Of course, the conversion of the power of the first and second to-be-digitized optical output signals Sout1, Sout2 to bit values of the first and second bit B1, B2 can also comprise a (simple) conversion rule, e.g., a rounding to next integer operation and / or a normalization operation and / or a simple threshold rule or similar.

[0236] In summary, according to the embodiment shown in Figures 2a and 2b a non-trivial be- haviour of the respective output power in the two optical channels as a function of in- put power is achieved by a clever design of the optical A / D converter combining opti- cal non-linear threshold behaviour with optical interference. In the following, further embodiments will be described that are based on a similar principle.

[0237] It is also note that the embodiment shown in Figure 2a can be scaled up for a higher number of power levels, e.g., N greater than 4, for encoding more bits in the analog optical input signal, e.g., by adding additional arms / branches / waveguides with corre- sponding additional nonlinear optical components and interferometric arms that are additionally split / branching out from the first node N1 (star-like topology). In this way, several optical A / D converter may be combined in one coherent design, e.g., in order to realize an optical N-level pulse amplitude modulation receiver, PAM-N signal re- ceiver. The attenuation factors of the linear attenuators in the respective arms / branches / waveguides may then be chosen and / or reconfigured depending on the number of power levels N. However, when the number of power levels increases, the number of required linear attenuators and optical nonlinear threshold devices etc also increases quickly, thereby possibly leading to comparably high power loss / dissipa- tion.

[0238] Therefore, the following embodiments describe alternative designs for the optical A / D converter that are more favourable when it comes to scalability.

[0239] Figure 3

[0240] Figure 3a shows an alternative embodiment of an optical A / D converter according to a first configuration. The embodiment shown in Figure 3a comprises less attenuators, waveguides and optical nonlinear threshold devices as compared to the embodiment shown in Figure 2a.

[0241] According to the embodiment shown in Figure 3a, the first optical channel comprises the input waveguide, the first node Nl, the first and second waveguide, the second node N2, the first intermediary waveguide, the third node N3 and the first output waveguide.

[0242] The second optical channel comprises the input waveguide, the first node Nl, the sec- ond waveguide, the second node N2 and the second output waveguide.

[0243] The first interferometer comprises the first node Nl, the first and second waveguide, the second node N2, the first intermediary waveguide and the third node N3. The first arm 1.1 of the first interferometer comprises the first waveguide and the phase shifter 1. The second arm 1.2 of the first interferometer comprises the second waveguide, the second node N2, the first intermediary waveguide and the first optical nonlinear threshold device 2.1.

[0244] More specifically, in Figure 3a, the input node Nl is configured to split the analog opti- cal input signal Sin into the first S1 and second S2 portion of the analog optical input signal. The input node Nl also splits the input waveguide into the first and second waveguide. Here, the first node Nl is the same as the input node Nin.

[0245] In Figure 3a, the first optical channel comprises the first interferometer configured to receive the first portion of the analog optical input signal S1 in the first arm 1.1 of the first interferometer. The first arm 1.1 comprises the phase shifter 1 to shift the phase of the first portion of the analog optical input signal S1. The phase shifter 1 is configured to shift the phase of the first portion of the analog optical input signal S1 byπ.

[0246] The first interferometer is also configured to receive the second portion of the analog optical input signal S2 in the second arm 1.2 of the first interferometer. In this embodi- ment, the second arm 1.2 comprises the first optical nonlinear threshold device 2.1 shown in Figure 3b. The first optical nonlinear threshold device 2.1 is configured to re- ceive and modulate the amplitude of the second portion of the analog optical input signal S2 to generate the amplitude-modulated second portion of the analog optical input signal S21.

[0247] In Figure 3a, the second node N2 is configured to split the amplitude-modulated sec- ond portion of the analog optical input signal S21 into the first portion of the ampli- tude-modulated second portion of the analog optical input signal S211 and the second portion of the amplitude-modulated second portion of the analog optical input signal S212. The second node N2 splits the second waveguide into the first intermediary waveguide and the second output waveguide. The first intermediary waveguide con- nects the second node N2 and the third node N3 and is configured to receive the first portion of the amplitude-modulated second portion of the analog optical input signal S211. The second output waveguide is connected to the second node N2 and config- ured to receive the second portion of the amplitude-modulated second portion of the analog optical input signal S212.

[0248] The third node N3 is configured to combine and interfere the phase-shifted first por- tion of the analog optical input signal Sil and the first portion of the amplitude-modu- lated second portion of the analog optical input signal S211 into the first interference signal. The third node N3 combines the first waveguide and the first intermediary waveguide into the first output waveguide.

[0249] In other words, the first optical channel is configured to generate the first interference signal by interfering the phase-shifted first portion of the analog optical input signal Sil from the first arm 1.1 with the first portion of the amplitude-modulated second portion of the analog optical input signal S211 coming from the second arm 1.2 of the first interferometer and the second node N2.

[0250] The first optical channel is further configured to generate the first portion of the to-be- digitized optical output signal Sout1 from the first interference signal. Here, the first portion of the to-be-digitized optical output signal Sout1 is identical with the first interference signal.

[0251] The optical A / D converter also comprises a second optical channel for generating from the analog optical input signal Sin a second portion of the to-be-digitized optical out- put signal Sout2 providing a second bit B2.

[0252] Here, the first and second optical channel spatially overlap significantly at the second arm 1.2 of the first interferometer and share the input waveguide, the first node Nl, the second waveguide, the first optical nonlinear threshold device 2.1 and the second node N2. Here, the first and second optical channel hot comprise the first optical non- linear threshold device 2.1.

[0253] The second optical channel is configured to generate the second portion of the to-be- digitized optical output signal Sout2 from the second portion of the amplitude-modu- lated second portion of the analog optical input signal S212.

[0254] In this embodiment, the second optical channel also comprises a linear attenuator 3.0 in the second output waveguide with an attenuation factor 1 / 2. Thus, the linear atten- uator 3.0 in the second output waveguide is configured to attenuate the second por- tion of the amplitude-modulated second portion of the analog optical input signal S212 by a factor 1 / 2 to generate the second portion of the to-be-digitized optical output sig- nal Sout2. In other words, the second portion of the to-be-digitized optical output sig- nal Sout2 is obtained from the second portion of the amplitude-modulated second portion of the analog optical input signal S212 coming from the second arm 1.2 of the first interferometer and from the second node N2.

[0255] In Figure 3b, the first optical nonlinear threshold device 2.1 comprises in series a linear attenuator 3.0 configured to attenuate the power of the second portion of the analog optical input signal S2, an optical nonlinear threshold device 2.0 with a normalized (monotonic) input-output power transfer characteristic as shown in Figure 2b and a linear amplifier 4.0. Here, the linear attenuator 3.0 is configured with an attenuation factor 1 / 3 and the linear amplifier 4.0 is configured with an amplification factor / gain 4.

[0256] However, it is noted that linear attenuation with the linear attenuator 3.0 merely has the function to rescale the input power of the optical nonlinear threshold device 2.0 into an appropriate range. The linear amplifier 4.0 has the function to rescale the out- put power of the optical nonlinear threshold device 2.0 into an appropriate range for further processing. Alternatively, linear attenuation and linear amplification may also be skipped and / or realized internally in the first optical nonlinear threshold device 2.1 or in other components of the A / D converter to ensure a proper scaling of the respec- tive input power(s).

[0257] The working principle of the embodiment of the optical A / D converter shown in Figure 3a can now be explained in more detail as follows:

[0258] Assuming without loss of generality that the analog optical input signal Sin features N=4 different power levels with input powers 0, 2, 4 and 6 (in dimensionless units), re- spectively. The first / input node Ml comprises an equal power splitter such that each one of the first S1 and second S2 portion of the analog optical input signal features N=4 power levels with power values 0, 1, 2 and 3 (in dimensionless units).

