Open Access
Add to BibSonomy
Abstract
An all-optical channel aggregation from two lower bit rate OOK and MPSK to a higher bit rate 2M-ary PSK signal based on cross phase modulation (XPM) effect and corresponding de-aggregation to recover the two original signals by phase sensitive amplification (PSA) are proposed and demonstrated aimed to improve efficient use of the fiber and transponder resources in elastic optical networking (EON). Moreover, for the PSA-based de-aggregation scheme, a black-box device to extract the phase locked pump is also designed to realize the desired phase-locking relationship for practical application. Both 20-Gbps aggregated QPSK signal (M=2) and 30-Gbps aggregated 8PSK signal (M=4) as examples are studied respectively. The feasibility and tunability of the scheme have been confirmed by the input-output constellations and waveforms. The phase noise (PN) of recovered B/QPSK signal and amplitude standard deviation (ASD) of recovered OOK signal over the varying optical signal-to-noise ratio (OSNR) of the Q/8PSK signal affected by the link noise are studied and analyzed. Based on the same link noise environment, the two system bit-error rate (BER) performance are also investigated and the corresponding OSNRs for error-free signal recovery are given respectively. Finally, some potential application scenarios are discussed based on the same setup.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Elastic optical networking (EON) is a promising network architecture which enables agile spectrum management and intelligent grooming [1]. It has the advantage to allocate variable granularity bandwidth to maximize spectral efficiency by adapting to the actual conditions of the network and data rate for each given traffic demand [2]. However, for each low-speed traffic, the resource utilization would be still low and costly if an end-to-end independent optical path was established for it, since the number of optical channels in the fiber and the number of optical transceivers in the network node are limited.
Channel aggregation and de-aggregation enabling traffic grooming are the key technologies in EON application [3]. At the source network node, multiple low-speed traffic channels can be aggregated into a single high-speed channel for the long haul transmission. At the destination node, inverse conversion will be achieved by channel de-aggregation, so that these low-speed traffic data can be recovered and sent to their own receivers. In this way, aggregation and de-aggregation functions greatly improve fiber and transponder utilization by sharing common optical channels with several low-speed traffic. However, conventional aggregation and de-aggregation are commonly performed in the electrical domain, where the signal demodulation and re-modulation are necessary. Some digital signal processing(DSP) algorithms are also required to mark each low-speed signal to ensure that the original traffic is correctly de-aggregated. However, this electrical method has limitations to handle with the high bit rate signal due to the electronic rate bottleneck.
All-optical approach based on fiber nonlinear effects enabling ultrahigh-speed signal processing has attracted desirable attention in recent years [4]. Some optical aggregations have been demonstrated, such as multiple on-off keying (OOK) signals to a M-ary phase shift keying (MPSK) signal [5] by cross phase modulation (XPM) in highly nonlinear fiber (HNLF), a quadrature phase shift keying (QPSK) and amplitude and phase shift keying (APSK) to an 8-ary quadrature amplitude modulation (8QAM) signal by four-wave mixing (FWM) in HNLF [6], two quadrature phase shift keying (QPSK) signals to a 16-ary quadrature amplitude modulation (16QAM) signal by XPM in HNLF [7], a OOK and an BPSK signal to a QPSK signal by XPM in nonlinear birefringent photonic crystal fiber (PCF) [8], and four OOK signals to a polarization division multiplexing (PDM) QPSK signal by XPM in HNLF [9]. However, all of these schemes are partly focused on optical aggregation, the corresponding de-aggregation function has not been performed.
Similarly, numerous single function optical de-aggregation schemes have also been demonstrated. For example, a QPSK signal can be de-aggregated to two BPSK signals using FWM [10] or phase sensitive amplification (PSA) [11–14]. An 8PSK signal can be de-aggregated into two QPSK signals [15] or three BPSK signals by PSA [16]. Based on the same PSA mechanism, a 16QAM signal can be de-aggregated to two 4-pulse amplitude modulation (PAM) signals [17,18] or two QPSK signals [19]. However, besides the lack of the inverse aggregation function, in these PSA-based de-aggregation schemes, the phase locking relationship is also neglected even though it is critical for practical applications.
