Open Access
Add to BibSonomyAbstract
Optical logic gate, especially exclusive-or (XOR) gate, plays important role in accomplishing photonic computing and various network functionalities in future optical networks. On the other hand, optical multicast is another indispensable functionality to efficiently deliver information in optical networks. In this paper, for the first time, we propose and experimentally demonstrate a flexible optical three-input XOR gate scheme for multiple input phase-modulated signals with a 1-to-2 multicast functionality for each XOR operation using four-wave mixing (FWM) effect in single piece of highly-nonlinear fiber (HNLF). Through FWM in HNLF, all of the possible XOR operations among input signals could be simultaneously realized by sharing a single piece of HNLF. By selecting the obtained XOR components using a followed wavelength selective component, the number of XOR gates and the participant light in XOR operations could be flexibly configured. The re-configurability of the proposed XOR gate and the function integration of the optical logic gate and multicast in single device offer the flexibility in network design and improve the network efficiency. We experimentally demonstrate flexible 3-input XOR gate for four 10-Gbaud binary phase-shift keying signals with a multicast scale of 2. Error-free operations for the obtained XOR results are achieved. Potential application of the integrated XOR and multicast function in network coding is also discussed.
© 2016 Optical Society of America
1. Introduction
With unabated exponential growing demand for higher throughput in next-generation optical networks, transparent and scalable all-optical processing is highly desirable to overcome the bottle-neck imposed by optical-electrical-optical conversion. Moreover, since phase modulation formats and high-order modulation have been deployed to improve the spectrum efficiency and transmission capacity in optical communication systems, all-optical signal processing for phase-modulated signals has attracted lots of research attention. Among various all-optical signal processing techniques, optical logic gate has been recognized as one of the key elements to realize all-optical network functionalities and photonic computing. In particular, optical exclusive-or (XOR) gate is one critical logic operation since it is used in many fundamental functionalities, such as header/label processing in optical packets, parity check, data encryption, error detection/correction, digital comparison and so on. Optical XOR gate operation has been experimentally demonstrated in a number of platforms like semiconductor optical amplifier (SOA) [1, 2], highly-nonlinear fiber (HNLF) [3], periodically-poled lithium niobate [4], and silicon nanowire waveguide [5]. On the other hand, with the emergence of high-bandwidth point-to-multipoint applications such as high-definition Internet TV, big-data sharing and data center migration, the need for optical multicast has arisen recently to improve the network throughput and decrease the blocking probability in optical networks. All-optical multicast of binary phase-shift keying (BPSK) signal through the nonlinearities in HNLF [6, 7], SOA [7] and silicon nanowire waveguide [8] has been reported. Obviously, optical XOR gate operation and all-optical multicast have been extensively studied in past decades. To lower power consumption, complexity and cost of all-optical processing systems, it is of particular interest to integrate multiple functionalities to a single unit in a more compact manner, which brings benefits in terms of efficiency in power consumption and implementation cost, and also creates new application in optical networks. For instance, optical XOR gate with optical multicast functionality has potential application in realizing optical physical-layer network coding to increase the network throughput, or protect link or node failures in optical networks [9]. Recently, we have experimentally demonstrated the function integration of optical three-input-XOR gate and multicast through FWM in quantum-dot SOA [10, 11]. It is desirable to further improve the flexibility and efficiency of the function-integrated XOR and multicast unit, especially when the network scale is increased.
Here, for the first time, a flexible three-input XOR gate among multiple input phase-modulated signals through self-pumped FWM in HNLF is presented. XOR operation among any three out of multiple input signals is simultaneously implemented by using single nonlinear media. Meanwhile, for each obtained optical XOR gate, the XOR result is multicasted to two wavelengths by sharing the same nonlinear media. The re-configurability of the proposed XOR gate and the function integration of optical XOR gate and optical multicast could be potentially used for the next-generation transparent optical networks. Particularly, application scenarios of XOR gate with multicast functionality in optical physical-layer network coding are discussed. The proposed scheme has the following features: (i) simultaneous realization of all of 3-input-XOR operations among input signals; (ii) function integration of optical logic gate and multicast in single device; (iii) flexibility in the three-input XOR operation configuration by selectively choosing the participant light from input signals; (iv) no need of additional pump to implement multicast by self-pumped FWM. In order to verify the proposed scheme, three-input XOR gate with optical multicast for binary phase-shift keying (BPSK) at 10Gbaud is experimentally demonstrated. Error-free operation of the generated XOR results with a multicast scale of 2 is achieved.
