All-optical logic with wired-OR multi-mode-interference combiners and semiconductor-optical-amplifier inverters

[0002] This invention relates to all-optical logic elements, and more particularly to the use of semiconductor optical amplifiers (SOA) and multi-mode interference (MMI) devices in optical logic gates.

[0003] Since photons travel faster than electrons, much effort has gone into trying to develop an optical computer or optical logic elements. For example, U.S. Pat. No. 4,262,992 sets forth a variable integrated optical logic element, U.S. Pat. No. 5,434, 701 discloses an all-optical inverter, U.S. Pat. No. 6,035,079 reveals a saturable absorber based optical inverter, and U.S. Pat. No. 5,004,314 sets forth a optical signal modulation device.

[0004] Different approaches have been taken to developing optical logic and other elements of a computer. Most of these approaches have proven to be difficult to produce, expensive, or bulky. Some examples are U.S. Pat. No. 3,849,740 sets forth an integrated optical AND/OR gate, and U.S. Pat. No. 5,315,422 discloses an optical logic element which performs an XOR operation. Optical logic gates are disclosed, for example, in U.S. Pat. Nos. 5,999,283; 5,742,045; 5,541,443; and 5,414,789. Optical flip-flops are described in U.S. Pat. Nos. 4,930,873 and 5,537,243. An optical latching device using an interferometer is shown in U.S. Pat. No. 5,999,284. A switching function using intrapulse Raman scattering is described in U.S. Pat. No. 5,600,479. Other optical switches include those described in U.S. Pat. Nos. 5,970,155; 5,703,975; and 5,119,227. An optical semiconductor image storage apparatus is disclosed in U.S. Pat. No. 6,081,470.

[0005] Another area of optical development is in telecommunications. Developments here have been more commercially successful. The types of optical elements are different since the emphasis is on transmission of optical data, switching, and combining multiple optical data streams. In addition, amplification and regeneration of signals for long-distance transmission is important.

[0006] One challenge in telecommunications is wavelength conversion in the transmission path. Wavelength conversion can be accomplished with a semiconductor optical amplifier (SOA) configured in a cross-gain modulation (XGM) mode. See, for example, the article Wavelength Converters by Allan Kloch et al, IEICE Trans. Electron., Vol. E82-C, No. 8, August 1999, pp. 1475-1486.

[0007] Another device is used to control power in an optical waveguide, and is called a multiple quantum well (MQW). Such a device can also be used as a switch. An example is set forth in U.S. Pat. Nos. 5,754,714 and 5,920,588. An example of a Y-junction is set forth in U.S. Pat. No. 5,580,818, showing the combining of two optical waveguides into one. Other examples of optical elements in communication systems are set forth in U.S. Pat. Nos. 5,798,853; 5,654,812; 5,751,758; and 5,825,517 (showing a cross-connect switch).

[0008] While these devices are somewhat useful, more flexible optical logic gates are desired.

[0009] FIG. 1 illustrates a prior-art wavelength converter.

[0010] FIG. 2 is a graph of the gain of a power-restoring semiconductor optical amplifier (SOA) as a function of input power.

[0011] FIG. 3A shows an optical inverter using an MMI device and a SOA.

[0012] FIG. 3B highlights that the SOA and filter of FIG. 3A act as an optical inverter.

[0013] FIG. 4 shows waveforms of a semiconductor optical amplifier (SOA) operating as an optical inverter.

[0014] FIG. 5 is an optical OR gate.

[0015] FIG. 6 is an optical NOR gate constructed from an optical MMI device and an optical inverter.

[0016] FIG. 7 shows an optical NAND gate using optical inverters and an MMI device.

[0017] FIG. 8 is an optical AND gate.

[0018] The present invention relates to an improvement in all-optical logic gates. The following description is presented to enable one of ordinary skill in the art to make and use the invention as provided in the context of a particular application and its requirements. Various modifications to the preferred embodiment will be apparent to those with skill in the art, and the general principles defined herein may be applied to other embodiments. Therefore, the present invention is not intended to be limited to the particular embodiments shown and described, but is to be accorded the widest scope consistent with the principles and novel features herein disclosed.

