Experimental demonstration and devices optimization of NRZ-DPSK amplitude regeneration scheme based on SOAs

Tong Cao; Liao Chen; Yu Yu; Xinliang Zhang

doi:10.1364/OE.22.032138

1. Introduction

The differential-phase-shift keying (DPSK) modulation format has been comprehensively used in optical fiber communication due to its lower optical signal-to-noise ratio requirement and robustness towards nonlinear impairments [1]. Balanced detection of DPSK signals offers a 3-dB improvement in receiver sensitivity in contrast to direct detection of on-off keying (OOK) signals. Therefore DPSK signals are particularly suitable for long-haul transmission. However, when a DPSK signal is transmitted, switched, the signal usually becomes degraded because of many possible effects, such as chromatic dispersion, nonlinear effects in optical fibers and amplified spontaneous emission from optical amplifiers. Thus an all-optical regeneration of DPSK signals is required. Unlike the on-off keying (OOK) format, the DPSK format regeneration can be realized by phase regeneration or amplitude regeneration. Phase regeneration has been realized by phase-sensitive amplification [2–4] or phase-to-amplitude conversion followed by amplitude regeneration and subsequent encoding into phase-modulated signal [5–9]. However, phase-sensitive amplification typically operates at high input power levels because of using four-wave mixing effect and requires complex phase-locking of the pump signal with the data signal. Schemes of phase-to-amplitude conversion usually alter the original differential coding when the signal is back-encoded into the phase domain [5] and generally work in an interferometric structure, which is not intrinsically stable. On the other hand, amplitude regeneration can be realized by fiber-based [10–12] or semiconductor optical amplifier (SOA)-based [13–18] all-optical regenerators. Among the above schemes, those based on phase-preserving amplitude regeneration of SOAs receive many interests and seem to be promising due to the benefits of low power operating power levels, compactness and cost-effectiveness.

The cross-gain compression (XGC) in SOAs has been recently investigated in all-optical signal processing [19–23]. Generally, the schemes based on the XGC effect require two SOAs. The first SOA (SOA1) is used to generate an intensity inverted and wavelength converted copy of input signal. The second SOA (SOA2) is used to experience the XGC effect. As the two SOAs are cascaded, not used in an interferometric structure, the schemes based on XGC effect offer intrinsic stability. The regeneration based on XGC effect was used by G. Contestabile to reamplify and reshape the distorted non-return-to-zero (NRZ)-OOK signal [19]. Later, an InP photonic integrated circuit (PIC) has been designed and realized to demonstrate effective all-optical regeneration for NRZ and return-to-zero (RZ) signals [22, 23]. To the best of our knowledge, there is no report on DPSK regeneration based on the XGC effect. In this paper, we demonstrate NRZ-DPSK amplitude regeneration with XGC effect in SOAs. For the amplitude regeneration, there exists extra phase noise induced by amplitude-to-phase conversion, which is similar to the Gordon-Mollenauer effect in fiber [24]. However, SOA’s amplitude-to-phase conversion occurs owing to the refractive index-change determined by the carrier density and temperature modulation from the amplitude noise. Due to the almost constant total power into the SOA2, the carrier density and temperature modulation is largely compressed. Thus the extra phase noise is significantly suppressed. Therefore the scheme based on the XGC effect is effective for NRZ-DPSK phase-preserving amplitude regeneration. On the other hand, two SOAs are exploited in this scheme, but the requirements for their characteristics are different. In order to further improve the regenerative performance, the two SOAs’ quantum-well (QW) structures are separately optimized, by combining QW band structure calculation with SOAs’ dynamic model. Regarding SOA1, the most important factor is to get an inverted signal without too much distortion at high bit rate. The maximal operating bit rate of the schemes based on XGC effect critically depends on the availability of the intensity inverted signal [19]. Thus in order to obtain higher operating bit rate, we use the transient cross-phase modulation (T-XPM) effect to eliminate the slow recovery process determined by the electron-hole recombination time (typically several tens to hundreds of picoseconds) [25] and optimize the QW structure of SOA1 to enhance the T-XPM effect. Regarding SOA2, the key factor is to preserve the phase information while realizing amplitude regeneration. The extra phase noise due to amplitude-to-phase conversion is mainly determined by the refractive index-change. We use the optimized rules [13] to design the QW structure to decrease the refractive index-change. In a word, we propose a novel scheme which can simultaneously realize wavelength-preserving and phase-preserving amplitude noise compression of a NRZ-DPSK signal and for each SOA used in the scheme, the QW structure is separately optimized to get its best performance.