[0259] The first portion S1 passes through the phase shifter 1 such that the amplitude be- comes multiplied with the phase factor eiπ= — 1. Correspondingly, the output of the phase shifter 1, i.e., the phase shifted first portion of the analog optical input signal S11 also has power values 0, 1, 2, 3 corresponding to amplitude values A=0, 1, √2, √3 with phase factor -1.

[0260] The second portion S2 first passes through the linear attenuator 3.0 of the first optical nonlinear threshold device 2.1 and thus its power values 0, 1, 2, 3 get first transformed to 0, 1 / 3, 2 / 3, 1 which then correspond to the input power values for the optical non- linear threshold device 2.0 of the first optical nonlinear threshold device 2.1 with in- put-output transfer characteristic again shown in Figure 2b. When the attenuated sec- ond portion S2 has passed the optical nonlinear threshold device 2.0, the power levels correspond approximately to 0, 0, 1, 1. Here, we have approximately set output power Pout=0 for input power Pin=0 and 1 / 3 and output power Pout=1 for input power Pin=2 / 3 and 1 in approximate correspondence with Figure 2b. The output of the optical nonlin- ear threshold device 2.0 then becomes amplified by the linear amplifier 4.0 in Figure 3b to power values 0, 0, 4, 4 of the amplitude-modulated second portion of the analog optical input signal S21.

[0261] At the second node N2 comprising an equal splitter the amplitude-modulated second portion S21 gets split into first S211 and second S212 portion of the amplitude-modu- lated second portion each with power levels 0, 0, 2, 2 corresponding to amplitude lev- els 0, 0, √2, √2. At the third node N3, the phase-shifted first portion Sil with amplitude levels A=0, 1, √2, √3 and phase factor -1 destructively interferes with the first portion of the ampli- tude-modulated second portion S211 with amplitude levels B=0, 0, √2, √2 such that the resulting first interference signal has amplitude values -A+B = 0, -1, 0, √2-√3 corre- sponding approximately to power levels 0, 1, 0, 0.1.

[0262] Here, the first interference signal is identical to the first portion of the to-be-digitized optical output signal. The conversion to bit values then comprises a simple threshold rule, e.g., according to which power values equal or below a simple conversion thresh- old 0.05 are converted to bit value 0 and power values higher than 0.05 are converted to bit value 1 resulting in bit values 0, 1, 0, 1 as shown for the first bit B1 in Figure 3a.

[0263] The second portion of the amplitude-modulated second portion S212 with with power levels 0, 0, 2, 2 becomes attenuated by the linear attenuator 3.0 in the second output waveguide of the second optical channel by an attenuation factor ½ resulting in a sec- ond portion of the of the to-be-digitized optical output signal Sout2 with power levels 0, 0, 1, 1. The bit value of the second bit B2 then simply correspond to the power level of the second to-be-digitized optical output signal Sout2, i.e., 0, 0, 1, 1.

[0264] The optical A / D converter shown in Figures 3a and 3b is more power efficient as com- pared to the embodiment in Figures 2a and 2b as can be appreciated by comparing the different input power values of the analog optical input signal Sin in Figure 2a (with in- put power values 0, 4, 8, 12) and Figure 3a (with lower input power values 0, 2, 4, 6) leading to the same output power values in the first and second to-be-digitized output signals Sout1, Sout2.

[0265] Figure 4

[0266] Figures 4a and 4b show an embodiment of the optical A / D converter which is similar to the embodiment shown in Figures 3a and 3b, but comprises uneven splitter at the first node N1 and the second node N2. In particular, when comparing the power levels of the analog input signal Sin in Figure 3a with the power values of the first and second portion of the to-be-digitized optical output signals Sout1, Sout2 in Figure 3a, it can be recognized that some power loss has occurred. This power loss can be further limited by using the uneven splitters as shown in Figure 4a.

[0267] In Figure 4a, the input node Ml is configured to unevenly split the analog optical input signal Sin into the first portion of the analog optical input signal S1 and the second portion of the analog optical input signal S2 such that the power of the first portion S1 is higher than the power of the second portion S2. In this way, the stronger first por- tion S1 passes th rough the phase shifter 1 with almost no loss while the weaker second portion S2 passes through the first optical nonlinear threshold device 2.1 and is subject to attenuation, thresholding and amplification. Thus, the efficiency of the optical A / D converter is increased since power is wasted mainly due to interference and artificially induced losses, e.g., through the use of an attenuator, are limited.

[0268] More specifically, the input node Ml is configured with a power splitting ratio e: 1, i.e., where e « 1 is a small number smaller than 1. An analog optical input signal Sin with power levels 0(1 + e), 1(1 + e), 2(1 + e), 3(1 + e) is splitted at the first node Ml into the first portion S1 with power levels 0, 1, 2, 3 and the second portion with power levels Oe, le, 2e, 3e.

[0269] After the first portion S1 has passed through the phase shifter 1, the power levels of the phase-shifted first portion of the analog optical input signal Sil are still 0, 1, 2, 3 with phase-shifted amplitude levels A=0, 1, √2, √3 and phase factor -1.

[0270] Figure 4b shows an appropriate design of the first optical nonlinear threshold device 2.1 comprising in series a linear attenuator 3.0 with attenuation factor l / (3e), an (cali- brated) optical nonlinear threshold device 2.0 with input-output power transfer char- acteristic as shown in Figure 2b and a linear amplifier 4.0 with amplification factor 3 (gain).

[0271] After the second portion S2 has passed through the first optical nonlinear threshold device 2.1, the power levels of the amplitude-modulated second portion of the analog optical input signal S21 are 0, 0, 3, 3 based on similar assumptions as discussed in the context of Figure 3a.

[0272] In Figure 4a, the second node N2 also comprises an uneven splitter configured to split the amplitude-modulated second portion of the analog optical input signal S21 with a power ratio 2:1 into the first portion of the amplitude-modulated second portion of the analog optical input signal S211 with higher power levels 0, 0, 2, 2 and correspond- ing amplitude levels 0, 0, √2, √2 and the second portion of the amplitude-modulated second portion of the analog optical input signal S212 here being identical with the second to-be-digitized optical output signal Sout2 with lower power levels 0, 0, 1, 1. Finally, at the third node N3 the first portion of the amplitude-modulated second por- tion of the analog optical input signal S211 with amplitude levels 0, 0, √2, √2 destruc- tively interferes with the phase-shifted first portion of the of the analog optical input signal Sil with amplitude levels 0, 1, √2, √3 and phase factor -1 to generate the first interference signal here being identical with the first portion of the to-be-digitized op- tical output signal Sout1 with amplitude levels 0, -1, 0, √2 — √3 corresponding to ap- proximate power levels 0, 1, 0, 0.1.

[0273] Again, the conversion to bit values then comprises a simple threshold rule according to which power values equal or below 0.05 are converted to bit value 0 and power values higher than 0.05 are converted to bit value 1 resulting in bit values 0, 1, 0, 1 as shown for the first bit B1 in Figure 4a. The bit value of the second bit B2 simply correspond to the power level of the second to-be-digitized optical output signal Sout2, i.e., 0, 0, 1, 1.

[0274] The optical A / D converter shown in Figures 2a, 3a and 4a each correspond to a PAM-4 signal receiver with n=2 bits encoded in N=2n=4 power levels of the analog optical in- put signal Sin.

[0275] A further advantage of the proposed design in Figures 3a and 4a is the scalability with respect to higher level encoding that would allow more bits corresponding to more power levels to be encoded in the analog optical input signal Sin.

