Signal processing is generally considered an efficient and powerful enabler for a host of communication functions as well as a system performance enhancer. The hope is that performing signal processing in the optical domain might reduce any optical-electronic conversion inefficiencies and take advantage of the ultrahigh bandwidth inherent in optics [1]. It is important that optical signal processing functions be capable of handling and operating on advanced modulation formats. In this manner, the bottleneck of midstream routing and processing may be alleviated. One of the basic building blocks to achieve efficient and reconfigurable signal processing is a continuously tunable optical delay line [2].

Given the present importance of spectrally efficient differential-quadrature phase-shift-keying (DQPSK) and the forthcoming IEEE $100\mathrm{Gbit}/\mathrm{s}$ Ethernet Standard, 802.3ba, which includes provisions for many potential formats [3], desirable characteristics of a tunable optical delay line may include (a) transparency to the data modulation format, (b) high-speed operation, and (c) the ability to accommodate a large range of delays.

One promising method of creating optical delays is conversion dispersion. An incoming data signal (i) is wavelength converted; (ii) the converted signal is passed through a high-chromatic-dispersion element, which produces a wavelength-dependent time delay; and (iii) the output can be wavelength-converted back to its original wavelength. Fine control of the delay resolution and high-speed, $160\mathrm{Gbit}/\mathrm{s}$, operation have been shown for relatively small delay values [4, 5], while recent results include delays $>0.5\mathrm{\mu s}$ at data rates of 10 and $40\mathrm{Gbit}/\mathrm{s}$ for data modulation formats of on–off keying (OOK) and differential phase-shift keying (DPSK) [6, 7, 8]. Recently, a combination of fixed and tunable delays demonstrated $>7\mathrm{\mu s}$ for a $10\mathrm{Gbit}/\mathrm{s}$ OOK signal [9]. A report also demonstrated the delay of an $80\mathrm{Gbit}/\mathrm{s}$ DQPSK channel by $105\mathrm{ns}$ using periodically poled lithium niobate wavelength converters [10]. A laudable goal would be to demonstrate the DQPSK format at the $100\mathrm{Gbit}/\mathrm{s}$ Ethernet standard rate and for a much longer, $>1\mathrm{\mu s}$, delay time.

Recently, we demonstrated an optically controlled tunable delay element using wavelength conversion in a highly nonlinear fiber, dispersion compensating fiber (DCF), and optical phase conjugation, enabling a continuous delay of up to $1.16\mathrm{\mu s}$ for $100\mathrm{Gbit}/\mathrm{s}$ non-return-to-zero (NRZ)-DQPSK and $50\mathrm{Gbit}/\mathrm{s}$ NRZ-DPSK modulation formats [11]. This corresponds to a time-delay bit-rate product exceeding 55,000 for $50\mathrm{Gbit}/\mathrm{s}$ DPSK and 110,000 for $100\mathrm{Gbit}/\mathrm{s}$ DQPSK. In this Letter, we further discuss the effect of dispersion slope on the delay element and the limiting factors of the delay range. Higher- order dispersion is shown to be a possible limitation for large conversion dispersion delays.

An experimental block diagram of our setup is shown in Fig. 1. $100\mathrm{Gbit}/\mathrm{s}$ NRZ-DQPSK is generated by driving two parallel integrated Mach–Zehnder modulators (MZMs) with $50\mathrm{Gbit}/\mathrm{s}$, $2^{31}-1$, PRBS (pseudorandom binary sequence) data. The data are shifted by 123 bits to decorrelate the two data streams. One data stream is removed to generate the $50\mathrm{Gbit}/\mathrm{s}$ NRZ-DPSK signal. The optical signal is then wavelength converted by using a $330\mathrm{m}$ piece of highly nonlinear fiber (HNLF) with a zero-dispersion wavelength (ZDW) of $\sim 1560\mathrm{nm}$. A one- pump degenerate four-wave mixing (FWM) approach is used to convert the signal (${\lambda }_{\mathrm{IN}}\approx 1536\mathrm{nm}$) from 1540 to $1583\mathrm{nm}$. The pump and signal powers, $12–18\mathrm{dBm}$, are kept low to prevent stimulated Brillouin scattering (SBS). The converted signal is then filtered out and sent through the DCF ($D\approx -13.3\mathrm{ns}/\mathrm{nm}$ at $1550\mathrm{nm}$) to impose a wavelength-dependent delay. Counterpropagating Raman amplification in the DCF is used to compensate for the $\sim 80\mathrm{dB}$ of loss. Following the DCF, the delayed signal is phase conjugated by using a second piece of HNLF ($\mathrm{ZDW}\approx 1558\mathrm{nm}$). Again, a one-pump FWM approach is used to upconvert the signal from $\sim 4$ to $14\mathrm{nm}$ in wavelength. Pump and signal powers were $10–15\mathrm{dBm}$ and required no SBS suppression. The signal is then pas sed through the DCF a second time to compensate for the intrachannel dispersion and to undergo a second wavelength-dependent delay. Both the original signal and the phase-conjugated signal copropagate through the DCF to allow for better balancing of the Raman amplification and an overall improved noise figure. While this may increase the nonlinear interaction between the signals, the effect on performance was less than the degradation caused by a lower noise figure in a counterpropagating configuration. A third wavelength conversion stage would be needed to return the output to its original wavelength but is not implemented. The signal is received by using a preamplified receiver with a $50\mathrm{GHz}$ delay-line interferometer and balanced photoreceiver.