The rest of this paper is organized as follows. In section 2, the principle of the NRZ-DPSK amplitude regeneration based on the XGC effect is presented and the SOA model is shown. In section 3, experimental demonstration of amplitude noise compression for a distorted 40Gb/s NRZ-DPSK signal is put forward. In section 4, we separately optimize QW structures of the two SOAs used in our scheme. Finally, conclusions are given in section 5.

2. Operating principle and QW SOA model

2.1 Operating principle

The NRZ-DPSK signal is generated by a push-pull Mach-Zehnder modulator (MZM). Due to the imperfect modulation of MZM, there are amplitude dips in the output waveform where phase changes between 0 and $π$ . The illustration of generating the intensity inverted and wavelength converted copy of input signal in SOA1 is described in Fig. 1(a). The pump signal, along with a continuous wave (CW) probe beam, is fed into SOA1. The carrier density and carrier temperature are modulated by the pump signal and detected by the probe beam. The purpose of the subsequent bandpass filter (BPF) is to take the desired blue chirp component to eliminate the slow recovery process determined by the electron-hole recombination time. The operating principle is similar to the non-inverted wavelength conversion (WC) [25], which is so-called T-XPM effect. Figure 1(b) shows the operating principle of the wavelength-preserving amplitude regeneration based on the XGC effect. The initial NRZ-DPSK signal and the converted signal obtained by SOA1 are coupled, synchronized, and injected into the SOA2. Controlling the relative power of the two signals, the total power into SOA2 can be set constant. The green line in Fig. 1(b) shows the total amplitude fluctuations are only induced by the input NRZ-DPSK amplitude fluctuations. During the propagation of the two intensity-inverted signals, the saturated gain of SOA2 works as a power equalizer. In other words, the input NRZ-DPSK amplitude can be regenerated without obvious carrier density and carrier temperature change. Thus the extra phase noise induced by amplitude-to-phase conversion effect can be significantly suppressed.

Fig. 1 (a) Schematic illustration of T-XPM effect in SOA1 to obtain the intensity inverted signal. (b) Schematic illustration of XGC effect in SOA2 between two signals having opposite intensity and different wavelength.

Download Full Size | PDF

2.2 QW SOA model

If we want to optimize the QW structure, a comprehensive SOA model should be developed. Here we adopt Huang’s model [13], which contains band structure calculation and SOA’s dynamic model. The intraband effects, such as spectrum hole burning (SHB) and carrier heating (CH) are considered in our simulated model. The T-XPM effect of SOA1 is mainly determined by the intraband effects.

The band structure is obtained by the $k \cdot p$ method. The conduction bands can be characterized by a parabolic-band model

- \frac{ℏ^{2}}{2} \frac{\partial}{\partial z} \frac{1}{m (z)} \frac{\partial}{\partial z} ϕ (z) + V (z) ϕ (z) = E_{c n} ϕ (z)

where

m (z)

are the electron effective masses perpendicular (t) and parallel (z) to the growth direction.

V (z)

is the potential energy of the strained conduction band edge.

E_{c n}

and

ϕ (z)

are the energy and envelope function of the nth conduction subband. The valence band structure is calculated using the

6 \times 6

Luttinger-Kohn Hamiltonian

H_{3 \times 3}^{σ} (k_{z} = - i \frac{\partial}{\partial z}) [\begin{matrix} g_{m, h h}^{σ} (z; k_{t}) \\ g_{m, l h}^{σ} (z; k_{t}) \\ g_{m, s o}^{σ} (z; k_{t}) \end{matrix}] = E_{σ m}^{v} [\begin{matrix} g_{m, h h}^{σ} (z; k_{t}) \\ g_{m, l h}^{σ} (z; k_{t}) \\ g_{m, s o}^{σ} (z; k_{t}) \end{matrix}]

where

H_{3 \times 3}^{σ}

and

E_{σ m}^{v}

are Luttinger-Kohn Hamiltonian and the energy of the mth valence band.

g_{m, h h}^{σ} (z; k_{t})

,

g_{m, l h}^{σ} (z; k_{t})