2. Flexible 3-input XOR gate with multicast functionality
2.1 Operation principle
Figure 1(a) depicts the concept of the proposed flexible three-input-XOR gate with multicast functionality. First, logic operation is simultaneously performed over any three out of N input data carried at different wavelengths with n replicas. After the logic operation and multicast, through an optical wavelength de-multiplexer like array-waveguide grating (AWG), the corresponding XOR results are led to different ports and delivered to the next network nodes distinguished by wavelengths. As an example of the proposed flexible XOR gate, Fig. 1(b) shows a flexible 3-input XOR gate with four input data with a multicast scale of 2, which will be experimentally demonstrated in next section.
Owing to the modulation transparency, FWM is deployed here to simultaneously realize all of possible XOR operations among input signals with multicast functionality. Figure 2 illustrates the operation principle of the proposed flexible three-input XOR gate with multicast. With amplified four input signals (S1~S4) at ω1, ω2, ω3, and ω4, after 3rd-order nonlinear media like HNLF or SOA, several FWM components generate behind the input signals. Note that the frequency spacing between ω2 and ω3 (Δω) is twice that between ω1 and ω2, i.e. 2Δω. It is to avoid spectrum overlapping among generated FWM components. Among these components, we are particularly interested in the components at ωxyz*, where x,y,z ∈ [1–4], x≠y≠z, and * denotes the conjugate operation, i.e. ω321*, ω213*, ω431*, ω413*, ω432*, ω423*, ω421*, and ω412*. The corresponding electrical field of these spurious components at ωxyz* is given by
The coefficient kxyz in the equation is a constant proportional to the FWM efficiency. It is clear that the phase of these components is given by θxyz* = θx + θy–θz. With the phases of input signals taking values either “0” or “π”, i.e. BPSK modulation, logically, the obtained components correspond to the XOR operation among any three out of multiple input signals, i.e. θxyz* = θx⊕θy⊕θz where x,y,z ∈ [1–4], x≠y≠z. With four input BPSK signals, we could simultaneously achieve four XOR operations over any three input signals. Moreover, with the configuration shown in Fig. 2, each three-input XOR result has two replicas, thus totally resulting in eight XOR components. The generated XOR results could be flexibly selected with desired XOR participators and number of replicas using wavelength selective components. Thus it is referred to as flexible optical XOR gate with multicast functionality here.2.2 Application in optical network coding
Recently, networking coding (NC) has been proposed to improve network efficiency and security [9, 12, 13]. Particularly, in optical networks, as a derivative of NC, optical physical-layer network coding (OPNC) has been proposed and demonstrated for boosting network efficiency, protecting the network against failures, or increasing system throughput. To implement OPNC, several schemes based on optical XOR gate have been demonstrated in simple network topologies where less than 4 network nodes are considered. However, with the increase of the network scale, a flexible and re-configurable XOR operation would help to further improve the network efficiency in OPNC. Figure 3(a) depicts a network with six nodes, where nodes A and B are going to broadcast data to nodes a and b, simultaneously. After the data λA and λB reach the node CX, conventionally, two separate wavelength channels are required in order to deliver the data to the destinations. With OPNC, NC could be performed at the node CX to form common data λNC, i.e. λNC = λA⊕λB. At the node CM, the NC data is forwarding to the destinations a and b, i.e. multicast. With the received data λA(B) and λNC, λB(A) is finally recovered at the sink node a(b) by performing another XOR operation. The procedure involving XOR-based coding/decoding and multicast here is referred to as OPNC. It is clear that the proposed XOR gate with multicast functionality could be deployed at the common channel from CX to CM to implement the OPNC. With the further increase of network scale, a flexible XOR operation subsystem becomes more desirable to efficiently configure network based on available resources in a network. Figure 3(b) and 3(c) shows two different topologies, where three and two common channels are deployed, respectively. In both topologies, the data from the source nodes (A~D) could be broadcasted to the sink nodes (a~d). With less common channels, more direct links from source to sink nodes are required in order to eventually recover all of the source data. For example, with three common channels shown in Fig. 3(b), one direct channel from source to sink node is required, whereas two direct links are desired in the case with two common channels (Fig. 3(c)). Note that the XOR patterns in Fig. 3(b) and 3(c), i.e. channel combination in XOR operations, are just shown as examples here. The XOR pattern could be adaptively configured among all of possible combinations. On the other hand, in our proposed flexible XOR gate, all of 3-input-XOR operations could be simultaneously obtained among any three input signals. For example, with four source nodes, there are total four XOR patterns. The desired XOR signals could be flexibly picked up by the followed wavelength selective component. If applying the flexible XOR gate to implement OPNC in networks, the flexibility of the proposed subsystem could allow the network operators dynamically establish network connections based on the available network resources, which effectively improves the network efficiency.