[0276] Figure 5

[0277] Figure 5a shows an embodiment of a PAM-N signal receiver for converting an analog optical input signal with an arbitrary number N=2npower levels into a respective digital output signal for n bits, where is an integer. The digital output signal can be obtained from n portions of to-be-digitized optical output signals each corresponding to one of the n bits, i.e., each corresponding to a PAM-2 signal.

[0278] The PAM-N signal receiver comprises (n-1) optical A / D converter, e.g., corresponding to one of the embodiments shown in Figure 3a, 4a , 8b or 9a. All (n-1) optical A / D con- verter are coupled in series, wherein a pair of neighboring optical A / D converter in the series is coupled via their first optical channels.

[0279] In particular, the analog optical input signal Sin received by the first optical channel of an optical A / D converter in the series is obtained from the first portion of the to-be- digitized output signal Sout1 of the first channel of the directly preceding optical A / D converter in the series. Correspondingly, the analog optical input signal Sin for the i-th A / D converter with i=l,..,n-1, is obtained from the first portion of the to-be-digitized optical output signal Sout1 of the (i+l)-th optical A / D converter. For example, in Figure 5a the analog optical input signal Sin for the first A / D converter AD1 (i=1) is obtained from the first portion of the to-be-digitized optical output signal Sout1 of the second optical A / D converter AD2 (i=2), wherein the second optical A / D converter AD2 directly precedes the first optical A / D converter AD1.

[0280] The analog optical input signal Sin of the (n-1)-th optical A / D converter in the series is the information carrying optical analog optical input signal with N-levels that is to be converted into bits and that corresponds to the analog optical input signal Sin of the PAM-N signal receiver. In other word, the first optical A / D converter is the last optical A / D converter in the series and the (n-1)-th optical A / D converter is the first optical A / D converter in the series in the propagation direction of the information carrying an- alog optical input signal Sin.

[0281] The first bit B1 of the first optical A / D converter AD1 corresponds to a first bit of the PAM-N signal receiver. The second bit B2 of the first optical A / D converter AD2 corre- sponds to a second bit B2 of the PAM-N signal receiver. The second bit B2 of the sec- ond optical A / D converter AD2 corresponds to a third bit B3 of the PAM-N signal re- ceiver etc. Therefore, the first optical channel of the first optical A / D converter AD1 provides the first bit of the PAM-N signal receiver and the second optical channel of the first optical A / D converter AD1 provides the second bit of the PAM-N signal re- ceiver. The other (n-2) bits of the PAM-N signal receiver, i.e., the 3rdbit to the nthbit, are provided by the (n-2) second optical channels of the (n-2) optical A / D converter other than the first optical A / D converter AD1 in the series.

[0282] Figure 5b shows an embodiment of a PAM-8 signal receiver as an example of a PAM-N level signal receiver with N=8 power levels, n=3 bits and two optical A / D converter, the first optical A / D converter AD1 and the second optical A / D converter AD2. In Figure 5b, the first AD1 and second AD2 optical A / D converter each correspond to the embodi- ment of Figure 3a.

[0283] In Figure 5c, each first optical nonlinear threshold device 2.1 of the first AD1 and sec- ond AD2 optical A / D converter comprises in series a linear attenuator 3.0, a (cali- brated) optical nonlinear threshold device 2.0 and a linear amplifier 4.0 as shown. Ad- ditionally, each input waveguide in Figure 5b comprises a linear amplifier 4.0 for amplifying the respective analog optical input signals Sin of the first AD1 and second AD2 optical A / D converter in the series and each second optical output waveguide comprises a linear attenuator 3.0 for attenuating the respective second portion of the to-be-digitized optical output signals Sout2. Note that in Figure 5b, the analog optical input signal Sin of the second optical A / D converter AD2 corresponds to the analog op- tical input Signal Sin of the PAM-8 signal receiver.

[0284] The attenuation and amplification (gain) factors of the linear attenuators and amplifi- ers in Figures 5b and 5c can be suitably chosen in dependence of the choice of power levels in the information carrying analog optical input signal Sin of the PAM-8 signal re- ceiver to ensure that the power levels of the first and second portions of the to-be-dig- itized output signal Sout1, Sout2 of the two optical A / D converter correspond to the three bits B1, B2, B3 of the PAM-8 signal receiver with bit values as shown in Figure 5b.

[0285] Theses linear optical components merely rescale / normalize the input power at the var- ious stages of the conversion to appropriate absolute values. They can also be re- moved or included in the nonlinear optical components and / or appropriate scaling can also be ensured by properly calibrating the first optical nonlinear threshold devices 2.1.

[0286] The working principle of the embodiment of the PAM-8 signal receiver with the first AD1 and second AD2 optical A / D converter as shown in Figure 5b can now be ex- plained in more detail as follows:

[0287] Without loss of generality, but for simplicity of illustration, we assume in the following example that N=8 power levels corresponding to 3-bit information have been encoded in an analog optical input signal Sin such that the corresponding amplitude levels are equally spaced. The assumption of equally spaced amplitude levels rather than equally spaced power levels simplifies the discussion below. However, the same technical ef- fects can be achieved for an analog optical input signal Sin with equally spaced power levels.

[0288] The N=8 amplitude levels of the analog optical input signal Sin of the PAM-8 signal re- ceiver are in dimensionless units 0, 1, 2, 3, 4, 5, 6, 7 (corresponding power levels are obtained by taking the absolute square of the amplitude levels).

[0289] The input waveguide of the second optical A / D converter AD2 receives said analog op- tical input signal Sin. In this example, the input waveguide of the second optical A / D converter AD2 comprises a linear amplifier 4.0 configured to amplify the power of the analog optical input signal Sin with an amplification factor 2 (i.e., multiplying the ampli- tude with a factor of √2).

[0290] The input node Ml of the second optical A / D converter AD2 comprises an equal power splitter such that each one of the first S1 and second S2 portion of the analog optical input signal Sin in the second optical A / D converter AD2 again features N=8 power / am- plitude levels with amplitude values 0, 1, 2, 3, 4, 5, 6, 7 (in dimensionless units).

[0291] The first portion S1 passes through the phase shifter 1 of the second optical A / D con- verter AD2 such that the amplitude becomes multiplied with the phase factor eiπ= — 1. Correspondingly, the output of the phase shifter 1, i.e., the phase shifted first por- tion of the analog optical input signal S11 of the second optical A / D converter AD2 has amplitude levels A=0, 1, 2, 3, 4, 5, 6, 7 with phase factor -1.

[0292] The second portion S2 of the analog optical input signal in the second optical A / D con- verter AD2 first passes through the linear attenuator 3.0 of the first optical nonlinear threshold device 2.1 of the second optical A / D converter AD2. In this example, the lin- ear attenuator 3.0 in the first optical nonlinear threshold device 2.1 of the second opti- cal A / D converter AD2 is configured to attenuate the power by an attenuation factor 1 / 49, i.e., to attenuate the amplitude (!) by a factor of 1 / 7.

[0293] Thus, its output amplitude values are 0, 1 / 7, 2 / 7, 3 / 7, 4 / 7, 5 / 7, 6 / 7, 1 which then cor- respond to the input amplitude values for the optical nonlinear threshold device 2.0 in the first optical nonlinear threshold device 2.1 of the second optical A / D converter AD2 with an input-output transfer characteristic similar to the one shown in Figure 2b, but for the amplitude rather than power, i.e., the output amplitude (square root of the output power) features a crossover from output amplitude 0 to 1 at an input ampli- tude (square root of the input power) of 0.5.