,

g_{m, s o}^{σ} (z; k_{t})

are the hole envelope functions. In [26], Chang et al. gives the details of the band structure calculation and material parameters. In the active region of QW SOAs, the optical material gain coefficient can be written as

\begin{array}{l} g_{i} (ℏ ω) = \frac{q^{2} π}{n_{r} c ε_{0} m_{0}^{2} ω L_{w}} \sum_{η = ↑, ↓} \sum_{σ = U, L} \sum_{n, m} \int {| \overset{\land}{e} \cdot M_{n m}^{η σ} (k_{t}) |}^{2} \times \frac{(γ / π)}{{(E_{σ, n m}^{c v} (k_{t}) - ℏ ω)}^{2} + γ^{2}} \\ \begin{matrix}  \end{matrix} \times (f_{n}^{c} (k_{t}) + f_{σ m}^{v} (k_{t}) - 1) \times \frac{k_{t} d k_{t}}{2 π} \end{array}

where

f_{n}^{c} (k_{t})

and

f_{σ m}^{v} (k_{t})

are the Fermi functions of electrons and holes. q is the magnitude of the electron charge,

m_{0}

is the electron rest mass in free space, c and

ε_{0}

are the velocity of light and permittivity in free space,

\overset{\land}{e}

is the polarization vector of the optical electric field,

n_{r}

and

L_{w}

are the refractive index and well width of the quantum well,

γ

is the half linewidth of the Lorentzian function,

k_{t}

is the real transverse wave number and

M_{n m}^{η σ} (k_{t})

is the momentum matrix element.

E_{σ, n m}^{c v} (k_{t}) {=E}_{n}^{c} (k_{t}) - E_{σ, m}^{v} (k_{t})

, where

E_{n}^{c} (k_{t})

and

E_{σ, m}^{v} (k_{t})

are eigenenergies in the conduction and the valence band, which are calculated by the above band structure calculation.

The refractive index-change can be described by

\begin{array}{l} Δ n_{i} = \frac{q^{2} ℏ^{2}}{ε_{0} m_{0}^{2} n_{r} L_{w}} \sum_{η = ↑, ↓} \sum_{σ = U, L} \sum_{n, m} \int {| \overset{\land}{e} \cdot M_{n m}^{η σ} (k_{t}) |}^{2} \times \frac{(f_{n}^{c} (k_{t}) + f_{σ m}^{v} (k_{t}) - 1)}{{(E_{σ, n m}^{c v} (k_{t}) - ℏ ω)}^{2} + γ^{2}} \\ \begin{matrix}  \end{matrix} \times \frac{E_{σ, n m}^{c v} (k_{t}) - ℏ ω}{E_{σ, n m}^{c v} (k_{t}) (E_{σ, n m}^{c v} (k_{t}) + ℏ ω)} \times \frac{k_{t} d k_{t}}{2 π} \end{array}

The traveling wave rate equations are modeled as

\frac{d S_{i}}{d z} = (Γ g_{i} - α_{F c} N - α_{int}) S_{i}

\frac{d Φ_{i}}{d z} = - Γ \frac{2 π}{λ} (Δ n_{i} + Δ n_{F c})

where

Γ

,

α_{F c}

,

N

,

α_{int}

,

Δ n_{F c}

are the confinement factor, absorption coefficient due to free carrier absorption, carrier density, internal loss, and refractive index-change due to plasma effect.

The time evolutions of the carrier density and carrier temperature are described by

\frac{d N}{d t} = \frac{I}{q V} - \sum_{i} v_{g} g_{i} S_{i} - A N - B N^{2} - C N^{3}

\frac{d T}{d t} = \frac{1}{\partial U / \partial T} [- \sum_{i} (ℏ ω_{i} - E_{g}) v_{g} g_{i} S_{i} + \sum_{i} ℏ ω_{i} v_{g} α_{F c} N S_{i} - \frac{\partial U}{\partial N} \frac{d N}{d t}] - \frac{T - T_{0}}{τ_{T}}

where I, V, v_g, A, B, C are the injection current, active waveguide volume, group velocity, the recombination constants caused by trapping sites, the bimolecular spontaneous radiative recombination coefficient and the Auger recombination coefficient, respectively. The details of Eq. (8) can be found in [27], while the modeling parameters can refer to [13].