[0019] The inventors have recognized that arbitrary logic can be constructed from an inverter and a 2-input OR gate. Large optical computers and smaller optical logic and control units could then be feasible. The inventors have realized that an optical Y-junction can operate like a wired-OR gate in digital-electronic logic, while a semiconductor optical amplifier (SOA) can operate like an optical inverter, providing an inversion function. Together, the optical-OR junction and the optical inverter allow for a wide variety of digital logic gates and functions to be implemented.

[0020] Rather than use a basic Y-junction, a multi-mode interference (MMI) device acts as a junction, combining two optical input signals to generate one or more optical output signals. MMI devices are optical directional couplers that use the self-imaging or interferometric properties of a small optical cavity or planar waveguide to pass optical signals from an input opening to one or more output openings. Constructive and destructive interference of the two inputs occurs within the waveguide. The length and width of the waveguide relative to the input wavelength are design parameters. The locations of output openings in the waveguide can be carefully designed to be located at an interference maxima.

[0021] An optical input wave-guide can be an optical fiber, such as a strand of a fiber-optic cable, or a guide formed on a substrate, such as a groove or a transparent layer formed on the substrate. Two wave-guides A, B can connect together at a 3-way MMI device that has a third wave guide C as an output.

[0022] The optical signals on input wave-guides are combined together or superimposed on each other in the MMI device and travel together down output wave-guide from the MMI device. For example, when a signal with wavelength λ₁is applied to wave-guide A and another signal with a different wavelength λ₂is applied to wave-guide B, the output C has both signals λ₁and λ₂.

[0023] The inventors realize that the MMI device performs the OR logic function in a way similar to a wired-OR gate that combines electrical currents. Although Y-junctions and MMI devices have been used before to combine signals in communication systems, the inventors have recognized the applicability to a binary logic circuit. In particular, when two signals combine, it doesn't matter from a binary-logic point of view that there is twice as much light. As long as the light is more than a high threshold, it is a logic 1, whether it is at the threshold or twice the threshold. When there is light from only one input, the output is still a logic one. When both inputs are dark, or 0, the output is 0. Thus, the Y junction or MMI device functions as a logic OR when placed between two logic inputs and an output.

[0024] FIG. 1 illustrates a prior-art wavelength converter. While SOA's are commonly used to boost optical power, they can also be used as wavelength converters. A modulated signal at wavelength λ₁is to be converted to a different wavelength λ₂. The modulated signal M*λ₁(or Mλ₁) is input from waveguide input 19 to active layer 12A of SOA 12. The modulation M is the signal information that is carried by the optical wave of wavelength λ₁and for binary signals M alternates high and low for different periods of time to carry the data.

[0025] A continuous-wave (CW) signal at wavelength λ₂is also input to active layer 12A of SOA 12 over input 14 waveguide. This signal is labeled as 1*λ₂since the CW signal is not modulated but always in the high (logic 1) state. The two signals can be combined by a Y-junction and then travel through active layer 12A together.

[0026] Normally, the optical weaves that emerge from active layer 12A are amplified by the lasing action. However, carrier depletion in active layer 12A occurs when too much optical power is introduced into active layer 12A. This can occur when both optical input signals are high. When carrier depletion occurs, most of the excited electrons at the higher energy levels have already been stimulated to fall to the lower level. Thus there are not enough excited electrons to emit photons. Amplification of the light waves passing through active layer 12A is thus reduced when carrier depletion occurs.

[0027] Carrier depletion in SOA 12 occurs when both inputs 19, 14 are at full power, and the result is a limiting of optical amplification. Since CW signal 1λ₂is always high, carrier depletion occurs when modulated signal Mλ₁is high, but carrier depletion does not occur when modulated signal Mλ₁is low. The result is that optical amplification by SOA 12 is reduced when M is high, but not when M is low. The optical signal gain emerging from SOA 12 is obtained most when M is low, but least when M is high. When appropriate power levels of signal inputs 19, 14 are applied, modulated signal Mλ₁is inverted.