[0294] Thus, when the attenuated second portion S2 has passed the optical nonlinear thresh- old device 2.0, the amplitude levels correspond approximately to 0, 0, 0, 0, 1, 1, 1, 1. Here, we have approximately set the output amplitude to 0 for input amplitudes 0, 1 / 7, 2 / 7 , 3 / 7 and the output amplitude to 1 for input amplitudes 4 / 7, 5 / 7, 6 / 7 and 1. The output of the optical nonlinear threshold device 2.0 then becomes amplified by the linear amplifier 4.0 in Figure 5c. In this example, the linear amplifier 4.0 in the first optical nonlinear threshold device 2.1 of the second optical A / D converter AD2 is con- figured to amplify the amplitude (!) by a factor of 4√2. Thus, the output of the first op- tical nonlinear threshold device 2.1 of the second optical A / D converter AD2 has amplitude values 0, 0, 0, 0, 4√2, 4√2, 4√2, 4√ 2. Said amplitude values correspond to those of the amplitude-modulated second portion of the analog optical input signal S21 of the second optical A / D converter AD2.

[0295] At the second node N2 of the second optical A / D converter AD2 comprising an equal (power) splitter the amplitude-modulated second portion S21 gets split into first S211 and second S212 portion of the amplitude-modulated second portion of the second optical A / D converter AD2 each with amplitude levels 0, 0, 0, 0, 4, 4, 4, 4 (because an amplitude 4√ 2 corresponds to power 32 that is split into 16 and 16 each correspond- ing to amplitude 4).

[0296] At the third node N3 of the second optical A / D converter AD2, the phase-shifted first portion Sil of the second optical A / D converter AD2 with amplitude levels A=0, 1, 2, 3, 4, 5, 6, 7 and phase factor -1 destructively interferes with the first portion of the ampli- tude-modulated second portion S211 with amplitude levels B= 0, 0, 0, 0, 4, 4, 4, 4 such that the resulting first interference signal of the second optical A / D converter AD2 has amplitude values -A+B = 0, -1, -2, -3, 0, -1, -2, -3.

[0297] The second portion of the amplitude-modulated second portion S212 of the second optical A / D converter AD2 with amplitude levels 0, 0, 0, 0, 4, 4, 4, 4 finally becomes at- tenuated by the linear attenuator 3.0 in the second output waveguide of the second optical channel of the second optical A / D converter AD2 by an amplitude attenuation factor 1 / 4 resulting in a second portion of the to-be-digitized optical output signal Sout2 of the second optical A / D converter AD2 with amplitude levels 0, 0, 0, 0, 1, 1, 1, 1 corresponding to the same power levels (in dimensionless units) for providing the third bit B3 of the PAM-8 signal receiver. The bit value of the third bit B3 of the PAM-8 signal receiver then simply correspond to the power level of the second to-be-digitized optical output signal Sout2.

[0298] The first interference signal of the second optical A / D converter AD2 is identical to the first portion of the to-be-digitized optical output signal of the second optical A / D con- verter AD2 and forms the analog optical input signal of the first optical A / D converter AD1 configured to providing the first bit B1 and the second bit B2 of the PAM-8 signal receiver.

[0299] The analog optical input signal of the first optical A / D converter AD1 now has ampli- tude levels 0, 1, 2, 3, 0, 1, 2, 3 with phase factor -1. The input waveguide of the first optical A / D converter AD1 receives said analog optical input signal. In this example, the input waveguide of the first optical A / D converter AD1 comprises a linear amplifier 4.0 configured to amplify the power of the analog op- tical input signal again with an amplification factor 2 (i.e., multiplying the amplitude with a factor of √2).

[0300] The input node N1 of the first optical A / D converter AD1 comprises an equal power splitter such that each one of the first S1 and second S2 portion of the analog optical input signal in the first optical A / D converter AD1 feature N=8 amplitude levels with amplitude values 0, 1, 2, 3, 0, 1, 2, 3 with phase factor -1.

[0301] The first portion S1 passes through the phase shifter 1 of the first optical A / D converter AD1 such that the amplitude become (again) multiplied with the phase factor e™ = — 1. Correspondingly, the output of the phase shifter 1, i.e., the phase shifted first por- tion of the analog optical input signal S11 of the first optical A / D converter AD1 has amplitude levels A=0, 1, 2, 3, 0, 1, 2, 3 with phase factor +1!

[0302] The second portion S2 of the analog optical input signal in the first optical A / D con- verter AD2 first passes through the linear attenuator 3.0 of the first optical nonlinear threshold device 2.1 of the first optical A / D converter AD1. In this example, said linear attenuator 3.0 is configured to attenuate the power by a factor of 1 / 9, i.e., the ampli- tude (!) by a factor of 1 / 3. Note, that the operational parameters of the first optical nonlinear threshold device 2.1 in the first AD1 and second AD2 optical A / D converter are different.

[0303] Thus, the output amplitude values of the linear attenuator 3.0 are 0, 1 / 3, 2 / 3, 1, 0, 1 / 3, 2 / 3, 1 with phase factor -1 which then correspond to the input amplitude values for the optical nonlinear threshold device 2.0 in the first optical nonlinear threshold device 2.1 of the first optical A / D converter AD1 with an input-output transfer characteristic similar to the one shown in Figure 2b but again with input / output power replaced by input / output amplitude, i.e., the output amplitude features a crossover from 0 to 1 for threshold input amplitude around 0.5.

[0304] When the attenuated second portion S2 of the first optical A / D converter AD1 has passed the optical nonlinear threshold device 2.0, the amplitude levels correspond ap- proximately to 0, 0, 1, 1, 0, 0, 1, 1 with phase factor -1. Here, we have approximately set the output amplitude to 0 for input amplitudes 0, 1 / 3 and the output amplitude to 1 for input amplitudes 2 / 3 and 1. The output of the optical nonlinear threshold device 2.0 then becomes amplified by the linear amplifier 4.0 in Figure 5c. In this example, the linear amplifier 4.0 in the first optical nonlinear threshold device 2.1 of the first optical A / D converter AD1 is configured to amplify the power by a factor 8, i.e., the amplitude (!) by a factor of 2√2. Thus, the output of the first optical nonlinear threshold device 2.1 of the first optical A / D converter AD1 has amplitude values 0, 0, 2√2, 2√2, 0, 0, 2√2, 2√2 with phase factor -1. This corresponds to the amplitude-modulated second portion of the analog optical input signal S21 of the first optical A / D converter AD1 with power levels 0, 0, 8, 8, 0, 0, 8, 8.

[0305] At the second node N2 of the first optical A / D converter AD1 comprising an equal split- ter the amplitude-modulated second portion S21 gets split into first S211 and second S212 portion of the amplitude-modulated second portion of the first optical A / D con- verter AD1 each with amplitude levels 0, 0, 2, 2, 0, 0, 2, 2 and phase factor -1.

[0306] At the third node N3 of the first optical A / D converter AD1, the phase-shifted first por- tion S11 of the first optical A / D converter AD1 with amplitude levels A=0, 1, 2, 3, 0, 1, 2, 3 destructively interferes with the first portion of the amplitude-modulated second portion S211 with amplitude levels B= 0, 0, 2, 2, 0, 0, 2, 2 and phase factor -1 such that the resulting first interference signal of the first optical A / D converter AD1 has ampli- tude values A-B = 0, 1, 0, 1, 0, -1, 0, 1 corresponding to power values 0, 1, 0, 1, 0, 1, 0, 1 that can be directly converted to bit values of the first bit B1 of the PAM-8 signal re- ceiver.