3. Experimental demonstration of NRZ-DPSK amplitude regeneration

The experimental setup for validating the amplitude regeneration based on XGC effect is reported in Fig. 2. The NRZ-DPSK data signal was generated by modulating a CW light at 1560 nm with a 40 Gb/s pseudo random bit sequence (PRBS) of length 2⁷-1, whose eye diagram and measured optical spectra are shown in Fig. 2(a1) and 2(b1). The amplitude distortion of the signal was emulated by slightly detuning the driving voltage of the MZM from the optimum value. Obviously, the phase information remains unchanged while the amplitude is different for signals “0” and “1”. Then the NRZ-DPSK signal was split in two parts. One part with the probe signal at 1558 nm was launched into SOA1. The powers of NRZ-DPSK signal and the CW signal were 12.27 and 5.00 dBm, respectively. The measured optical spectra after SOA1 is shown in Fig. 2(b2), where the spectra of probe signal is broadened due to cross phase modulation in SOA1. After the blue detuning BPF, the converted signal could be obtained, as Fig. 2(a2) and Fig. 2(b3) show. The blue detuning and 3-dB bandwidth were 0.32 and 0.26 nm. The other part passed an optical delay line (ODL) and optical attenuator (ATT) and was coupled together with the converted signal into SOA2. The powers of the second part NRZ-DPSK signal and the converted signal were 7.03 and 4.68 dBm. The total input and output eye diagrams into SOA2 are shown in Fig. 2 (a3) and 2(a4). In order to obtain better regenerative performance, fine enough time-synchronization and almost constant total input power should be carefully adjusted. Generally, longer length of PRBS means much more difficult for time-synchronization. If the PRBS length of 2⁷-1 is long enough to evaluate the performance, there is no necessary to choose longer PRBS. In [28], Xu et al. has analyzed the minimum bit pattern length of the PRBS which can characterize pattern effect versus the carrier recovery time. Xu’s results show the PRBS length of 2⁷-1 is long enough to characterize the pattern effect as long as the carrier recovery time is shorter than 141 ps. In our scheme, for SOA1, we use the T-XPM effect to get the converted signal. In [25], Liu et al. has demonstrated based on T-XPM effect the carrier recovery time can reduce to about 3 ps, which is much shorter than 141 ps. For SOA2, due to using XGC effect, there is no pattern effect. That is to say, in our scheme using PRBS length of 2⁷-1 can effectively characterize the performance of the NRZ-DPSK amplitude regeneration. The optical spectra of the amplitude regenerative NRZ-DPSK is shown in Fig. 2(b4), which is almost the same as the input spectra in Fig. 2(b1). This partly verifies that our scheme can regenerate the signal while the phase information is preserved. For the spectra measurement, an ATT is used before the optical spectrum analyzer (OSA) to control the input power, and by adjusting the ATT we set the maximum peak of the measured optical spectra equal 0 dBm. It is shown that the maximum power of the spectra in Fig. 2(b1), 2(b2), 2(b3) and 2(b4) all equal 0 dBm. The two SOAs are CIP nonlinear devices (CIP SOA-NL-1550) which operated at bias current of 210 and 240 mA, respectively.

Fig. 2 Experimental setup.

Download Full Size | PDF

The first validation of our scheme can be seen in Fig. 3. The eye diagrams of the original distorted NRZ-DPSK signal, back to back (B2B) constructive demodulation NRZ-OOK signal, and B2B destructive demodulation RZ-OOK signal are shown in Fig. 3(a1), 3(a2), and 3(a3). The eye diagrams shown in Fig. 3(b1), 3(b2) and 3(b3) are obtained from the regenerative NRZ-DPSK signal and its demodulation signals. For the original signals, the strongly degraded signals with large amplitude fluctuations can be seen in Fig. 3(a1) and 3(a2). Based on our scheme, clear eye opening, reduction of amplitude fluctuations have been obtained in Fig. 3(b1) and 3(b2).

Fig. 3 Eye diagrams of the regenerative process.

Download Full Size | PDF

We further validate the regenerative properties by the bit-error rate (BER) measurement as shown in Fig. 4. Power penalties up to −1.75dB for constructive demodulation and −1.01dB for destructive demodulation at BER of 10⁻⁹ are obtained, as Fig. 4(a) and 4(b) show.

Fig. 4 Measured BER curves of constructive demodulation (a) and destructive demodulation (b).