To overcome the issue, few simultaneous all-optical channel aggregation and de-aggregation schemes are subsequently proposed. For instance, optical aggregation for QPSK and corresponding de-aggregation to the two original BPSK signal and phase shift keying (PSK) signal are demonstrated in a FWM-based bidirectional architecture [20]. Based on the XPM and FWM effects, an optical aggregation for two 8QAM formats and its de-aggregation into the original QPSK signal and APSK signal are also reported [21]. However, the PSK and APSK formats to be aggregated in these schemes are rarely used, which is the limitation for practical application in EON. Compared with them, OOK and MPSK are well known to be the mainstream modulation formats in EON, which can be considered for deployment flexibly in different transmission links. For instance, OOK format is desirable in short-distance transmission because of its simple implementation. BPSK has been applied to ultra-long haul transmission because of its superior performance. QPSK and 8PSK with high spectral efficiency are also promising for long-haul transmission systems. However, to our best knowledge, there is no research on the simultaneous optical aggregation and de-aggregation for these widely exploited formats.
In this paper, a simultaneous all-optical channel aggregation and de-aggregation based on nonlinear effects for OOK and MPSK formats are designed. In detail, the incoming OOK signal and MPSK signal can be aggregated to higher bit rate 2M-ary PSK (2MPSK) signal by XPM in HNLF. Subsequent de-aggregation function will be executed to recover the original MPSK signal and OOK data by PSA. Both functions are realized in the optical domain without conventional photoelectric conversion. Besides, for the application in practice, a black-box device to extract the phase locked pump for PSA-based de-aggregation is also designed. Both 20-Gbps aggregated QPSK signal (M=2) and 30-Gbps aggregated 8PSK (M=4) signal as examples are studied, and the input-output constellations and waveform shows the feasibility and tunability of the whole scheme. Finally, the quality of recovered OOK and MPSK signals are evaluated with the varying optical signal-to-noise ratio (OSNR) of the two aggregated Q/8PSK signal transmitted in the link, and the whole system bit-error rate (BER) performance is analyzed in the same link noise environment.
2. Operation principle
2.1 Optical channel aggregator
As shown in Fig. 1, the OOK (Fig. 1(a)) is a pump signal and the MPSK (Fig. 1(b) or Fig. 1(c) as an example) is a probe signal. After the XPM effect in HNLF1, the aggregated 2MPSK signal at ${W_2}$ ((Fig. 1(d) or Fig. 1(e))) satisfies the relations:
Fig. 1. Scheme of optical channel aggregator with constellation diagrams (a) OOK b) BPSK (M=2) c) QPSK (M=4), (d) aggregated QPSK (M=2) or (e) aggregated 8PSK (M=4).
The mapping and logic relations between inputs and aggregated Q/8PSK signal in the two cases are summarized in Table 1. By observing the binary data encoded on the aggregated 2MPSK signal, we can find the first bit of 2MPSK has an identical logic relation with OOK, and the other bits are the same with MPSK.
Table 1. The optical mapping and logic pattern among the signals in aggregation.
To study the phase characteristic of aggregated 2MPSK, $\gamma = 13.1{{{W^{ - 1}}} \mathord {\left / {\vphantom {{{W^{ - 1}}} {km}}} \right.} {km}}$ and L=1km, the relationship between the OOK power ${P_{OOK}}$ and induced phase shift $\Delta \varphi$ on MPSK can be verified by the theoretical value derived from Eq. (2) and actual measured value of the simulation setup. As shown in Fig. 2(a), to obtain the aggregated QPSK (M=2) and 8PSK signal (M=4) , the required OOK average power should be 15mW and 31mW to induce a phase shift of ${\pi \mathord {\left /{\vphantom {\pi 2}} \right.} 2}$ and ${\pi \mathord {\left /{\vphantom {\pi 4}} \right.} 4}$ respectively. Meanwhile, the measured value is lower because there is small power flow between MPSK and every occurrence of the bit ’1’ of OOK signal in the additional parametric gain process. The trend will become more obvious as the power of OOK increases. Therefore, the amplitude of aggregated 2MPSK signal may be not completely constant like MPSK as described in Eq. (1) but with a jitter.
Fig. 2. (a) Theoretical value and actual measured value of the phase shift $\Delta \varphi$ on MPSK varies with the OOK power; (b) The amplitude jitter of Q/8PSK varies with the corresponding B/QPSK power under the fixed OOK power of 31mW and 15mW respectively.
The amplitude jitter of 2MPSK, which is defined the ratio of its amplitude ones with phase shift 0 and ${\pi \mathord {\left /{\vphantom {\pi M}} \right.} M}$. To observe the amplitude characteristic of aggregated 2MPSK, the amplitude jitter of Q/8PSK as examples varies with the corresponding B/QPSK power under the fixed OOK power are studied respectively. As shown in Fig. 2(b), the 2MPSK signal would have a constant amplitude envelope when the power of OOK and MPSK are comparable since that the power flow between them is the weakest. In our case, they are 27mW BPSK with 31mW OOK for aggregated QPSK signal, and 14mW QPSK with 15mW OOK for aggregated 8PSK signal.