Fig. 3 Application scenarios of the integrated XOR and multicast in optical physical-layer network coding. OPNC networks with (a) one, (b) three, and (c) two common channels.
3. Experiments and results
To verify the proposed scheme, an experiment operated at 10Gbaud is carried out. The experimental setup is shown in Fig. 4. Light from four lasers with linewidths of around 150 kHz at 1544.2 nm (ω1), 1543.4 nm (ω2), 1541.8 nm (ω3) and 1550.6 nm (ω4) are combined by polarization-maintained coupler and launched to phase modulator driven by 10Gbaud pseudorandom bit sequence (PRBS) with length of 215-1 from pulse-pattern generator (PPG) to perform BPSK modulation. In order to de-correlate these four input signals, after power amplification, they are separated by using an AWG, and three of them are delayed by different integral periods between each other via optical delay lines (ODL1, ODL2 and ODL3). With respect to the input signal at ω1, 137, 109, and 31 period relative delay were obtained for the signals at ω2, ω3 and ω4, respectively. Polarization controllers (PCs) are used for each branch to optimize FWM efficiencies. After power amplification, four input signals are re-combined and led to a piece of HNLF, which has a length of 200 meter, an attenuation coefficient of 1.5 dB/km, a nonlinear coefficient of 18/W/km, a zero-dispersion wavelength of 1551 nm and a dispersion slope of around 0.025ps/nm2/km. After HNLF, optical band-pass filter (BPF) is used to select the desired FWM components. Then the selected components are de-modulated by a differential phase-shift keying (DPSK) receiver, consisting of two cascaded EDFAs, Mach-Zehnder delay interferometer (MZDI) and balanced photo-detector.
In the experiment, the total power launched to HNLF is first optimized by measuring the conversion efficiency and bit-error-rate (BER) of the converted signal at ω432* when tuning the launched power. Note that the launched power here is referred to the total power of input signals launched into the HNLF. As shown in Fig. 5, when the launched power is increased till around 20.3 dBm, the measured BER reaches the minimum value with a reasonable conversion efficiency of around −22 dB. Here, the conversion efficiency is defined as the power ratio between the converted signal and the input signal at ω4. With the launched power of around 20.3 dBm, the optical spectrum after HNLF is shown in Fig. 6. Taking the power of S4 as reference, the measured conversion efficiency of the generated components at ω321*, ω213*, ω431*, ω413*, ω432*, ω423*, ω421*, and ω412* varies from −18 dB to −25 dB, which ensures the performance of the multicasted XOR signals. Further performance improvement of the XOR results is expected if the launched power of each input signal could be individually optimized.
To verify the XOR operation and multicast function, bit patterns of input and generated XOR components after de-modulation are measured and depicted in Fig. 7. It is clear that the measured data pattern carried by the FWM components follows the XOR logic relationship with the corresponding input signals, fulfilling the expectations of XOR operation. Note that the polarity of the measured data pattern is dependent on the bias of the MZDI. With four input BPSK signals, three-input XOR operations are obtained with four possible combinations. Meanwhile, two replicas of each XOR operation result are obtained to realize optical multicast. To further verify the performance of the proposed XOR gate, BER curves of input and converted XOR results are measured and plotted in Fig. 8. A sensitivity of around −39.8 dBm at BER of 10−9 is observed for the input data streams. After XOR and multicast, less than 2 dB power penalty is obtained for eight XOR signals with error-free operations. This verifies the feasibility of the proposed flexible XOR gate with multicast function. The power penalty obtained here could be attributed to the crosstalk induced by the existence of multiple FWM spurious components, especially behind the bundle of high-power input signals (S1 to S3). This also explains the slightly larger penalty for the generated XOR results at ω321* and ω213*. As mentioned above, since all of the input signals shared single EDFA before the HNLF, only the launched total power into the HNLF could be optimized. If the launched power of each input could be individually optimized, it will be helpful to minimize the possible crosstalk induced by surrounding FWM spurious components, thus further improving the performance of the XOR results.
Fig. 8 Measured BER curves of input signals and XOR results. The denotation XORxyz* represents the obtained XOR signals at ωxyz*, and INPUTx represents the input signal at ωx.
4. Conclusion
We have proposed and experimentally demonstrated a flexible XOR gate with multicast functionality through self-pumped FWM in single piece of HNLF. All of the possible XOR operations are simultaneously obtained after HNLF. By using wavelength selective component, the participators in XOR operation could be flexibly selected among input signals with a desired number of replicas. A flexible XOR operation among three out of four input 10Gbaud BPSK signals with a 1-to-2 multicast scale has been experimentally demonstrated with less than 2dB power penalty for all of the XOR results. Error-free operation has been successfully achieved. The proposed function integration of flexible XOR operation and multicast in single device offers flexibility in network management, improves the network efficiency, and shows potential applications in OPNC.