[0028] This phenomenon is known as cross-gain modulation (XGM). The optical signal emerging from SOA 12 on waveguide output 16 contains both wavelengths λ₁and λ₂. However, since carrier depletion affects both wavelengths, and occurs when M is high, the modulation on wavelength λ₁reduces gain of wavelength λ₂. When M is high, carrier depletion occurs and the gain of CW signal λ₂is reduced. Thus optical output signal λ₂has reduced power when the optical input signals Mλ₁is high. The modulation M is inverted on the λ₂signal that emerges from SOA 12 on output 16.

[0029] Filter 15 blocks wavelength λ₁from output 16, but passes wavelength λ₂. The optical signal output by filter 15 is the λ₂signal inversely modulated by M. Thus output 20 is optical signal M′λ₂.

[0030] FIG. 2 is a graph of the gain of a power-restoring semiconductor optical amplifier (SOA) as a function of input power for optical logic computation. The novelty in the design is such that the SOA operates in a desired saturation region such that the output optical logic 1 level will maintain a target power level 1*P_Hregardless of the input power level. In the FIG. 2, a gain of 0 dB occurs at input power 1*P_H, a gain of 3.01 dB occurs at input power 2*P_H, a gain of 6.02 dB occurs at input power 4*P_Hand a gain of 9.032 dB occurs at input power 8*P_H, etc. The wavelengths used in optical logic computation are not limited to UV-visible or near infrared. The operating power consumed in optical logic computation is small. Of course, the SOA can deviate from the ideal gain curve and still be useful in practical applications.

[0031] FIG. 3A shows an optical inverter using an MMI device and a SOA. The modulated (signal) input 19 is combined with continuous-wave input 14 by MMI device 22. Interference within MMI device 22 combines both wavelengths λ₁and λ₂in the output of MMI device 22.

[0032] The combined output from MMI device 22 is input to SOA 23. The input power of inputs 19, 14 are carefully chosen so that the output from MMI device 22 causes SOA 23 to have gain saturation when input 19 is high (lit), but no gain saturation when input 19 is low (dark). Thus gain saturation in SOA 23 causes its combined input to be logically inverted. Output 16 from SOA 23 is then filtered by filter 15 which passes wavelength λ₂but blocks wavelength λ₁.

[0033] Gain saturation in SOA 23 causes cross-gain modulation (XGM). The optical output signal emerging from SOA 23 on waveguide output 16 contains both wavelengths λ₁and λ₂. However, since carrier depletion affects both wavelengths, and occurs when M is high, the modulation on wavelength λ₁reduces gain of wavelength λ₂. When M is high, carrier depletion occurs and the gain of CW signal λ₂is reduced. Thus optical signal has reduced power when signals Mλ₁is high. The modulation is inverted on the λ₂signal that emerges from SOA 22 on output 16. SOA 23 also operates to restore the power level to a target high/low level.

[0034] Filter 15 blocks wavelength λ₁from output 16, but passes wavelength λ₂. The optical signal output by filter 15 is the λ₂signal inversely modulated by M. Thus output 20 is optical signal M′λ₂.

[0035] FIG. 3B highlights that SOA 23 and filter 15 of FIG. 3A act as optical inverter 10. The modulated signal M on input 19 is inverted to produce output 20, which carrier signal M′λ₂. However, wavelength conversion also occurs, since the modulation M is transferred from wavelength λ₁to wavelength λ₂.

[0036] FIG. 4 shows waveforms of a semiconductor optical amplifier (SOA) operating as an optical inverter. Two optical signals are input to the active region of the SOA: a modulated signal Mλ₁and a continuous-wave (CW) signal 1λ₂. The first waveform, Mλ₁, is high when M is one, indicating the presence of an optical signal of wavelength λ₁is present, and low when the optical signal of wavelength λ₁is dark (not present). Thus the first waveform indicated the presence or absence of the optical signal at wavelength λ₁. Two pulses of the optical signal at wavelength λ₁occur during the time shown in FIG. 4, one wide and one narrow pulse high.