[0307] The second portion of the amplitude-modulated second portion S212 of the first opti- cal A / D converter AD1 with amplitude levels 0, 0, 2, 2, 0, 0, 2, 2 and phase factor -1 fi- nally becomes attenuated by the linear attenuator 3.0 in the second output waveguide of the second optical channel of the first optical A / D converter AD1 by a power attenu- ation factor 1 / 4 corresponding to an amplitude attenuation factor 1 / 2 resulting in a second portion of the to-be-digitized optical output signal Sout2 of the first optical A / D converter AD2 with amplitude levels 0, 0, 1, 1, 0, 0, 1, 1 and phase factor - 1 corre- sponding to power levels levels 0, 0, 1, 1, 0, 0, 1, 1 for providing the second bit B2 of the PAM-8 signal receiver.

[0308] Figure 6

[0309] Figure 6a shows input-output power transfer characteristics of a linear amplifier 4.0 and a linear absorber 3.0. Each input-output power transfer characteristic of the linear absorber 3.0 (solid lines) corresponds to a different constant attenuation factor acl with Pout= (l-a) Pin. Each input-output power transfer characteristic of the linear ampli- fier 4.0 (dashed lines) corresponds to a different constant amplification / gain factor g>l with Pout= g Pin-

[0310] All input-output power transfer characteristics describe linear functions. Such linear optical components alone cannot generate the desired non-linear optical output shown in Figures 1b and 1c.

[0311] Figure 6b shows input-output power transfer characteristics of a saturable absorber 3. The dashed line shows an ideal behavior of a saturable absorber 3 that can absorb al- most all input power up to a threshold input power around Pin=1. For higher input power the saturable absorber 3 starts to transmit and the corresponding output power increases almost linearly with input power. However, the output power Poutis still smaller than the input power Pin. The solid line shows a more realistic behavior of the saturable absorber 3 where the threshold is smeared and absorption is reduced gradu- ally.

[0312] The saturable absorber 3 can be described by the (power-dependent) absorption / at- tenuation coefficient / factor:

[0313] Here, αsis the (linear) absorption of the saturable part of the total absorption, αnsis the non-saturable part of the absorption and Psatis the saturation power at high signal strength / power. In Figure 6b, these parameters for the realistic solid line are αs= 1, αns= 0 / Psat= 1 anD the power P corresponds to the input power Pin. These param- eters represent realistic physical values.

[0314] Figure 6c shows input-output power transfer characteristics of a saturable amplifier 4. The dashed line shows an ideal behavior of a saturable amplifier 4 that generates an output power at several multiples of the input power up to a threshold input power around Pin=1. For higher input power the saturable amplifier 4 cannot generate more power and stabilizes / saturates at a constant output power around 10 (in dimension- less units). The solid line shows a more realistic behavior of the saturable amplifier 4 where the threshold is smeared and the output power saturates gradually. The saturable amplifier is described by the (power-dependent) amplification factor (gain):

[0315] Here, gssis the (linear) gain for small signal strength / power and Psatis the saturation power at high signal strength / power. In Figure 6c, these parameters for the solid line are gss= 10 and Psat= 1 and the argument P corresponds to the input power Pin. These parameters represent realistic physical values.

[0316] A saturable absorber 3 and a saturable amplifier 4 can be combined to realize various implementations of an optical nonlinear threshold device with a comparably sharp threshold or threshold region. These devices can be integrated in a same photonic in- tegrated chip platform or be distributed on different chips and coupled together via evanescent coupling, Bragg gratings, photonic wire bonding, or edge coupling.

[0317] For example, the first optical nonlinear threshold device 2.1 and the second optical nonlinear threshold device 2.2 comprise a saturable absorber 3 in series with a satura- ble amplifier 4 as shown in Figures 8b, 9a and 10a. The combination of the respective input-output power transfer characteristics as shown in Figures 6b and 6c in such a se- ries then gives rise to an input-output power transfer characteristic for the first optical nonlinear threshold device 2.1 and the second optical nonlinear threshold device 2.2 similar to the one shown in Figures 1b and 1c. Also, the optical nonlinear threshold de- vice 2.0 shown in Figure 2b as part of the first and / or second optical nonlinear thresh- old devices 2.1, 2.2 may comprise a saturable absorber 3 in series with a saturable am- plifier 4.

[0318] The parameters for the saturable absorber 3 (saturable absorption, non-saturable ab- sorption, saturation power) and the saturable amplifier 4 (linear gain for small power, saturation power) can be calculated analytically. Alternatively, these parameters can be calculated using optimization techniques such as covariance matrix adaptation (CMA) or gradient descent to find optimal values for these components / devices. This can help to properly scale up various implementations, e.g., when the number of power levels used to encode information in the analog optical input signal increases and / or shall be equally spaced in power, e.g., rather than being equally spaced in am- plitude. For example, in order to ensure that the analog optical input signal of the first optical A / D converter AD1 in Figure 5b is obtained from the first interference signal or the first to-be-digitized optical output signal of the second analog A / D converter AD2 with proper power values, the operational parameters of the various optical compo- nents of the first AD1 and second AD2 optical A / D converter may be determined nu- merically using said optimization techniques.

[0319] Figure 7

[0320] In a preferred embodiment, the saturable absorber 3 comprises graphene G or a gra- phene layer G as optical nonlinear material that exhibits saturable absorption due to interband dynamics. Inside graphene G each absorbed photon from the input signal may generate a pair of electron and hole that will recombine and release heat. When many photons arrive within a short time, i.e., at high input power, there is not enough space for further electron-hole pairs to be generated and the material cannot absorb anymore and will start to transmit at a certain threshold input power. Figure 7 shows coupling designs between a waveguide W and graphene G inside the first and second optical nonlinear threshold device 2.1, 2.2. Here, first 2.1 and second 2.2 optical non- linear threshold device comprise are CMOS-based / integrated with a CMOS layered structure. More specifically, Figure 7 shows a cut through the first and second optical nonlinear threshold device 2.1, 2.2, where at least one graphene layer G covers cou- ples to the waveguide W. In all shown examples the graphene layer G is arranged di- rectly on (a surface of) the waveguide W. However, side-coupling and evanescent cou- pling may also be employed.

[0321] Graphene is deposited on top of the waveguide W or at least near a surface of the waveguide W. Light or signals propagating in the waveguide W have an evanescent field extending outside of the waveguide. The evanescent field overlaps with the gra- phene layer that couples to it and cause the desirable nonlinear interactions.

[0322] The upper panel shows a coupling from one side, where the graphene layer G is ar- ranged on top of the waveguide W.

[0323] The middle panel shows a coupling from two sides, where the graphene layer G is ar- ranged and sandwiched between two parts of the waveguide W. The first (bottom) part of the waveguide W couples / provides the input to the first and second optical nonlinear threshold device 2.1, 2.2 while the second (upper) part of the waveguide W provides / receives the output of the first and second optical nonlinear threshold device 2.1, 2.2. In this way, the signal passing through the first and second optical nonlinear threshold device 2.1, 2.2 also has to pass directly through the graphene layer G ensur- ing strong optical nonlinear behavior.

[0324] The lower panel shows two graphene layer G with a passivation layer GP in between the two graphene layers G.

[0325] In case of the first optical nonlinear threshold device 2.1 the waveguide W corresponds to the second waveguide that couples to and is arranged between input node N1 and second node N2 in Figures 3a and 4a or between the input node N1 and the third node N3 in Figure 2a. In case of the second optical nonlinear threshold device 2.2 in Figure 2a the waveguide W corresponds to the third waveguide coupling to input node Nl.

[0326] It is also noted that a saturable amplifier 4 that comprises an active material such as InP or GaAs and a saturable absorber 3 based on Graphene may have to be fabricated on different substrates. The saturable amplifier 4 and the saturable absorber 3 may then be coupled via edge coupling or evanescent waveguide coupling, where two edges of the different substrates with respective waveguide portions are arranged close to each other such that light jumps / tunnels from one waveguide or edge to the other. Alternatively, one may also employ a flip chip architecture for the series cou- pling of the saturable amplifier 4 and the saturable absorber 3, where these two are glued together such that the positions of the relevant waveguides match and couple to each other.