The total measured delay of $1.16\mathrm{\mu s}$ is shown in Fig. 2a as a function of the first stage converted wavelength. The remaining residual dispersion (picosecond per nano meter) is removed by using a switchable fiber-Bragg- grating-based dispersion compensator to predisperse the input signal with either $+95\mathrm{ps}/\mathrm{nm}$, for converted wavelengths of 1540–$1571\mathrm{nm}$, or by $0\mathrm{ps}/\mathrm{nm}$, for wavelengths of 1568–$1583\mathrm{nm}$. The phase conjugation distance can then be adjusted to match the dispersion of both passes through the DCF as determined by the amount of predispersion and the dispersion slope of the DCF [12, 13]. The use of two different predispersion values and tunable phase conjugation allowed the residual dispersion to be kept below $\pm 20\mathrm{ps}/\mathrm{nm}$ ($\sim 1\mathrm{dB}$ penalty for $100\mathrm{Gbit}/\mathrm{s}$ DQPSK) over the range of 1540–$1583\mathrm{nm}$. The use of lower-dispersion-slope fiber, or additional predispersion values, could allow for the wavelength operation range, and thus the achievable delay, to be increased. The delay resolution of $\sim 26\mathrm{ps}$ was limited by the $1\mathrm{pm}$ wavelength resolution of our pump lasers. Finer wavelength control has been shown to greatly increase the delay resolution [4].

Since the intrachannel dispersion slope (picoseconds per square nanometer) is not compensated during the phase conjugation process, several different types of dispersive fiber are combined to create a flatter dispersion profile. This slope matching of fibers allowed the residual dispersion slope to be kept below $80\mathrm{ps}/{\mathrm{nm}}^2$, $\sim 1.2\mathrm{dB}$ penalty at $100\mathrm{Gbit}/\mathrm{s}$, in the C and L bands, Fig. 2b. Better matching of dispersive fibers may further flatten the dispersion profile, allowing for higher-rate operation and lower penalties.

Figure 3a shows the received optical duobinary signal after the delay-line interferometer as the first stage pump is varied in $10\mathrm{pm}$ steps for a converted wavelength of $\sim 1575.8\mathrm{nm}$. The expected $\sim 275\mathrm{ps}$ changes ($2\times 0.01\mathrm{nm}\times 13.75\mathrm{ns}/\mathrm{nm}$ at $1575.8\mathrm{nm}$) in delay are easily seen. On the right, the experimental spectra of both wavelength conversion stages for the case of minimum delay are shown. Both stages are phase conjugating; however, only the second stage is required to be conjugating for dispersion compensation. SBS suppression was not required in either stage. A maximum 0.3 and $0.5\mathrm{dB}$ power penalty was measured for each stage due to the lowered conversion efficiency when being operated away from the ZDW of the fiber. HNLF with a lower dispersion slope or a two-pump FWM scheme may provide better conversion efficiency and reduced power penalties.

Bit-error-rate (BER) measurements were used to assess the performance of our delay system. Figure 4a shows the performance of both the in-phase (I) and quadrature-phase (Q) components of the $100\mathrm{Gbit}/\mathrm{s}$ NRZ-DQPSK signal at the minimum, middle, and maximum delay values. Power penalties from 2.3 to $5.4\mathrm{dB}$ are observed across the delay range. In all cases, the I channel had a slightly higher penalty, likely caused by imperfect modulation resulting in a slightly rectangular constellation. Figure 4b shows the performance of $50\mathrm{Gbit}/\mathrm{s}$ NRZ-DPSK under the same conditions. Penalties of 1.6–$3.9\mathrm{dB}$ are observed for DPSK.