Download Full Size | PDF

4. Optimized QW SOAs for better NRZ-DPSK amplitude regeneration

In above experiments, we used two SOAs. SOA1 was used to generate the inverted signal based on SOA’s T-XPM effect and SOA2 was used to regenerate the distorted amplitude of the NRZ-DPSK signal using SOA’s XGC effect. The maximum operating bit rate of our scheme is only determined by SOA1’s maximum operating bit rate. Using the T-XPM effect can effectively reduce the impact of slow recovery process. Stronger T-XPM effect means better performance at higher bit rate. Enhancing the T-XPM effect in SOA1 can be realized by optimizing the QW structure to get the maximal refractive index-change. However, in SOA2, the large refractive index-change is undesirable. We want to regenerate the degraded amplitude without introducing extra phase noise. That is to say, the requirements of the two SOAs are totally different. The QW structure of each SOA should be separately optimized to get its best performance. The quantum well is In_1-xGa_xAs which is surrounded by In_0.7322Ga_0.2678As_0.5810P_0.4190 barrier. The barrier’s lattice is matched to the InP substrate and bandgap wavelength is 1.3 um. When the gallium mole fraction (x) equals 0.47, the well is lattice-matched to the InP substrate. Larger x than 0.47 means the tensile strain is introduced in QW structure and smaller than 0.47 means compressive strain. As the gain of the device is mainly determined by the gain of transverse electric (TE) mode in our material system, only TE mode gain is taken into account in this paper.

4.1 SOA model fit for experiment

If we want to characterize improved performance of the optimized QW SOAs, we must find one type of QW SOA in our material system to describe the SOAs in our experiments. However, we don’t know the quantum structure of the experimental SOAs. To solve this problem, we fit simulation results with the experimental results. Firstly, we use the experimental spectra of probe signal after SOA1 to define a parameter P_TXPM in Fig. 5(a), which determines the intense of the T-XPM effect. Stronger T-XPM effect means lower parameter P_TXPM. Experimental P_TXPM equals 17.82 dB. Setting other operating parameters same as the experimental parameters, we theoretically calculate the P_TXPM as a function of quantum well width at x = 0.47, which is shown in Fig. 5(b). The well widths at point A and point B are 7 and 10.2 nm, which both agree well with the experiment. The experimental amplitude regeneration based on XGC effect is shown in Fig. 6(a). In order to further find the right point, we simulate the amplitude regeneration based on XGC effect at both points as Fig. 6(b) and 6(c) show. In contrast to the amplitude regeneration at point B, the amplitude regeneration at point A is more similar to the experimental amplitude regeneration. Thus the type of SOAs in our experiment can be modeled using the unstrained (x = 0.47) QW SOA with 7 nm well width. There is another point to say, the mole fraction x is not necessary to choose 0.47. If choose other value, the optimized well will not be 7 nm any more. The work in this section is just used to find the reference SOA in contrast to the optimized QW SOAs. In the following section, we call the unstrained (x = 0.47) QW SOA with 7 nm well width as the initial QW SOA.

Fig. 5 (a) Experimental spectrum of probe signal after SOA1 and definition of parameter P_TXPM (b) calculated P_TXPM as a function of quantum well width at x = 0.47.

Download Full Size | PDF

Fig. 6 Constructive demodulation eye diagrams of experiment (a), at point A (b), and at point B (c). (a1), (b1) and (c1) are the distorted signal demodulation results, while (a2), (b2), and (c2) are the regenerative demodulation results.

Download Full Size | PDF

4.2 Optimized QW structure for SOA1

To improve the performance of SOA1 based on T-XPM effect, the quantum structure and the optical confinement factor are carefully designed. In Fig. 7, we analyze the strain and well width effects on the maximal probe blue chirp component and the output power after the BPF, which are determined by the T-XPM effect. The BPF’s blue detuning and 3-dB bandwidth are fixed to 0.32 and 0.26 nm, which are the same as the parameters used in our experiment. For a fixed well width (> 8 nm), the tensile strain means larger maximal blue chirp component and output power. For a fixe mole fraction x, there is an optimal well width to get the maximum of the blue chirp component and output power. And the optimal well width of compressive strain is narrower than tensile strain. However, the fabricating process of narrow well width is much more difficult. Thus we choose the tensile strained QW SOA at the optimized well width of 9 nm to enhance the T-XPM effect.

Fig. 7 The maximal blue chirp component (a) and average power of converted signal (b) as functions of quantum well width and strain. The optical confinement factor is fixed to 0.30.