2.2 Optical channel de-aggregator
The de-aggregator is designed at the destination node to recover the original MPSK and OOK signal from the aggregated 2MPSK signal, so that both low-speed traffic can be received by their own users respectively.
According to Eq. (2), for aggregated 2MPSK, the information of MPSK and OOK corresponds to ${\varphi _{\textrm {MPSK}}}$ and $\Delta \varphi$ respectively. Therefore, the MPSK signal ${\varphi _{\textrm {MPSK}}}$ can be recovered by squeezing two adjacent phase points ${\varphi _{\textrm {MPSK}}}$ and ${\varphi _{\textrm {MPSK}}} + \Delta \varphi$ to eliminate OOK information $\Delta \varphi$ with phase squeezing axis ${\varphi _{\textrm {PS}}} = {\varphi _{\textrm {MPSK}}} + {{\Delta \varphi } \mathord {\left / {\vphantom {{\Delta \varphi } 2}} \right.} 2}$. For example, the squeezing process of QPSK (M=2) and 8PSK (M=4) are illustrated in Figs. 3(a)–3(b) and Figs. 3(d)–3(e) respectively. Similarly, the mapping and logic relations between input and recovered signals are summarized in Table 2. Combined with Table 1, the recovered binary data are the same with the original data encoded on B/QPSK.
Fig. 3. Constellation diagrams of (a) aggregated QPSK signal (M=2), (b) recovered BPSK and (c) ’OOK’ signals; (d) aggregated 8PSK signal (M=4), (e) recovered QPSK and (f) ’OOK’ signal.
Table 2. The optical mapping and logic pattern among the signals in de-aggregation.
As for OOK data, $\Delta \varphi$ can be recovered when phase squeezing axis is aligned to ${\varphi _{MPSK}} + \Delta \varphi$. In this way, different amplitude gain axes ${\varphi _{\textrm {AG1}}}$ and ${\varphi _{\textrm {AG2}}}$ would appear and align to ${\varphi _{\textrm {MPSK}}}$ and ${\varphi _{\textrm {MPSK}}} + \Delta \varphi$ as shown in Figs. 3(a)–3(c) and Figs. 3(d)–3(f) respectively. With the help of photodiode, the OOK amplitude information $\Delta \varphi$ can be detected and the MPSK phase information ${\varphi _{\textrm {MPSK}}}$ would be eliminated. As shown in Table 2, the recovered binary data are also consistent with the original data encoded on OOK.
The designed de-aggregator is applicable for 2MPSK format, which can recover MPSK and OOK signals simultaneously as depicted in Fig. 4. It has two basic building block, the phase conjugator and $90^\circ$ optical hybrid.
Fig. 4. (a) Scheme of optical channel de-aggregator from aggregated 2MPSK signal into MPSK and OOK signals; (b) and (c) are the corresponding schematic signal spectra before and after the cascaded FWM process in the HNLF at port C and port D respectively
The phase conjugator is used to obtain the original signal S and phase conjugated signal (M-1)S* at different ports A2 and B2 separately by dual-pump FWM in HNLF with a nonlinear optical loop mirror (NOLM) [22]. As shown in Fig. 4(a), the 2MPSK signal (S) at ${f_s}$ combined with a continuous wave(CW) as pump1(P1) at ${f_s} - \Delta f$ are incident to port A, and CW pump2 (P2) at ${f_s} + \left ( {M - 1} \right )\Delta f$ is input to port B in the opposite direction. The electric field of the launched signals at input ports A1 and B1 can be expressed as:
The $90^\circ$ optical hybrid is used to obtain required phase squeezing axis ${\varphi _{\textrm {PS}}}$ and different amplitude gain axis ${\varphi _{\textrm {AG}}}$ simultaneously to recover MPSK and OOK signal by interfering signal and idler in two different paths. The output signals at ports G and H can be calculated as:
In order to calculate the exact value of m for the de-aggregation scheme, one phase state ${\varphi _\textrm {m}} = {\pi \mathord {\left / {\vphantom {\pi M}} \right.} M}$ of 2MPSK squeezed toward ${\varphi _{\textrm {PS}}} = {\pi \mathord {\left / {\vphantom {\pi {2M}}} \right.} {2M}}$ as an example is illustrated in Fig. 5.