[0327] Figure 8

[0328] Figure 8a shows an embodiment of the first optical nonlinear threshold device 2.1 comprising a saturable absorber 3 and a saturable amplifier 4 in series. In an alterna- tive embodiment, the first optical nonlinear threshold device 2.1 may also comprise a plurality of pairs of a saturable absorber 3 and a saturable amplifier 4 in series.

[0329] Figure 8b shows an embodiment of the optical A / D converter with the first optical non- linear threshold device 2.1 as shown in Figure 8a. As compared to the embodiment shown in Figure 3a, in this embodiment the first and second optical channels are spa- tially separated to a greater extent. The input node NO and the first node N1 are differ- ent and separated by a second intermediary waveguide connecting these two nodes. The first 2.1 and second 2.2 optical nonlinear threshold devices are also different and arranged spatially separated. More specifically, the first optical channel comprises the input waveguide, the input node NO, the second intermediary waveguide, the first node Nl, the first and second waveguide, the third node N3 and the first output waveguide.

[0330] The second optical channel comprises the input waveguide, the input node NO, the third waveguide and the second output waveguide.

[0331] The first interferometer comprises the first node Nl, the first and second waveguide and the third node N3. The first arm 1.1 of the first interferometer comprises the first waveguide. The second arm 1.2 of the first interferometer comprises the second wave- guide.

[0332] The input node NO is configured to split the analog optical input signal Sin into the third S3 and the intermediary portion of the analog optical input signal. The input node NO splits the input waveguide into the third waveguide and the first intermediary waveguide.

[0333] The first node Nl is different from the input node NO. The first node Nl is configured to split the intermediary portion of the analog optical input signal into the first S1 and second S2 portion of the analog optical input signal. In this case, the first node Nl is configured to split the first intermediary waveguide into the first and second wave- guide.

[0334] The third node N3 is configured to combine the phase-shifted first portion of the ana- log optical input signal Sil and the amplitude-modulated second portion of the analog optical input signal S211 into the first interference signal. The third node N3 combines the first and second waveguide into the first output waveguide.

[0335] The second optical nonlinear threshold device 2.2 comprises two pairs of a saturable absorber 3 and a saturable amplifier 4 in series, i.e., the first pair of a saturable ab- sorber 3 and a saturable amplifier 4 is coupled in series with the second pair of a satu- rable absorber 3 and a saturable amplifier 4. The corresponding input-output power transfer characteristic is shown in Figure 8c. The dashed line corresponds to an ideal curve obtained from a combination of respective ideal input-output power transfer characteristics shown in Figures 6b and 6c. The solid line corresponds to a more realis- tic threshold behaviour with a gradual monotonic increase. The shape of the input-out- put power transfer characteristic shown in Figure 8c for the second optical nonlinear threshold device 2.2 also corresponds to the input-output power transfer characteristic of the second optical channel in Figure 8b.

[0336] The second optical nonlinear threshold device 2.2 is configured to modulate the ampli- tude of the third portion of the analog optical input signal S3. The amplitude-modu- lated third portion of the analog optical input signal S31 is identical with the second portion of the to-be-digitized optical output signal Sout2.

[0337] In Figure 8c, the parameters for the second optical nonlinear threshold device 2.2 are as follows (from left to right): First saturable absorber 3 of the first pair in the series: αs= 0.3 and αns= 0.7 and Psat= 5; First saturable amplifier 4 of the first pair in the series: gss= 0.66 and Psat= 30; Second saturable absorber 3 of the second pair in the series: αs= 1 and αns= 0 and Psat= 5; Second saturable amplifier 4 of the sec- ond pair in the series: gss= 100 and Psat= 1.3. Here, the first pair precedes the sec- ond pair in signal propagation direction. These parameters represent realistic physical values that are obtained via an optimization algorithm employing a covariance matrix adaptation evolution strategy (CMA-ES).

[0338] Figure 8d shows an input-output power transfer characteristic of the first interferome- ter that is similar in shape to the input-output power transfer characteristic of the first optical channel in Figure 8b (up to rescaling of the x-axis, i.e., the input power). Again, the dashed line corresponds to the ideal behavior with sharp thresholds and the solid line corresponds to a more realistic behavior. Note, that the second threshold at input power Pin=2 is still very sharp also for the realistic solid curve since it is generated through interference (at node N3 in Figure 8b).

[0339] In Figure 8d, the parameters for the first optical nonlinear threshold device 2.1 are as follows: Saturable absorber 3: αs= 1 and αns= 0 and Psat= 5; Saturable amplifier 4 in the series: gss= 57 and Psat= 2.2. These parameters represent realistic physical values that are obtained via an optimization algorithm employing a covariance matrix adaptation evolution strategy (CMA-ES).

[0340] Figure 9

[0341] Figure 9a shows an embodiment of the optical A / D converter, where the second opti- cal nonlinear threshold device 2.2 has been modified with respect to the embodiment shown in Figure 8b. In this embodiment, the second optical nonlinear threshold device 2.2 comprises a sec- ond interferometer. The second interferometer comprises the intermediate node Nl, the first and second further waveguide and the fourth node N4.

[0342] The intermediary node Nl is configured to split the third portion of the analog optical input signal S3 into a first and a second portion of the third portion of the analog opti- cal input signal. The intermediary node Nl splits the third waveguide into the first fur- ther and the second further waveguide.

[0343] The first arm of the second interferometer comprises the first further waveguide and a saturable amplifier 4 in series with a phase shifter 1. The saturable amplifier 4 in series with a phase shifter 1 can also be denoted as a non-linear phase shifter. The saturable amplifier 4 in series with a phase shifter lis configured to amplify and phase-shift the first portion of the third portion of the analog optical input signal.

[0344] The second arm of the second interferometer comprises the second further waveguide and a saturable absorber 3 in series with a saturable amplifier 4. The saturable ab- sorber 3 in series with a saturable amplifier 4 is configured to modulate the amplitude of the second portion of the third portion of the analog optical input signal.

[0345] The fourth node N4 is configured to combine the (amplified and) phase-shifted first portion of the third portion of the analog optical input signal from the first arm of the second interferometer with the amplitude-modulated second portion of the third por- tion of the analog optical input signal from the second arm of the second interferome- ter into a second interference signal. The fourth node N4 is configured to combine the first further and second further waveguide into the second output waveguide.

[0346] Here, the second interference signal is identical to the amplitude-modulated third por- tion of the analog optical input signal S31, because the second interferometer is con- figured to modulate the amplitude of the third portion of the analog optical input sig- nal S3. Again, the amplitude-modulated third portion of the analog optical input signal S31 is identical with the second portion of the to-be-digitized optical output signal Sount2.

[0347] In this embodiment, the second interference mediated by the second interferometer in the second optical channel is used here to further sharpen the transition / threshold from low output power (0) to high output power (1) in the saturated regime as shown in Figure 9b. Figure 9b shows an input-output power transfer characteristic of the second optical nonlinear threshold device 2.2 or, equivalently, the second interferometer of the sec- ond optical channel shown in Figure 9a. Here, the input-output power transfer charac- teristic is not monotonic over the whole range of input power values / levels Pin=0...3. However, compared to the solid line in Figure 8c, the solid line in Figure 9b is some- what closer to the ideal dashed curve in particular near and in the threshold region be- tween input power values Pin= 1 and Pin=2. Note, that the dashed curve in Figure 9b has been reproduced from Figure 8c for better comparison.