Download Full Size | PDF

Figure 8 shows the optical confinement factor impact on the maximal blue chirp component and the output power of the tensile strained QW SOA. It is clear to choose large optical confinement factor to enhance the T-XPM effect. The large optical confinement factor can be realized by increasing the number of the quantum well.

Fig. 8 The maximal blue chirp component (a) and average power of converted signal (b) as functions of quantum well width and optical confinement factor.

Download Full Size | PDF

In a word, we have designed optimized QW SOA to enhance the T-XPM effect using 9 nm tensile strained quantum well and large optical confinement factor. The optimized result is shown in Fig. 9. The output average power of the converted signal can be improved up to 2.46 dB. Our optimized QW SOA for SOA1 is not only suitable in our scheme but also suitable for all applications based on T-XPM effect.

Fig. 9 Converted eye diagrams based on the initial QW SOA (a) and the optimized QW SOA for SOA1 (b).

Download Full Size | PDF

4.3 Optimized QW SOA for SOA2

In this section, we investigate the well structure effect on the performance of the amplitude regeneration. Figure 10 shows the results for SOA2. The initial NRZ-DPSK amplitude fluctuations, phase fluctuations, constructive demodulation Q factor, and destructive demodulation Q factor are 40.72%, 0.56%, 6.86, and 11.14. The amplitude fluctuations can be largely suppressed by using larger well width and introducing compressive strain as Fig. 10(a) shows. At the same time, the amplitude-to-phase conversion effect which distorts the phase can also be impressively compressed in Fig. 10(b). Thus the demodulation Q factor can be significantly improved in Fig. 10(c) and 10(d). We choose the compressively strained QW SOA at the optimized well width of 11 nm to regenerate the distorted signal. Our optimized QW SOA for SOA2 is not only suitable in our scheme but also suitable for all DPSK amplitude regeneration based on gain saturation in SOA, such as Porzi’s scheme [18].

Fig. 10 Regenerative NRZ-DPSK amplitude fluctuations(a), phase fluctuations(b), constructive demodulation Q factor (c) and destructive demodulation Q factor (d) as functions of quantum well width and strain. The optical confinement factor is fixed to 0.30.

Download Full Size | PDF

We further validate the regenerative performance of our optimized QW SOAs as the Fig. 11 shows. Figure 11(a) gives the input constellation diagram. The constellation diagrams of using initial QW SOAs and optimized QW SOAs are shown in Fig. 11(b) and 11(c). In contrast to the initial output constellation diagram, the optimum output verifies that the amplitude fluctuations can be significantly suppressed without introducing much extra phase noise. Better wavelength-preserving and phase-preserving amplitude regeneration of NRZ-DPSK signal can be obtained using the optimized QW SOAs.

Fig. 11 Constellation diagrams of input NRZ-DPSK signal (a), initial QW SOAs output (b) and optimized QW SOAs output (c).

Download Full Size | PDF

5. Conclusion

We have proposed and experimentally demonstrated a novel scheme for wavelength-preserving and phase-preserving amplitude regeneration of a 40 Gb/s distorted NRZ-DPSK signal based on SOAs. Power penalties up to −1.75dB for constructive demodulation and −1.01dB for destructive demodulation at BER of 10⁻⁹ have been obtained. We have designed different optimized QW SOAs for NRZ-DPSK amplitude regeneration. Regarding SOA1, using the optimized tensile strained QW and large optical confinement factor SOA, the output average power can be improved 2.46 dB. Regarding SOA2, using compressively strained QW SOA, the amplitude fluctuations can be largely suppressed. In a word, better wavelength-preserving and phase-preserving amplitude regeneration of NRZ-DPSK signal can be obtained using the optimized QW SOAs. The scheme is intrinsically stable comparing with the interferometer structure and can be integrated on a chip, making it a practical candidate for all-optical amplitude regeneration of high-speed NRZ-DPSK signal.

Acknowledgment

This work was supported by the National Basic Research Program of China (Grant No. 2011CB301704), the National Natural Science Found for Distinguished Yong Scholars (61125501), NSFC Major International Joint Research Project (61320106016) and Scientific and Technological Innovation Cross Team of Chinese Academy of Sciences.