According to the triangle sine theorem, the relation satisfies:
then the MPSK signal can be recovered when $\textrm {m} = \tan \left ( {{\pi \mathord {\left / {\vphantom {\pi {\textrm {2M}}}} \right.} {\textrm {2M}}}} \right )$, and the extinction ratio of the recovered OOK data is ${{\left ( {1 + \sin \left ( {{\pi \mathord {\left / {\vphantom {\pi M}} \right.} M}} \right )} \right )} \mathord {\left / {\vphantom {{\left ( {1 + \sin \left ( {{\pi \mathord {\left / {\vphantom {\pi M}} \right.} M}} \right )} \right )} {\left ( {1 - \sin \left ( {{\pi \mathord {\left / {\vphantom {\pi M}} \right.} M}} \right )} \right )}}} \right.} {\left ( {1 - \sin \left ( {{\pi \mathord {\left / {\vphantom {\pi M}} \right.} M}} \right )} \right )}}$ in such condition.For example, in the case of aggregated QPSK signal, M=2, ${\varphi _\textrm {i}} = {\varphi _{P1}} - {\varphi _S} + {\varphi _{P2}}$, $m$=1 and ${\varphi _{\textrm {P1}}} + {\varphi _{\textrm {P2}}} - 2{\varphi _{\textrm {S0}}} = 0 $, the output signals at ports G and H can be expressed as:
Fig. 6. (a) Phase-to-phase and (b) phase-to-amplitude gain transfer functions for QPSK (M=2) de-aggregation to BPSK and ’OOK’ respectively;(c) Phase-to-phase and (d) phase-to-amplitude gain transfer functions for 8PSK (M=4) de-aggregation to QPSK and ’OOK’ respectively.
Moreover, in the case of aggregated 8PSK signal, M=4, ${\varphi _\textrm {i}} = 3{\varphi _{P1}} - 3{\varphi _S} + {\varphi _{P2}}$, $m$=0.414 and $ 3{\varphi _{\textrm {P1}}} + {\varphi _{\textrm {P2}}} - 4 {\varphi _{\textrm {S0}}} = 0 $, the output signals at ports G and H can be expressed as:
2.3 Black box
According to Eq. (7), the critical phase relationship among signal carrier, CW P1 and P2 needs to be stable by employing phase-locked waves rather than three independent lasers for de-aggregation in practical system. P2 is related to the M-th signal harmonic ${\varphi _{\textrm {MH}}}$ generated by the interaction of both P1 and the signal after FWM, which can be expressed as:
The principle of extracting the carrier from carrier-less BPSK signal by feed-forward based modulation stripping with a DPSK optical-electrical-optical (OEO) regenerator has been reported in [23]. In our case, as shown in Fig. 7, the 2MPSK signal is first sent to the pump phase-locking device to generate M-th harmonic as BPSK signal by mixing with a free-running P1 in HNLF. Second, the BPSK signal is split to path a and path b. In path b, the signal is fed to a DPSK demodulator consisting of a 1-bit delay interferometer and then a balanced photodiode to detect the differentially demodulated signal. Finally, the received electrical signal passes to a differential encoder to restore the original logic sequence, which is used to drive a BPSK modulator consisting of a phase modulator in path a. Therefore, the data modulation can be stripped off the BPSK signal by re-modulating it with a complimentary data pattern and the phase-locked P2 can be obtained at the output of pump phase-locking device.
Fig. 7. (a) Black box schematic of phase locked Pump2 extraction device; (b) Optical spectrum of input 2MPSK signal (M=2); (c) Measured self-heterodyned linewidth for output P2.
Taking M=2 as an example, an aggregated QPSK (Fig. 7(b)) with 20 Gbps is input and the line width of output P2 is 110 kHz (Fig. 7(c)), which is broadened from the laser (5KHz). It is mainly due to both the noise generated in the aggregation and the residual modulation components in this phase-locking process. In practice, an injection locked laser (ILL) is often used further to narrow the line width [24].
3. Setup and results
Figure 8 shows the whole simulation setup of channel aggregation and de-aggregation for OOK and MPSK signals.
Fig. 8. (a) The whole setup for optical channel aggregator and de-aggregator for OOK and MPSK signals and (b-e) the corresponding optical spectra of aggregated QPSK (M=2) and 8PSK (M=4) as examples before and after the HNLF2 respectively.