[0348] In Figure 9b, the parameters for the second optical nonlinear threshold device 2.2 are as follows: Saturable absorber 3 in the (upper) second arm of the second interferome- ter: αs= 0.1 and αns= 0.9 and Psat= 5; Saturable amplifier 4 in the (upper) second arm of the second interferometer gss= 86 and Psat= 8.5; Saturable amplifier 4 in the (lower) first arm of the second interferometer gss= 1.44 and Psat= 30. These parameters represent realistic physical values that are obtained via an optimization al- gorithm employing a covariance matrix adaptation evolution strategy (CMA-ES).

[0349] Figure 9c shows the input-output power transfer characteristic of the first optical chan- nel in Figure 9a and is identical to Figure 8d.

[0350] Figure 10

[0351] For an improved scalability and ease of manufacturing it can also be advantageous to use the same type of unit cell / block in the various components and branches / channels of the optical A / D converter and set the operational parameters of the devices in such a unit cell individually to different values to achieve a desired functionality of the re- spective unit cell in each branch / channel. The operational parameters can be deter- mined through simulations to make sure that a similar functionality as shown in Fig- ures 1b, 1c for first and second Bit B1, B2, respective first and second optical channel, is ensured.

[0352] Figure 10a shows an embodiment of such a unit cell. The unit cell comprises in series a first linear attenuator 3.0, a saturable absorber 3, a saturable amplifier 4, a second lin- ear attenuator 3.0 and a (tunable / configurable) phase shifter 1.0.

[0353] The attenuation factors of the first and second linear attenuators 3.0, the input-output power transfer characteristic of the saturable absorber 3 in series with the saturable amplifier 4 and / or the phase shift of the (tunable / configurable) phase shifter 1.0 are operational parameters that can be configured differently depending on where the re- spective unit cell is arranged in the circuit shown in Figure 10b. For example, the phase shifter 1.0 can produce a phase shift between 0 and 2 π . These components can also be reordered within the unit cell with only small changes to the effectiveness of the unit cell. For example, in some branches / waveguides or optical channels of the optical A / D converter a configurable phase shifter 1.0 may merely be used to compensate for un- desirable phase shifts.

[0354] Figure 10b shows an embodiment of the optical A / D converter comprising several unit cells as shown in Figure 10a with different operational parameters. This embodiment can be viewed as a modification of the embodiment shown in Figure 3a.

[0355] Here, the phase shifter 1 of Figure 3a has been replaced with the unit cell of Figure 10a. The first optical nonlinear threshold device 2.1 here comprises three unit cells as shown in Figure 10a and has a more complex structure as described further below.

[0356] In Figure 10b, the first optical channel comprises the input waveguide, the input node Nl, the first waveguide, the second waveguide, the intermediary node Nl, the first fur- ther waveguide, the second further waveguide, the fourth node N4, the third interme- diary waveguide, the second node N2, the first intermediary waveguide, the third node N3 and the first output waveguide. Here, the input node Nl is identical to the first node.

[0357] The first interferometer comprises the input node Nl, the first waveguide, the second waveguide, the intermediary node Nl, the first further waveguide, the second further waveguide, the fourth node N4, the third intermediary waveguide, the second node N2, the first intermediary waveguide, the third node N3.

[0358] The first arm of the first interferometer comprises the first waveguide and a unit cell as phase shifter 1.

[0359] The second arm of the first interferometer comprises the second waveguide, the sec- ond interferometer, the third intermediary waveguide, the second node N2 and the first intermediary waveguide. The second interferometer comprises the intermediary node Nl.

[0360] The second optical channel comprises the input waveguide, the input node Nl, the second waveguide, the second interferometer, the third intermediary waveguide, the second node N2 and the second output waveguide.

[0361] In Figure 10b, the first and the second optical channel each comprise the second inter- ferometer. The second interferometer comprises the intermediate node Nl, the first and second further waveguide and the fourth node N4. The first arm of the second in- terferometer comprises the first further waveguide. The second arm of the second in- terferometer comprises the second further waveguide.

[0362] In Figure 10b, the input node Nl is configured to split the analog optical input signal Sin into the first S1 and second S2 portion of the analog optical input signal. The input node Nl splits the input waveguide into the first and second waveguide.

[0363] The intermediary node Nl is configured to split the second S2 portion of the analog op- tical input signal into a first and a second portion of the second portion of the analog optical input signal. The intermediary node Nl splits the second waveguide into the first further and the second further waveguide.

[0364] The first and second optical channel and the second waveguide comprise a unit cell or nonlinear phase shifter as shown in Figure 10a between the input node Nl and the in- termediary node Nl.

[0365] The second portion of the analog optical input signal S2 is amplitude-modulated and / or phase-shifted before it reaches the intermediary node Nl. This may be ac- counted for by noting that the intermediary node Ni is configured to split the ampli- tude-modulated and / or phase-shifted second portion of the analog optical input signal into a first and a second portion of the second portion of the analog optical input sig- nal.

[0366] The fourth node N4 is configured to combine the (amplitude-modulated and / or phase- shifted) first and second portion of the (amplitude-modulated and / or phase-shifted) second portion of the analog optical input signal into a second interference signal. The fourth node N4 is configured to combine the first and second further waveguide into the third intermediary waveguide.

[0367] In other words: The input to the first optical nonlinear threshold device 2.1 is the sec- ond portion of the analog optical input signal S2. The output of the first optical nonlin- ear threshold device 2.1 is the second interference signal as the amplitude-modulated (and possibly phase-shifted) second portion of the analog optical input signal S21. The second node N2 is configured to split the second interference signal S21 into the first S211 and second S212 portion of the amplitude-modulated (and possibly phase- shifted) second portion of the analog optical input signal. The second node N2 splits the third intermediary waveguide into the first intermediary and the second output waveguide. Here, the second interferometer and the third intermediary waveguide may be considered as part of the second waveguide. The second portion of the ampli- tude-modulated (and possibly phase-shifted) second portion of the analog optical in- put signal S212 is identical to the second portion of the to-be-digitized optical output signal Sout2.

[0368] The third node N3 is configured to combine the phase-shifted (and possibly also ampli- tude-modulated) first portion of the analog optical input signal Sil and the first por- tion of the amplitude-modulated (and possibly also phase-shifted) second portion of the analog optical input signal S211 into the first interference signal. The first interfer- ence signal is identical to the first portion of the to-be-digitized optical output signal Sout1.

[0369] It is noted that not all the elements / components shown in Figure 10b are needed in every unit cell to achieve the desired functionality of the first and second optical chan- nel, but using a standardized cell as shown in Figure 10a facilitates ease of fabrication.

[0370] However, all phase shifter 1.0 and other components of the optical A / D converter are configured to ensure an overall phase shift of it at node N3, i.e., between the phase- shifted (and possibly also amplitude-modulated) first portion of the analog optical in- put signal Sil and the first portion of the amplitude-modulated (and possibly also phase-shifted) second portion of the analog optical input signal S211. In this way, de- structive interference with a phase difference of it is ensured for the first interference signal at the output of the first optical channel. Also, all nonlinear components of the optical A / D converter are configured to ensure a threshold behavior for the output power of the second optical channel.

[0371] Features of the different embodiments which are merely disclosed in the exemplary embodiments as a matter of course can be combined with one another and can also be claimed individually.