References and links

1. A. H. Gnauck, G. Raybon, S. Chandrasekhar, J. Leuthold, C. Doerr, L. Stulz, A. Agarwal, S. Banerjee, D. Grosz, S. Hunsche, A. Kung, A. Marhelyuk, D. Maywar, M. Movassaghi, X. Liu, C. Xu, X. Wei, and D. M. Gill, “2.5 Tb/s (64x42.7 Gb/s) transmission over 40x100 km NZDSF using RZ-DPSK format and all-Raman-amplified spans,” in Optical Fiber Communications Conference, Vol. 70 of 2002 OSA Trends in Optics and Photonics Series (Optical Society of America, 2002), paper FC2. [CrossRef]

2. K. Croussore, C. Kim, and G. Li, “All-optical regeneration of differential phase-shift keying signals based on phase-sensitive amplification,” Opt. Lett. 29(20), 2357–2359 (2004). [CrossRef] [PubMed]

3. K. Croussore, I. Kim, C. Kim, Y. Han, and G. Li, “Phase-and-amplitude regeneration of differential phase-shift keyed signals using a phase-sensitive amplifier,” Opt. Express 14(6), 2085–2094 (2006). [CrossRef] [PubMed]

4. Y. Meng, J. Lian, S. Fu, M. Tang, P. Shum, and D. Liu, “All-optical DPSK regenerative one-to-nine wavelength multicasting using dual-pump degenerate phase sensitive amplifier,” J. Lightwave Technol. 32(15), 2605–2612 (2014). [CrossRef]

5. I. Kang, C. Dorrer, L. Zhang, M. Rasras, L. Buhl, A. Bhardwaj, S. Cabot, M. Dinu, X. Liu, M. Cappuzzo, L. Gomez, A. Wong-Foy, Y. F. Chen, S. Patel, D. T. Neilson, J. Jacques, and C. R. Giles, “Regenerative all optical wavelength conversion of 40-Gb/s DPSK signals using a SOA-MZI,” in Proceedings of the European Conference and Exhibition on Optical Communication (ECOC, Glasgow, Scotland, 2005), paper Th.4.3.3.

6. V. S. Grigoryan, S. Myunghun, P. Devgan, J. Lasri, and P. Kumar, “SOA-based regenerative amplification of phase-noise-degraded DPSK signals: dynamic analysis and demonstration,” J. Lightwave Technol. 24(1), 135–142 (2006). [CrossRef]

7. P. Johannisson, G. Adolfsson, and M. Karlsson, “Suppression of phase error in differential phase-shift keying data by amplitude regeneration,” Opt. Lett. 31(10), 1385–1387 (2006). [CrossRef] [PubMed]

8. P. Vorreau, A. Marculescu, J. Wang, G. Bottger, B. Sartorius, C. Bornholdt, J. Slovak, M. Schlak, C. Schmidt, S. Tsadka, W. Freude, and J. Leuthold, “Cascadability and regenerative properties of SOA all-optical DPSK wavelength converters,” IEEE Photon. Technol. Lett. 18(18), 1970–1972 (2006). [CrossRef]

9. C. Kouloumentas, M. Bougioukos, A. Maziotis, and H. Avramopoulos, “DPSK regeneration at 40 Gb/s and beyond using a fiber-Sagnac interferometer,” IEEE Photon. Technol. Lett. 22(16), 1187–1189 (2010). [CrossRef]

10. A. Striegler and B. Schmauss, “All-optical DPSK signal regeneration based on cross-phase modulation,” IEEE Photon. Technol. Lett. 16(4), 1083–1085 (2004). [CrossRef]

11. M. Matsumoto, “Regeneration of RZ-DPSK signals by fiber-based all-optical regenerators,” IEEE Photon. Technol. Lett. 17(5), 1055–1057 (2005). [CrossRef]

12. A. G. Striegler, M. Meissner, K. Cvecek, K. Sponsel, G. Leuchs, and B. Schmauss, “NOLM-based RZ-DPSK signal regeneration,” IEEE Photon. Technol. Lett. 17(3), 639–641 (2005). [CrossRef]

13. X. Huang, Z. Zhang, C. Qin, Y. Yu, and X. Zhang, “Optimized quantum well semiconductor optical amplifier for RZ-DPSK signal regeneration,” IEEE J. Quantum Electron. 47(6), 819–826 (2011). [CrossRef]