At the transmitter one 10 Gbaud B/QPSK signal is generated by modulating a CW laser of 5 KHz line width in a phase modulator/IQ modulator at 193.1 THz with ${2^{15}} - 1$ pseudo-random binary sequence (PRBS). Another 10-Gbps OOK signal is generated at 193.05 THz in an amplitude modulator with ${2^{15}} - 1 $ PRBS. Synchronization between the B/QPSK and OOK signals is adjusted by the optical tunable delay line (TDL), and the polarization controllers (PC) is used to adjust the polarization of OOK for the best aggregation performance. Both signals are amplified, combined and launched into the 2MPSK aggregator simultaneously. For QPSK (M=2) aggregator, as illustrated in Fig. 2, the optimal power of BPSK signal and corresponding OOK signal are 31mW and 27mW respectively. Similarly, for 8PSK (M=4) aggregator, the optimal power of QPSK signal and corresponding OOK signal should be 14mW and 15mW respectively. The employed HNLF1 has a nonlinear coefficient $\gamma = 13{{{\textrm {W}^{ - 1}}} \mathord {\left / {\vphantom {{{\textrm {W}^{ - 1}}} {\textrm {km}}}} \right.} {\textrm {km}}}$, length $L = 1km$, and zero-dispersion frequency at 193.1 THz with its slope of ${{{{0.02ps} \mathord {\left / {\vphantom {{0.02ps} {n{m^2}}}} \right.} {n{m^2}}}} \mathord {\left / {\vphantom {{{{0.02ps} \mathord {\left / {\vphantom {{0.02ps} {n{m^2}}}} \right.} {n{m^2}}}} {km}}} \right.} {km}}$. An band pass filter (BPF) with 3-dB bandwidth of 0.32 nm at 193.1 THz is inserted after HNLF1 to filter the aggregated Q/8PSK signal. To simulate the practical situation of the signal transmitting in fiber links, variable amplifier spontaneous emission (ASE) noise is added at node A corresponding to different optical signal-to-noise ratio (OSNR) of the aggregated signal.
At the de-aggregator, the aggregated Q/8PSK signal with a PC and a local CW P1 at 193.05 THz are first launched into a 3-dB coupler. Then the output signal and CW P1 on the upper arm, and phase locked CW P2 generated through phase locking device on the lower arm are sent to the phase conjugator with the desired polarization. The phase conjugator consists of a bidirectional HNLF2 with same parameters as HNLF1, a 3-dB coupler, two circulators and two BPFs centered at 193.1THz. The corresponding optical spectra of input Q/8PSK signals along with P2 at 193.15THz and 193.25THz before the HNLF2 at node C are shown in Figs. 8(b) and 8(d) respectively, and the corresponding optical spectra after the HNLF2 at node D are shown in Figs. 8(c) and 8(e) respectively. It can be measured that the new generated conjugated idler signals ${S^*}$ (M=2) and $3{S^*}$ (M=4) have power level around 24 dB and 49 dB less than the input QPSK signal and 8PSK signal respectively, showing the conversion efficiency (CE) of the FWM process. Some spikes in the spectra are the FWM components, which are generated solely by the pump interactions. After that, a variable optical attenuator (VOA) is used to adjust the relative intensity of the signal to the conjugated idler with a power ratio as ${1 \mathord {\left / {\vphantom {1 {{{\tan }^2}\left ( {{\pi \mathord {\left / {\vphantom {\pi {2M}}} \right.} {2M}}} \right )}}} \right.} {{{\tan }^2}\left ( {{\pi \mathord {\left / {\vphantom {\pi {2M}}} \right.} {2M}}} \right )}}$ and the phase shifter (PS) introduces a static phase compensation between them. We need to point out that the stable phase relationship of the phase-locked waves needs to be maintained when they are processed and launched to the $90^\circ$ optical hybrid separately. In practice, an active phase-locking loop (PLL) is usually employed to stabilize the phase relationship [25]. Finally, signal and conjugated idler are launched into $90^\circ$ optical hybrid for required ${\varphi _{\textrm {PS}}}$ and ${\varphi _{A\textrm {G}}}$ by interference in two different paths at the same time, so that B/QPSK signal and OOK data can be recovered simultaneously and received by their own receivers.
When no noise is added to the link, the constellations of the input-output signals and waveforms of input-output OOK data corresponding to aggregated QPSK and 8PSK signals are shown in Fig. 9. A BPSK signal (bit rate: 10 Gbps, phase noise (PN): $2.8^\circ$) and a OOK signal (bit rate: 10 Gbps, amplitude standard deviation (ASD): 0.66$\%$) as inputs to the system are first aggregated to the QPSK signal (bit rate: 20 Gbps, PN: $4.8^\circ$, ASD: 0.74$\%$), and then de-aggregated towards phase squeeze axis ${\varphi _{\textrm {PS}}}$ as BPSK signal (bit rate: 10 Gbps, PN: $4.8^\circ$) and amplitude gain axis ${\varphi _{A\textrm {G}}}$ as ’OOK’ signal (bit rate: 10 Gbps, ASD: 2.12$\%$). Similarly, a QPSK signal (bit rate: 20 Gbps, PN: $1.5^\circ$) and the OOK signal (bit rate: 10 Gbps, ASD: 0.68$\%$) can be first aggregated to the 8PSK signal (bit rate: 30 Gbps, PN: $1.5^\circ$, ASD: 0.72$\%$) and then de-aggregated into the original QPSK signal (bit rate: 20 Gbps, PN: $3.8^\circ$) and OOK data (bit rate: 10 Gbps, ASD: 2.01$\%$). The waveforms of both input-output OOK data also shows the feasibility of simultaneous all-optical channel aggregation and de-aggregation with the proposed architecture.