The delay system performs best when operated close to the ZDW of the HNLFs, and the wavelength conversion stages are most efficient. The peak residual dispersion occurs at the edges of the delay range, adding an extra $\sim 1\mathrm{dB}$ of penalty. The worst performance occurs at the maximum delay value, where $\sim 68\mathrm{ps}/{\mathrm{nm}}^2$ of residual intrachannel dispersion slope contributes an additional $\sim 1\mathrm{dB}$ penalty. As intrachannel dispersion slope is not compensated by phase conjugation, it may prove to be a limiting factor for achieving larger delays and higher data rates.

This material is based on research sponsored by the Air Force Research Laboratory (AFRL) and by the Defense Advanced Research Projects Agency (DARPA) under agreements FA8650-08-1-7820 and N00014-05-1-0053.

 

Fig. 1 Block diagram. DCF, dispersion compensating fiber; FBG, fiber Bragg grating; BPF, bandpass filter; EDFA, erbium-doped fiber amplifier; RX, receiver; HNLF, highly nonlinear fiber.

Download Full Size | PDF

 

Fig. 2 (a) Measured delay of $1.16\mathrm{\mu s}$. (b) Calculated residual intrachannel dispersion slope after two passes through the dispersive element.

Download Full Size | PDF

 

Fig. 3 (a) Received optical duobinary signal of the I channel showing $\sim 275\mathrm{ps}$ delay changes for $10\mathrm{pm}$ wavelength shifts at a converted wavelength of $\sim 1575.8\mathrm{nm}$. (b) Experimental spectra of the first stage (top) and phase conjugation stage (bottom) at the minimum delay value.

Download Full Size | PDF

 

Fig. 4 Measured bit-error-rate performance of (a) $100\mathrm{Gbit}/\mathrm{s}$ DQPSK and (b) $50\mathrm{Gbit}/\mathrm{s}$ DPSK for the minimum, middle, and maximum delay values.

Download Full Size | PDF

1. A. E. Willner, in The 20th Annual Meeting of the IEEE Lasers and Electro-Optics Society, 2007. LEOS 2007 (IEEE, 2007), paper WFF1.

2. R. S. Tucker, P.-Ch. Ku, and C. J. Chang-Hasnain, J. Lightwave Technol. 23, 4046 (2005). [CrossRef]  

3. G. Raybon and P. Winzer, in Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest (CD) (Optical Society of America, 2008), paper OTuG1.

4. S. R. Nuccio, O. F. Yilmaz, X. Wu, and A. E. Willner, Opt. Lett. 35, 523 (2010). [CrossRef]   [PubMed]  

5. X. Wu, J. Wang, O. F. Yilmaz, S. R. Nuccio, A. Bogoni, and A. E. Willner, Opt. Lett. 34, 3926 (2010). [CrossRef]  

6. Y. Dai, X. Chen, Y. Okawachi, A. C. Turner-Foster, M. A. Foster, M. Lipson, A. L. Gaeta, and C. Xu, Opt. Express 17, 7004 (2009). [CrossRef]   [PubMed]  

7. E. Myslivets, N. Alic, S. Moro, B. P. P. Kuo, R. M. Jopson, C. J. McKinstrie, M. Karlsson, and S. Radic, Opt. Express 17, 11958 (2009). [CrossRef]   [PubMed]  

8. N. Alic, E. Myslivets, S. Moro, B. P. P. Kuo, R. M. Jopson, C. J. McKinstrie, and S. Radic, in Optical Fiber Communication Conference, OSA Technical Digest (CD) (Optical Society of America, 2009), paper PDPA1.

9. Y. Dai, Y. Okawachi, A. C. Turner-Foster, M. Lipson, A. L. Gaeta, and C. Xu, Opt. Express 18, 333 (2010). [CrossRef]   [PubMed]  

10. L. C. Christen, I. Fazal, O. Yilmaz, X. Wu, S. R. Nuccio, A. E. Willner, C. Langrock, and M. Fejer, in Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest (CD) (Optical Society of America, 2008), paper OTuD1.

11. S. R. Nuccio, O. F. Yilmaz, X. Wang, J. Wang, X. Wu, and A. E. Willner, in Conference on Lasers and Electro-Optics/ International Quantum Electronics Conference, OSA Technical Digest (CD) (Optical Society of America, 2010), paper CFJ2. [PubMed]  

12. S. Namiki and T. Kurosu, in European Conference on Optical Communications (2008), paper Th.3.C.3.

13. T. Kurosu and S. Namiki, Opt. Lett. 34, 1441 (2009). [CrossRef]   [PubMed]