Acknowledgments
We acknowledge the helpful discussion with Xun Guan from the Chinese University of Hong Kong. This work was supported in part by Grant-in-Aid for Scientific Research (C) (15K06033) from the Ministry of Education, Science, Sports and Culture (MEXT), Japan.
References and links
1. K. Chan, C.-K. Chan, L. K. Chen, and F. Tong, “Demonstration of 20-Gb/s all-optical XOR gate by four-wave mixing in semiconductor optical amplifier with RZ-DPSK modulated inputs,” IEEE Photonics Technol. Lett. 16(3), 897–899 (2004). [CrossRef]
2. N. Deng, K. Chan, C.-K. Chan, and L.-K. Chen, “An all-optical XOR logic gate for high-speed RZ-DPSK signals by FWM in semiconductor optical amplifier,” IEEE J. Sel. Top. Quantum Electron. 12(4), 702–707 (2006). [CrossRef]
3. J. Wang, Q. Sun, and J. Sun, “All-optical 40 Gbit/s CSRZ-DPSK logic XOR gate and format conversion using four-wave mixing,” Opt. Express 17(15), 12555–12563 (2009). [CrossRef] [PubMed]
4. J. Wang, J. Sun, X. Zhang, D. Huang, and M. M. Fejer, “Ultrafast all-optical three-input Boolean XOR operation for differential phase-shift keying signals using periodically poled lithium niobate,” Opt. Lett. 33(13), 1419–1421 (2008). [CrossRef] [PubMed]
5. F. Li, T. D. Vo, C. Husko, M. Pelusi, D.-X. Xu, A. Densmore, R. Ma, S. Janz, B. J. Eggleton, and D. J. Moss, “All-optical XOR logic gate for 40Gb/s DPSK signals via FWM in a silicon nanowire,” Opt. Express 19(21), 20364–20371 (2011). [CrossRef] [PubMed]
6. G.-W. Lu, K. S. Abedin, and T. Miyazaki, “DPSK multicast using multiple-pump FWM in Bismuths highly nonlinear fiber with high multicast efficiency,” Opt. Express 16(26), 21964–21970 (2008). [CrossRef] [PubMed]
7. D. Wang, T.-H. Cheng, Y.-K. Yeo, Z. Xu, Y. Wang, G. Xiao, and J. Liu, “Performance comparison of using SOA and HNLF as FWM medium in a wavelength multicasting scheme with reduced polarization sensitivity,” J. Lightwave Technol. 28(24), 3497–3505 (2010).
8. M. Pu, H. Hu, H. Ji, M. Galili, L. K. Oxenløwe, P. Jeppesen, J. M. Hvam, and K. Yvind, “One-to-six WDM multicasting of DPSK signals based on dual-pump four-wave mixing in a silicon waveguide,” Opt. Express 19(24), 24448–24453 (2011). [CrossRef] [PubMed]
9. Y. An, F. D. Ros, and C. Peucheret, “All-optical network coding for DPSK signals, ” Proc. OFC’13, paper JW2A.60, 2013.
10. J. Qin, G.-W. Lu, T. Sakamoto, K. Akahane, N. Yamamoto, D. Wang, C. Wang, H. Wang, M. Zhang, T. Kawanishi, and Y. Ji, “Simultaneous multichannel wavelength multicasting and XOR logic gate multicasting for three DPSK signals based on four-wave mixing in quantum-dot semiconductor optical amplifier,” Opt. Express 22(24), 29413–29423 (2014). [CrossRef] [PubMed]
11. D. Wang, M. Zhang, G.-W. Lu, J. Qin, T. Sakamoto, K. Akahane, N. Yamaoto, T. Kawanishi, Z. Li, Y. Cui, “Multifunctional all-optical signal processing scheme for simultaneous multichannel WDM multicast and XOR logic gates based on FWM in QD-SOA,” Prof. OFC’15, paper Th2A.5, 2015. [CrossRef]
12. X. Guan and C.-K. Chan, “Physical-layer network coding in coherent optical OFDM systems,” Opt. Express 23(8), 10057–10063 (2015). [CrossRef] [PubMed]
13. L.-K. Chen, M. Li, and S. C. Liew, “Breakthroughs in photonics 2014: optical physical-layer network coding, recent developments, and challenges,” IEEE Photonics J. 7(3), 1–6 (2015).