To further study the dependence of the recovered signal quality over the link noise, ASE noise is added to the Q/8PSK signal by setting the OSNR at node A. Figure 10 presents the PN of recovered B/QPSK signal and ASD of recovered OOK signal with the varying OSNR of the Q/8PSK signal. In Fig. 10(a), comparing with aggregated QPSK signal, the PN of recovered BPSK is squeezed to a quite low level due to the perfect two-level phase-to-phase transfer curve as shown in Fig. 6(a). Meanwhile, the ASD of the recovered OOK shows an overall degradation due to the sinusoidal amplitude response curve in Fig. 6(b), which means more extra amplitude noise will generate when a noisy signal is input. In Fig. 10(b), the PN of recovered QPSK is very comparable with the aggregated 8PSK signal due to the imperfect four-level phase-to-phase transfer curve in Fig. 6(c). Similarly, the ASD of the recovered OOK also has an overall degradation trend, which can be explained by the sinusoidal amplitude response curve in Fig. 6(d).
Fig. 10. (a) The PN of the aggregated QPSK (M=2) and recovered BPSK signals and the ASD of the aggregated QPSK and recovered OOK signal with the varying OSNR of the QPSK signal; (b) The PN of the aggregated 8PSK (M=4) and recovered QPSK signals and the ASD of the aggregated 8PSK and recovered OOK signal with the varying OSNR of the 8PSK signal.
The system bit-error rate (BER) performance versus different link noise at node A is also investigated by error counting with Monte-Carlo approach. As shown in Fig. 11(a), three 10-Gbaud back-to-back (B2B) OOK, BPSK signal and QPSK which have the same OSNR with the aggregated QPSK signal transmitted in the link, are taken into consideration for reference. On the forward error correction (FEC) threshold of $3.8 \times {10^{ - 3}}$, comparing with the B2B QPSK signal, there is 1.7 dB OSNR penalty for aggregated QPSK signal, which means a small amount of noise is introduced in the aggregation process. Compared with other B2B signals, 2.7 dB OSNR penalty for recovered BPSK and 2.5 dB OSNR penalty for recovered OOK can be observed respectively. The former is mainly caused by the worse BER tolerance of its aggregated QPSK format than the BPSK format under the same OSNR, which in turn degrades the OSNR of recovered BPSK signal. The latter is mainly due to the degraded ASD curve as shown in Fig. 10(a). Comparing the aggregated QPSK with the recovered signals, the required OSNR for QPSK is 6.5 dB on the FEC threshold, while the required OSNRs reduces to 5.2 dB for BPSK and increases to 7.9 dB for OOK respectively. The result is consistent with Fig. 10(a) that the OSNR improvement is benefited from the perfect two-level phase squeezing effect and the OSNR penalty is caused by the added AN. For error-free recover, the BER of the recovered BPSK and OOK signals should be below the FEC threshold with OSNR of aggregated QPSK transmitted in link over 7.9 dB.
Fig. 11. BER curves versus varying OSNR of (a) the aggregated QPSK (M=2) and (b) the aggregated 8PSK (M=4) transmitted in the link. R-B/QPSK: Recovered B/QPSK, R-OOK: Recovered OOK, A-Q/8PSK: aggregated Q/8PSK, FEC thr: forward error correction threshold.
The system BER performance corresponding to the aggregated 8PSK format has a similar trend as shown in Fig. 11(b). On the FEC threshold, comparing with the B2B 8PSK signal, the OSNR penalty for aggregated 8PSK signal is still low as 1.3 dB, showing the good BER performance during aggregation. Compared with B2B QPSK signal, the OSNR penalty for recovered QPSK raises to 5 dB. Besides the worse BER tolerance of its aggregated 8PSK format mentioned above, it is also due to the imperfect four-level phase squeezing effect as shown in Fig. 10(b). Meanwhile, the OSNR penalty for recovered OOK is also increased to 11.2 dB. On one hand, it is due to the added AN as shown in Fig. 10(b). On the other hand, it is caused by the closer symbol distance, since the recovered signal has a lower extinction ratio and worse noise margin than the original OOK signal. Comparing the aggregated 8PSK with the recovered signals, the required OSNR for 8PSK is 11.3 dB on the FEC threshold, and is reduced to 9.6dB for QPSK and increased to 16.4dB for OOK respectively. The OSNR improvement is benefited from the lower modulation format of QPSK signal with better noise tolerance, and the OSNR penalty is caused by the added AN as shown in Fig. 10(b). For error-free recover, the BER of the recovered QPSK and OOK signals should be below the FEC threshold with OSNR of 8PSK over 16.4 dB.