[0037] The second optical signal, 1λ₂, is always present, since it is continuous. Thus a high waveform is shown for all time periods in FIG. 4 for the second waveform.

[0038] The carrier density becomes depleted when two high optical power inputs are applied to the active layer of the SOA. The power levels of the two signals of wavelengths λ₁and λ₂are selected so that each signal alone is not powerful enough to deplete the carriers in the SOA. However, when both signals are high, carrier depletion occurs.

[0039] The carrier density drops off when both inputs are high. When Mλ₁is high (the first optical signal is present) the carriers become depleted and the carrier density falls. Since 1λ₂is a continuous wave (always present), carrier density is a function of the first input Mλ₁.

[0040] When the carrier density falls, the number of photons emitted is reduced. The amplification of the optical signals depends on photon emission in the SOA, so the amplification or optical gain falls when the carrier density falls.

[0041] The filtered optical signal has a low power when carrier depletion and gain reduction occur, but a higher power when carrier depletion does not occur. Thus the output signal M′λ₂pulses low when modulated input Mλ₁pulses high. The modulated input signal Mλ₁is inverted in logical state and converted from wavelength λ₁to wavelength λ₂by cross-gain modulation.

[0042] FIG. 5 is an optical OR gate. Modulated optical signals X and Y applied to inputs 19, 14 respectively. MMI device 22 combines these inputs to drive SOA 23. SOA 23 operates to restore the power level to a target high level when either or both of inputs 19, 14 are high. The final output 20 is the OR function, (X+Y)λ₂.

[0043] FIG. 6 is an optical NOR gate constructed from an optical MMI device and optical inverter. Two modulated inputs both at wavelength λ₁are input on optical lines X, Y. The two optical signals Xλ₁and Yλ₁are combined by MMI device 22 and restored to a target power level by SOA 23 to produce first input 19 to optical inverter 10. MMI device 22 and SOA 23 act as an optical equivalent to a wired-OR gate, outputting signal (X+Y)λ₁onto first input 19.

[0044] SOA 23 is also operated to restore the power levels to a target high power level P_H. When both inputs X and Y are at the high power level P_H, the combined signal from MMI device 22 can have a double-power level 2*P_H. SOA 23 restores the optical signal to the 1*P_Hlevel. SOA 23 thus acts to restore the power levels to the target level, allowing for better logic integration. SOA 23 can be designed and operated to operate with the desired gain for power levels 1*P_Hand 2*P_Hto output 1*P_H. The injected bias current to SOA 23 can be adjusted to operate SOA 23 in the desired region of the non-linear gain curve. This maintains the desired extinction ratio.

[0045] Optical inverter 10 is another SOA that uses cross-gain modulation to logically invert the modulated input signal on first input 19 and convert it to wavelength λ₂. A continuous-wave signal at wavelength λ₂is input to optical inverter 10 as the second input.

[0046] Optical inverter 10 uses cross-gain modulation to invert the modulation of first input 19, and apply the inverted modulation to output 20. Thus optical inverter 10 outputs signal (X+Y)′λ₂. The modulated signals X and Y input to MMI device 22 each independently pulse high and low over time to carry data. Output 20 likewise pulses low and high as a function of X and Y. The function is the logical NOR of X, Y, (X+Y)′. Thus a logical NOR gate is constructed from MMI Adevice 22, SOA 23, and optical inverter 10. SOA 23 operates to restore to a target level. SOA 10 operates to restore to the target power level as well as acting as an inverter.

[0047] FIG. 7 shows an optical NAND gate using optical inverters and an MMI device. Two optical inverters 10, 11 have their inputs coupled to inputs X and Y, respectively, and their outputs 24, 26 connected to MMI device 22 which drives SOA 23. SOA 23 thus operates to restore to the target level.

[0048] Optical inverter 10 receives modulated optical signal Xλ₁and continuous signal λ₂and generates X′λ₂on intermediate output 24. Likewise, optical inverter 11 receives modulated optical signal Yλ₁and continuous signal λ₂and generates Y′λ₂on intermediate output 26. X and Y are modulation coefficients that vary with time and are normally independent of each other.

