Research on optical logic gates based on nonlinear effects

All-optical logic gates are the core components for all-optical signal processing. They can realize all-optical signal extraction, all-optical address recognition, all-optical multiplexing/demultiplexing, and all-optical switching. Therefore, they will have great application potential in future all-optical high-speed communication networks and new generation optical computers. At present, extensive and in-depth research has been carried out both at home and abroad. Semiconductor optical amplifiers have become the main functional devices in various all-optical logic gates due to their small size, good spectral performance, wide operating wavelength range, short response time, and good nonlinear characteristics. This paper introduces several all-optical logic gates based on the nonlinear optical effect in semiconductor optical amplifiers and compares their respective characteristics.

1. The nonlinear optical principle for realizing optical logic gates

The realization of all-optical logic gates is mainly based on the cross-gain modulation wavelength conversion principle in semiconductor optical amplifiers. Under the condition of ignoring the carrier consumption caused by amplified spontaneous radiation, the wavelength conversion process can be described by the following two equations:


Where N is the carrier concentration in the active region, I is the injection current, e is the electron charge, V is the volume of the active region, Γ is the mode field confinement factor, A is the cross-sectional area of ​​the active region, h is Planck's constant, c is the speed of light in vacuum, subscripts S and c correspond to signal light and detection light respectively, gi(N, vi) is the gain coefficient of the corresponding light wave, v is the frequency of the light wave, P+i and Pi correspond to the optical power of forward and reverse propagation respectively, αint is the loss coefficient inside the active region, and R(N) is the carrier consumption caused by non-radiative recombination and spontaneous radiation recombination. In order to accurately simulate the distribution of carriers along the length direction of the active region of the semiconductor optical amplifier, a segmented model can be used for numerical simulation. The active area is divided into M segments, and the carrier concentration in each segment is uniform. Given the incident optical power, the carrier concentration N1 of the first segment can be solved according to formula (1), and then the optical power P2 of the first segment can be calculated according to formula (2). Substituting it into formula (1) can be used to obtain N2. By analogy, the static distribution of the carrier concentration N and optical power P in the entire active area can be obtained in space. Finally, the output optical power that changes with time is calculated using the Runge-Kutta method.

2 Working principle of optical logic gate

2.1 Using semiconductor optical amplifier to realize optical logic AND gate

Using semiconductor optical amplifier (Semiconductor Optical Amplifier, SOA) to realize optical logic AND gate is to use cascaded cross-gain modulation wavelength conversion to realize all-optical logic AND gate. The working principle is: the signal light of a specific rate is amplified by the erbium-doped fiber amplifier and then divided into two paths by coupler 1. One of the signal light A and the continuous light (detection light) provided by the tunable laser are combined by coupler 2 and then sent to SOA1 through a circulator. The two beams of light can produce a wavelength conversion phenomenon based on the cross-gain modulation effect in SOA1, and the information carried by the signal light is converted to the detection light, but in reverse phase with the original information. The signal light output by the first-stage SOA1 is output through the circulator, and then amplified by EDFA2, and then the wavelength-converted signal is filtered out by the bandpass filter 1; the other optical signal is delayed by the tunable delay line, and then coupled into SOA2 through the coupler and circulator together with the signal output by the bandpass filter 1. The power of the first-stage conversion output is properly controlled to be much greater than the signal light power after delay. Therefore, when the bit output by the first-stage conversion is "1", the gain of SOA2 is suppressed, and the output is "0" regardless of whether the signal light is "1" or "0"; on the contrary, when the bit output by the first-stage conversion is "0", the signal light is "1" when it is "1", and "0" when it is "0". Therefore, the signal filtered out by the bandpass filter 2 (aligned with the signal light wavelength) is the logical AND operation result of the signal light A and the delayed signal light B.