4. Discussion
Furthermore, the proposed scheme has also several potential applications in EON. Firstly, the single optical aggregator can be used as an optical modulator of high-speed 2MPSK formats. In comparison with conventional IQ modulators driven by multilevel electrical signals, multiple modulator combinations can release the device bandwidth requirement. Secondly, the single de-aggregator has possible application to enhance the security of communication. The information from one user can be de-aggregated into two tributaries, and they can propagate using multi-core fibers on different spatial paths. Reliable communication is allowed only when both signals are simultaneously detected.
5. Conclusion
We designed a simultaneous all-optical channel aggregation and de-aggregation system for OOK and MPSK formats. Considering the actual application, a black-box phase locking pump device is also demonstrated. In order to verify the feasibility of the proposed scheme, two aggregated QPSK and 8PSK signals as examples are studied respectively by analyzing the input-output constellations and waveforms, the quality of recovered signals, and the BER performance of the whole system. For error-free recover of the OOK and B/QPSK signals, the OSNR of aggregated QPSK and 8PSK transmitted in the link should be over 7.9 and 16.4 dB respectively. Besides the intelligent grooming function, the proposed scheme may also be useful for optical modulator of MPSK signal and secure communication, which is promising to enhance network spectral efficiency, flexibility and security in EON.
Funding
National Natural Science Foundation of China (61831003, 61871051); National Key R&D Program of China (2018YFB1800802); BUPT Excellent Ph.D Students Foundation (CX2019216).
References
1. Y. F. Ji, J. W. Zhang, Y. l. Zhao, X. S. Yu, J. Zhang, and X. Chen, “Prospects and research issues in multi-dimensional all optical networks,” Sci. China Inf. Sci. 59(10), 101301 (2016). [CrossRef]
2. O. Gerstel, M. Jinno, A. Lord, and S. J. B. Yoo, “Elastic optical networking: a new dawn for the optical layer?” IEEE Commun. Mag. 50(2), s12–s20 (2012). [CrossRef]
3. Y. F. Ji, J. W. Zhang, X. Wang, and H. Yu, “Towards converged, collaborative and co-automatic (3C) optical networks,” Sci. China Inf. Sci. 61(12), 121301 (2018). [CrossRef]
4. Y. F. Ji, H. X. Wang, J. B. Cui, M. T. Yu, Z. T. Yang, and L. Bai, “All-optical signal processing technologies in flexible optical networks,” Photonic Netw. Commun. 38(1), 14–36 (2019). [CrossRef]
5. K. Mishina, S. Kitagawa, and A. Maruta, “All-optical modulation format conversion from on-off-keying to multiple-level phase-shift-keying based on nonlinearity in optical fiber,” Opt. Express 15(13), 8444–8453 (2007). [CrossRef]
6. B. Zhang, H. Zhang, C. Yu, X. Cheng, K. Y. Yong, P. K. Kam, J. Yang, H. Zhang, Y. H. Wen, and K. M. Feng, “An all-optical modulation format conversion for 8QAM based on FWM in HNLF,” IEEE Photonics Technol. Lett. 25(4), 327–330 (2013). [CrossRef]
7. X. Wu, J. Wang, H. Huang, and A. E. Willner, “Experimental optical multiplexing of two 20-Gbit/s QPSK data channels from different wavelengths onto a single 40Gbit/s star 16-QAM using fiber nonlinearities,” in CLEO:2011 - Laser Applications to Photonic Applications, (Optical Society of America, 2011), p. CThH4.
8. B. M. Cannon, T. Mahmood, W. Astar, P. Boudra, T. Mohsenin, and G. M. Carter, “Polarization-insensitive phase-transmultiplexing of CSRZ-OOK and RZ-BPSK to RZ-QPSK via XPM in a PCF,” in Optical Fiber Communication Conference, (Optical Society of America, 2014), p. W2A.52.