[0049] Optical MMI device 22 combines the optical signals from intermediate outputs 24, 26. SOA 23 restores the power level to generate the final output 28. MMI device 22 performs an OR function, combining signals on outputs 24, 26. When one or both of the signals on intermediate outputs 24, 26 are high (optical wave present or above a threshold power), then output 28 is also high. When both intermediate outputs 24, 26 are low (dark or below a threshold power), then their combination, output 28, is also low (dark).

[0050] Since the signal on output 28 is (X′+Y′)λ₂, a logical NAND function is performed. DeMorgan's theorem shows than (X′+Y′) is equivalent to (A AND B)′.

[0051] FIG. 8 is an optical AND gate. First modulated signal Xλ₂is inverted by optical inverter 10 to generate signal X′λ₁on intermediate signal 24. Second modulated signal Yλ₂is input on intermediate signal 26. Optical MMI device 22 combines intermediate outputs 24, 26, and the combined output is inverted to X′*λ₁and also restored to the target power level by SOA 23. Filter 15 blocks wavelength λ₁to generate (X*Y)λ₁on final output 34. A logical AND function is performed.

[0052] From these basic optical gates, OR, NOR, NAND, AND and inverter, arbitrary combinatorial logic functions can be constructed. Thus relatively complex logic can be implemented in the optical domain, with requiring conversion to electrical signals.

[0053] Several other embodiments are contemplated by the inventors. Rather than use a MMI device with 2 inputs and 1 output, larger MMI devices could be used, such as one that has two, three or more inputs. This allows for 3-input logic gates, or even larger logic functions with multiple inputs. The output can be split into 2 or more outputs. Semiconductor optical amplifiers (SOA's) can be constructed from a variety of materials and are not restricted to the semiconductor materials described in the example.

[0054] One benefit of using semiconductor optical amplifiers (SOAs) for optical inverters is that SOA's are inherently amplifying devices. Power levels are restored or boosted by the optical inverters, compensating for any losses. Thus the optical logic using optical inverters is self-restoring. The power levels of the inputs to the optical inverters can be adjusted to vary the amplification, and the SOA's themselves may be modified or optimized in a variety of ways to adjust amplification.

[0055] Rather than be completely dark when the modulated input or output is low, the optical power may be low but still somewhat present. The low optical power can be detected as a low if the optical power is below a threshold of a sensor. The optical inverter can still be useful even when the output is still somewhat on for logic 0.

[0056] Instead of using two complementary types of inverters, three or more types of inverters could be used with three or more wavelengths. For example, a first inverter type could convert from a first to a second wavelength, while a second inverter type could convert from the second to a third wavelengths. A third inverter type could convert from the third wavelength back to the first wavelength. Additional inverter types can be added as additional wavelengths are used.

[0057] The abstract of the disclosure is provided to comply with the rules requiring an abstract, which will allow a searcher to quickly ascertain the subject matter of the technical disclosure of any patent issued from this disclosure. It is submitted with the understanding that it will not be used to interpret or limit the scope or meaning of the claims. 37 C.F.R. §1.72(b). Any advantages and benefits described may not apply to all embodiments of the invention. When the word ‘means’ is recited in a claim element, Applicant intends for the claim element to fall under 35 USC §112, paragraph 6. Often a label of one or more words precedes the word ‘means’. The word or words preceding the word ‘means’ is a label intended to ease referencing of claims elements and is not intended to convey a structural limitation. Such means-plus-function claims are intended to cover not only the structures described herein for performing the function and their structural equivalents, but also equivalent structures. For example, although a nail and a screw have different structures, they are equivalent structures since they both perform the function of fastening. Claims that do not use the word means are not intended to fall under 35 USC §112, paragraph 6. Signals are typically electronic signals, but may be optical signals such as can be carried over a fiber optic line.

[0058] The foregoing description of the embodiments of the invention has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. It is intended that the scope of the invention be limited not by this detailed description, but rather by the claims appended hereto.