14. C. Porzi, A. Bogoni, and G. Contestabile, “Regeneration of DPSK signals in a saturated SOA,” IEEE Photon. Technol. Lett. 24(18), 1597–1599 (2012). [CrossRef]

15. Y. Yu, W. Wu, X. Huang, B. Zou, S. Hu, and X. Zhang, “Multichannel all-optical RZ-PSK amplitude regeneration based on the XPM effect in a single SOA,” J. Lightwave Technol. 30(23), 3633–3639 (2012). [CrossRef]

16. C. Porzi, A. Bogoni, and G. Contestabile, “Regenerative wavelength conversion of DPSK signals through FWM in an SOA,” IEEE Photon. Technol. Lett. 25(2), 175–178 (2013). [CrossRef]

17. B. Zou, Y. Yu, W. Wu, X. Huang, and X. Zhang, “All-optical amplitude regeneration of non-return-to-zero differential-phase-shift-keying signal,” Opt. Commun. 298(1–2), 83–87 (2013). [CrossRef]

18. C. Porzi, G. Serafino, A. Bogoni, and G. Contestabile, “Phase-preserving amplitude noise compression of 40 Gb/s DPSK signals in a single SOA,” J. Lightwave Technol. 32(10), 1966–1972 (2014). [CrossRef]

19. G. Contestabile, R. Proietti, N. Calabretta, and E. Ciaramella, “Reshaping capability of cross-gain compression in semiconductor amplifiers,” IEEE Photon. Technol. Lett. 17(12), 2523–2525 (2005). [CrossRef]

20. G. Contestabile, N. Calabretta, R. Proietti, and E. Ciaramella, “Double-stage cross-gain modulation in SOAs: an effective technique for WDM multicasting,” IEEE Photon. Technol. Lett. 18(1), 181–183 (2006). [CrossRef]

21. G. Contestabile, R. Proietti, N. Calabretta, and E. Ciaramella, “Cross-gain compression in semiconductor optical amplifiers,” J. Lightwave Technol. 25(3), 915–921 (2007). [CrossRef]

22. F. Bontempi, S. Faralli, N. Andriolli, and G. Contestabile, “An InP monolithically integrated unicast and multicast wavelength converter,” IEEE Photon. Technol. Lett. 25(22), 2178–2181 (2013). [CrossRef]

23. N. Andriolli, S. Faralli, F. Bontempi, and G. Contestabile, “A wavelength-preserving photonic integrated regenerator for NRZ and RZ signals,” Opt. Express 21(18), 20649–20655 (2013). [CrossRef] [PubMed]

24. J. P. Gordon and L. F. Mollenauer, “Phase noise in photonic communications systems using linear amplifiers,” Opt. Lett. 15(23), 1351–1353 (1990). [CrossRef] [PubMed]

25. Y. Liu, E. Tangdiongga, Z. Li, S. Zhang, H. Waardt, G. D. Khoe, and H. J. S. Dorren, “Error-free all-optical wavelength conversion at 160 Gb/s using a semiconductor optical amplifier and an optical bandpass filter,” J. Lightwave Technol. 24(1), 230–236 (2006). [CrossRef]

26. C. Chih-Sheng and C. Shun-Lien, “Modeling of strained quantum-well lasers with spin-orbit coupling,” IEEE J. Sel. Top. Quantum Electron. 1(2), 218–229 (1995). [CrossRef]

27. J. M. Dailey and T. L. Koch, “Simple rules for optimizing asymmetries in SOA-based Mach-Zehnder wavelength converters,” J. Lightwave Technol. 27(11), 1480–1488 (2009). [CrossRef]

28. J. Xu, X. L. Zhang, and J. Mork, “Investigation of patterning effects in ultrafast SOA-based optical switches,” IEEE J. Quantum Electron. 46(1), 87–94 (2010). [CrossRef]

Experimental demonstration and devices optimization of NRZ-DPSK amplitude regeneration scheme based on SOAs

Abstract

1. Introduction

2. Operating principle and QW SOA model

2.1 Operating principle

2.2 QW SOA model

3. Experimental demonstration of NRZ-DPSK amplitude regeneration

4. Optimized QW SOAs for better NRZ-DPSK amplitude regeneration

4.1 SOA model fit for experiment

4.2 Optimized QW structure for SOA1

4.3 Optimized QW SOA for SOA2

5. Conclusion

Acknowledgment

References and links

Cited By

Figures (11)

Equations (8)

Optics Express