9. Y. H. Wen and K. M. Feng, “A simple NRZ-OOK to PDM RZ-QPSK optical modulation format conversion by bidirectional XPM,” IEEE Photonics Technol. Lett. 27(9), 935–938 (2015). [CrossRef]
10. Y. H. Wen, J. W. Ho, and K. M. Feng, “Phase transparent demultiplexer for QPSK signal based on nonlinear response in a double-pass architecture,” IEEE Photonics J. 8, 1–7 (2016). [CrossRef]
11. M. Baillot, M. Gay, C. Peucheret, J. Michel, and T. Chartier, “Phase quadrature discrimination based on three-pump four-wave mixing in nonlinear optical fibers,” Opt. Express 24(23), 26930–26941 (2016). [CrossRef]
12. H. Kishikawa, N. Goto, and L. R. Chen, “All-optical wavelength preserved modulation format conversion from PDM-QPSK to PDM-BPSK using FWM and interference,” J. Lightwave Technol. 34(23), 5505–5515 (2016). [CrossRef]
13. F. Da Ros, K. Dalgaard, Y. Fukuchi, J. Xu, M. Galili, and C. Peucheret, “Simultaneous QPSK-to- $2\times$ BPSK wavelength and modulation format conversion in PPLN,” IEEE Photonics Technol. Lett. 26(12), 1207–1210 (2014). [CrossRef]
14. F. Parmigiani, G. Hesketh, R. Slavík, P. Horak, P. Petropoulos, and D. J. Richardson, “Polarization-Assisted Phase-Sensitive Processor,” J. Lightwave Technol. 33(6), 1166–1174 (2015). [CrossRef]
15. H. Liu, H. Wang, J. Cui, and Y. Ji, “Design of all-optical modulation format converter from one 8PSK to two QPSK signals based on phase sensitive amplification in elastic optical network,” IEEE Access 7, 51379–51385 (2019). [CrossRef]
16. L. Pan, H. Wang, and Y. Ji, “All-optical deaggregation from 8PSK to 3×BPSK based on FWM in HNLF,” Appl. Opt. 58(5), 1246–1252 (2019). [CrossRef]
17. A. Fallahpour, M. Ziyadi, A. Mohajerin-Ariaei, Y. Cao, A. Almaiman, F. Alishahi, C. Bao, P. Liao, B. Shamee, L. Paraschis, M. Tur, C. Langrock, M. M. Fejer, J. Touch, and A. E. Willner, “Experimental demonstration of tunable optical de-aggregation of each of multiple wavelength 16-QAM channels into two 4-PAM channels,” in Optical Fiber Communication Conference, (Optical Society of America, 2017), p. Th4I.6.
18. J. Cui, G. Lu, H. Wang, and Y. Ji, “On-chip optical vector quadrature de-multiplexer proposal for QAM de-aggregation by single bi-directional SOA-based phase-sensitive amplifier,” IEEE Access 7, 763–772 (2019). [CrossRef]
19. A. Bogris, “All-optical demultiplexing of 16-QAM signals into QPSK tributaries using four-level optical phase quantizers,” Opt. Lett. 39(7), 1775–1778 (2014). [CrossRef]
20. Y.-H. Wen, J.-W. Ho, and K.-M. Feng, “Simultaneous all-optical transparent phase multiplexing/de-multiplexing based on FWM in a HNLF,” in Optical Fiber Communication Conference, (Optical Society of America, 2016), p. W4D.1.
21. H. Liu, H. Wang, and Y. Ji, “Simultaneous all-optical channel aggregation and de-aggregation for 8QAM signal in elastic optical networking,” IEEE Photonics J. 11(1), 1–8 (2019). [CrossRef]
22. H. C. Lim, F. Futami, and K. Kikuchi, “Polarization-independent, wavelength-shift-free optical phase conjugator using a nonlinear fiber sagnac interferometer,” IEEE Photonics Technol. Lett. 11(5), 578–580 (1999). [CrossRef]
23. S. K. Ibrahim, S. Sygletos, R. Weerasuriya, and A. D. Ellis, “Novel real-time homodyne coherent receiver using a feed-forward based carrier extraction scheme for phase modulated signals,” Opt. Express 19(9), 8320–8326 (2011). [CrossRef]
24. R. Weerasuriya, S. Sygletos, S. K. Ibrahim, R. Phelan, J. O’Carroll, B. Kelly, J. O’Gorman, and A. D. Ellis, “Generation of frequency symmetric signals from a BPSK input for phase sensitive amplification,” in Optical Fiber Communication Conference, (Optical Society of America, 2010), p. OWT6.
25. A. Lorences-Riesgo, L. Liu, S. L. I. Olsson, R. Malik, A. Kumpera, C. Lundström, S. Radic, M. Karlsson, and P. A. Andrekson, “Quadrature demultiplexing using a degenerate vector parametric amplifier,” Opt. Express 22(24), 29424–29434 (2014). [CrossRef]