Claims

Claims1. Optical analog to digital converter for converting an analog optical input signal (Sin) into a digital output signal, comprising: a first optical channel for generating from the analog optical input signal (Sin) a first portion of a to-be-digitized optical output signal (Sout1) providing a first bit (B1), wherein the first optical channel comprises a first interferometer configured to receive a first portion of the analog optical input signal (S1) in a first (1.1) arm of the first interferometer, the first arm (1.1) comprising a phase shifter (1), wherein the phase shifter is configured to shift the phase of the first portion of the analog optical input signal (S1); and receive a second portion of the analog optical input signal (S2) in a sec- ond arm (1.2) of the first interferometer, the second arm (1.2) compris- ing a first optical nonlinear threshold device (2.1), wherein the first opti- cal nonlinear threshold device (2.1) is configured to receive and modu- late the amplitude of the second portion of the analog optical input sig- nal (S2); and generate a first interference signal by interfering the phase-shifted first portion of the analog optical input signal (Sil) from the first arm (1.1) of the first interferometer with the amplitude-modulated second portion of the analog optical input signal (S21) or at least with a first portion of the amplitude-modulated second portion of the analog optical input sig- nal (S211) from the second arm (1.2) of the first interferometer; and the first optical channel is configured to generate the first portion of the to-be- digitized optical output signal (Sout1) from the first interference signal; a second optical channel for generating from the analog optical input signal (Sin) a second portion of the to-be-digitized optical output signal (Sout2) providing a second bit (B2), wherein the second optical channel is configured as a comparator.

2. Optical analog to digital converter according to claim 1, wherein the second optical channel also comprises the first optical nonlinear threshold device (2.1) and is configured to generate the second portion of the to-be-digit- ized optical output signal (Sout2) from at least a second portion of the ampli- tude-modulated second portion of the analog optical input signal (S212) from the second arm (1.2) of the first interferometer; or the second optical channel comprises a second optical nonlinear threshold de- vice (2.2) configured to receive and modulate the amplitude of a third portion of the analog optical input signal (S3), and the second optical channel is config- ured to generate the second portion of the to-be-digitized optical output signal (Sout2) from the amplitude-modulated third portion of the analog optical input signal (S31).

3. Optical analog to digital converter according to one of the preceding claims, wherein the phase of the phase-shifted first portion of the analog optical input signal (Sil) from the first arm (1.1) of the first interferometer is shifted by a phase shift of K with respect to the phase of the amplitude-modulated second portion of the analog optical input signal (S21) or the phase of the first portion of the amplitude-modulated second portion of the analog optical input signal (S211) in order to ensure destructive interference when generating the first in- terference signal.

4. Optical analog to digital converter according to one of the preceding claims, wherein the first optical channel and / or the first interferometer is configured with a non-monotonic input-output power transfer characteristic that has par- tially a negative slope, wherein along the negative slope the power of the first portion of the to-be-digitized optical output signal (Sout1) and / or the power of the first interference signal decreases when the power of the analog optical in- put signal (Sin) and / or the sum of the power of the first (S1) and second (S2) portion of the analog optical input signal increases.

5. Optical analog to digital converter according to claim 2, wherein the second op- tical channel and / or the second optical nonlinear threshold device (2.2) is con- figured with a monotonic input-output power transfer characteristic that has aconsistently positive slope, wherein along the positive slope the power of the second portion of the to-be-digitized optical output signal (Sout2) and / or the power of the amplitude-modulated second portion of the analog optical input signal (S21) monotonously increase as a function of the power of the analog optical input signal (Sin) and / or as a function of the power of the second portion of the analog optical input signal (S2) or along the positive slope the power of the second portion of the to-be-digitized optical output signal (Sout2) and / or the power of the amplitude-modulated third portion of the analog optical input signal (S31) monotonously increase as a function of the power of the analog optical input signal (Sin) and / or as a func- tion of the power of the third portion of the analog optical input signal (S3).

6. Optical analog to digital converter according to one of the preceding claims, wherein the first and / or second optical nonlinear threshold device (2.1, 2.2) comprises at least one saturable absorber (3) in series with a saturable ampli- fier (4).Optical analog to digital converter according to one of the preceding claims, wherein the first and / or second optical nonlinear threshold device (2.1, 2.2) comprises graphene as a saturable absorber (3).

8. Optical analog to digital converter according to one of the preceding claims, wherein the first interferometer comprises a first node (Nl) comprising an une- ven splitter configured to unevenly split at least a portion of the analog optical input signal (Sin) into the first and second portion of the analog optical input signal (S1, S2) such that the power of the first portion of the analog optical in- put signal (S1) is higher than the power of the second portion of the analog op- tical input signal (S2).

9. N-level pulse amplitude modulation signal receiver, PAM-N signal receiver, comprising one or a plurality of optical analog to digital converter according to one of the preceding claims, wherein the number of signal amplitude levels N is an integer equal or greater than 4.

10. N-level pulse amplitude modulation signal receiver according to claim 8, wherein N is equal or greater than 8 and the PAM-N signal receiver comprises at least a first (AD1) and a second (AD2) optical analog to digital converter ac- cording to any one of claims 1 to 8, wherein the first optical channels of the first (AD1) and second (AD2) optical analog to digital converters are coupled in series such that the analog optical input signal (Sin) of the first optical analog to digital converter (AD1) is obtained from the first portion of the to-be-digitized optical output signal (Sout1) of the second optical analog to digital converter (AD2).

11. N-level pulse amplitude modulation signal receiver according to claim 10 com- prising a digital processing unit that is configured to convert the first portion of the to-be-digitized optical output signal of the first optical analog-to-digital con- verter (AD1) to a first bit of the PAM-N signal receiver, the second portion of the to-be-digitized optical output signal of the first optical analog to digital con- verter (AD1) to a second bit of the PAM-N signal receiver and the second por- tion of the to-be-digitized optical output signal of the second optical analog to digital converter (AD2) to a third bit of the PAM-N signal receiver.

12. Method for converting an analog optical input signal (Sin) into a digital output signal, the method comprising: generating in a first optical channel a first portion of a to-be-digitized optical output signal (Sout1) from the analog optical input signal (Sin) for providing a first bit (B1), wherein the generating of the first portion of a to-be-digitized op- tical output signal (Sout1) comprises receiving a first portion of the analog optical input signal (S1) in a first arm (1.1) of a first interferometer, shifting in the first arm (1.1) of the first interferometer the phase of the first portion of the analog optical input signal (S1) by using a phase shifter; and receiving a second portion of the analog optical input signal (S2) in a second arm (1.2) of the first interferometer, andmodulating in the second arm (1.2) of the first interferometer the ampli- tude of the second portion of the analog optical input signal (S2) by us- ing a first optical nonlinear threshold device (2.1); and generating a first interference signal by interfering the phase-shifted first portion of the analog optical input signal (Sil) from the first arm (1.1) of the first interferometer with the amplitude-modulated second portion of the analog optical input signal (S21) or with at least a first portion of the amplitude-modulated second portion of the analog opti- cal input signal (S211) from the second arm (1.2) of the first interferom- eter; and generating the first portion of the to-be-digitized optical output signal (Sout1) from the first interference signal; and generating in a second optical channel a second portion of the to-be-digitized optical output signal (Sout2) from the analog optical input signal (Sin) providing a second bit (B2), wherein the second optical channel is configured as a com- parator.

13. Method of claim 12, wherein the generating of the second portion of the to-be- digitized optical output signal (Sout2) comprises generating the second portion of the to-be-digitized optical output sig- nal (Sout2) from at least a second portion of the amplitude-modulated second portion of the analog optical input signal (S212) from the second arm (1.2) of the first interferometer; or receiving and modulating the amplitude of a third portion of the analog optical input signal (S3) by using a second optical nonlinear threshold device (2.2), and generating the second portion of the to-be-digitized optical output signal (Sout2) from the amplitude-modulated third por- tion of the analog optical input signal (S31).

14. A computer program comprising instruction which, when the program is exe- cuted by a computer, cause the computer using an optical analog to digital con- verter, to carry out the steps of the method according to to claim 12 or 13.

15. A computer-readable data carrier having stored thereon the computer program of claim 14.

Images