2.2 Realization of all-optical logic gate using terahertz optical asymmetric demultiplexer

The principle of realizing all-optical logic gate using terahertz optical asymmetric demultiplexer (TOAD) is shown in Figure 2. Coupler 1 connects a section of optical fiber end to end, and SOA, as a nonlinear element, is asymmetrically placed in the optical fiber line. Its optical path away from the center of the loop is T/2. The control pulse is introduced into the loop from port A through coupler 2, and the detection pulse is injected from port C. The control signal light is strong enough to cause nonlinear effects in SOA, while the detection light is very weak and does not cause nonlinear optical effects in SOA. The working process of this logic gate is as follows: the detection light is input from port C and is divided into two parts with equal amplitudes by coupler 1, which are transmitted in clockwise (CW) and counterclockwise (CCW) directions respectively. In the absence of control light, both CW and CCW light can obtain the small signal gain of SOA, and the phase shifts obtained when they return to coupler 1 again are also equal. Therefore, the two beams of light are coherently destructed at port D, and all the light is reflected from port C; on the contrary, when control light is input from port A, the control light is injected into the loop through coupler 2, and the time delay between the detection light and the control light is appropriately adjusted so that the control light reaches SOA after CCW and before CW. In this way, under the action of the control light, CW will obtain additional nonlinear phase shift, and after being coupled again by coupler 1, it is output from port D, which is equivalent to realizing the logical AND operation of the detection light and the control light.

2.3 All-optical logic gate based on Mach-Zehnder interferometer

The principle of the all-optical logic gate based on Mach-Zehnder interferometer (MZI) is shown in Figure 3. SOA1 and SOA2 are symmetrically placed in the two arms of the interferometer. The continuous detection light is decomposed into two beams through a coupler and injected into the two arms of the interferometer. Two intensity-modulated signal lights with a wavelength of λ1 are injected into them respectively. The peak power of the signal light is higher than the maximum linear input power of the SOA. When the input power exceeds the maximum linear input power of the SOA, the carrier density in the active region of the SOA will change, causing the effective refractive index in the active region to change, resulting in changes in the intensity and phase of the detection light passing through the SOA. The detection light will carry the information of the signal light after passing through the SOA. The two phase-modulated detection lights interfere in the coupler, transferring the phase into amplitude modulation, and completing the XOR operation of the two signals.

2.4 All-optical logic gate based on ultrafast nonlinear interferometer

The working principle of ultrafast nonlinear interferometer (UNI) is shown in Figure 4. The signal light passes through the polarizer to maintain the polarization state in a certain direction. After passing through the birefringent fiber, it is separated into two pulses with different polarization states that are orthogonal to each other and have a certain delay. One pulse enters the SOA first, and then the control pulse is input to the SOA through the coupler 1, and then the orthogonal latter pulse enters the SOA again. Since the intensity of the previous pulse is small, the SOA will not produce gain nonlinearity, while the latter pulse will encounter the SOA gain nonlinearity caused by the strong control pulse, thereby obtaining an additional phase shift.

Therefore, when the two pulses pass through BRF2, whose fast and slow axes are orthogonal to the birefringent fiber BRF1, they overlap again in time. Since the two pulses have a phase difference, they will interfere when they pass through the 45° analyzer, resulting in output; conversely, if there is no control pulse, the two pulses will encounter the same gain characteristics, have no phase difference, and cannot form interference in the analyzer, so there is no output. When using an ultrafast nonlinear interferometer as a logic gate, the clock signal is input as the signal light into the ultrafast nonlinear interferometer, and then the coupler 2 is used to input two logic control signals A and B to replace the original control signal of the ultrafast nonlinear interferometer, and an OR gate and an XOR gate can be obtained.

3 Conclusion

The working principles of several typical all-optical logic gates have been analyzed above. Among them, in the all-optical logic gates realized by semiconductor optical amplifiers, the input signal power and extinction ratio of the erbium-doped fiber amplifier in front of the second-stage semiconductor optical amplifier play a decisive role in the output performance of the logic and operation. The all-optical logic gate scheme realized by using terahertz optical asymmetric demultiplexer has the advantages of simple structure and strong operability. It realizes wavelength conversion while realizing logical operation. Finally, the detection light is used as a carrier to carry the logic result output. At the same time, this scheme is also scalable, that is, it can realize the operation of multiple data streams with different wavelengths. If polarization-independent semiconductor optical amplifiers are used instead, polarization-independent logic gates can be realized. The all-optical logic gate based on ultrafast nonlinear interferometer uses reverse control light to realize all-optical OR gates and XOR gates. At the same time, the signal light and control light can realize single wavelength operation. In addition, the use of semiconductor optical amplifiers makes the structure compact. Together with the all-optical logic gate based on Mach-Zehnder interferometer, it has the advantage of easy integration and is promising for future all-optical signal processing.