Open Access
Add to BibSonomyAbstract
A dual-quadrature coherent receiver based on a polymer planar lightwave circuit (PLC) is presented. This receiver comprises two separate optical 90°-hybrid chips made of polymer waveguides and hybridly integrated with InGaAs/InP photodiode (PD) arrays. The packaged receiver was successfully operated in 112 Gbit/s dual-polarization quadrature phase-shift keying (QPSK) transmission experiments. In back-to-back configuration the OSNR requirement for a BER value of 10−3 was 15.1 dB which has to be compared to a theoretical limit of 13.8 dB.
©2011 Optical Society of America
1. Introduction
The introduction of higher-order modulation formats in optical communication systems, especially in combination with coherent detection schemes, has increased the complexity of both transmitter and receiver assemblies. A key issue for the successful deployment of such advanced modulation formats is the availability of cost-efficient and compact integrated optoelectronic receivers. While the delay-line interferometer is typically used with direct-detection schemes [1], the 90° optical hybrid represents the key component in coherent receivers for demodulating phase-encoded signals [2]. An integrated dual polarization (DP) coherent receiver for quadrature phase-shift keying (QPSK) will be comprised of the following basic parts as shown in Fig. 1 : eight photodetectors (PDs), four transimpedance amplifiers (TIAs), two 90° hybrid mixers, and polarization beam splitting (PBS) elements. The size and requirements of these components are specified in [3].
Fig. 1 Schematic layout of the DP-QPSK-receiver concept operated in single ended configuration. SIG: signal; LO: local oscillator; PBS: polarization beam splitter; BS: beam splitter; PD: photodetector; TIA: transimpedance amplifier. XI, XQ, YI and YQ are the respective I-/Q- channels of the X-/Y-polarization channels.
Several integrated receivers using different material platforms have been published recently; e.g. a polarization-diverse dual-quadrature coherent receiver based on InP-monolithic integration and free space optics [4]; a monolithic InP multi-wavelength polarization-diversity coherent receiver [5]; a polarization-diversity coherent receiver photonic integrated circuit on Si [6]; and a polarization-diversity coherent receiver module using a silica-based hybrid integration platform [7]. While monolithic solutions may face high fabrication costs and yield issues, the hybrid integration solutions may benefit from using optimized individual components and hence superior overall performance. A prerequisite is however that photodetectors can be easily integrated with minimal insertion loss.
Among the various available material platforms polymer technology has recently gained interest due to aspects regarding facile processing [8], low-cost packaging [9] including passive fiber attachment [10], and the implementation of passive optical functions by applying thin-film elements [10]. Polymer planar lightwave circuits (PLCs) accommodate space consuming waveguide structures such as 90°-hybrids and delay-line interferometers at an affordable price per square unit, while active III-V components can be easily integrated via 45° turning mirrors [10,11].
This paper describes the implementation of polymer coherent receiver units and presents first results on a fully functional DP-QPSK receiver employing such subassemblies. These devices have been successfully tested in 112 Gbit/s DP-QPSK experiments, believed to be the first demonstration of polymer chip based coherent receivers until to date.
The paper is organized as follows: In Section 2 we cover the integration concept and outline the performance of the key building blocks i.e. optical hybrids and photodiodes. The characterization of the receiver assembly will be given in Section 3. In Section 4 the results of system measurements involving the receiver assembly are shown and evaluated. Finally, Section 5 concludes this work and briefly addresses directions of future work.
2. Integration concept and characterization of building blocks
The coherent receiver presented in this work is based on hybrid integration technology. It comprises two 90°-hybrids made of polymer waveguides and two PD-arrays, each consisting of four individual mesa-type InGaAs/InP p-i-n diodes. These two components are hybrid-integrated and co-packaged with four differential linear transimpedance amplifiers (Fig. 1). For polarization diversity operation the receiver can be extended by adding a polarization beam splitter (PBS) and a power beam splitter (BS).
On our polymer based hybrid integration platform, PDs are mounted on top of the waveguides using integrated 45° turning mirrors formed in the waveguide, as sketched in Fig. 2(a) , by a simple sawing process. Subsequently a thin Au metal layer is deposited onto the 45° slope of the mirror structure. This vertical coupling scheme offers the possibility of semi-automated surface-mount assembly by means of a fine placer machine. The typical mirror loss is estimated to be around 1 dB. Special index matching glue is applied to reduce back-reflection and also to fix the PD in place.
Fig. 2 (a) The integration unit comprising fiber-chip coupling, polymer PLC waveguides and vertical photodetectors (PD), and a micrograph of an integrated PD-array on polymer PLC (top-view). (b) Frequency response of a single PD.
The PD-array has been appropriately adapted to this hybrid integration technology. The PD-array comprises four vertical backside illuminated PDs with an active area of 22 µm diameter each and substrate thickness of 120 µm after thinning. The bottom-illuminated PDs are designed with a full p-metal contact on its top side facilitating uniform current flow and concurrently serving as optical reflector. Thus, the incoming light double-passes the absorbing layer. The typical PD responsivity amounts to ~0.6 A/W at 1550 nm, while the 3 dB bandwidth reaches 24 ~27 GHz (Fig. 2(b). Associated with each PD are bias circuits that are monolithically integrated on the array chip.
Two types of optical 90° hybrids incorporated on the polymer PLC were tested: a 2x4 multimode interference (MMI) splitter, Fig. 3(a) , and an arrangement of four 2x2 MMIs with active phase adjustment accomplished by thermo-optical phase shifters, Fig. 4(a) . Although purely passive operation of the 90° hybrid structures will be preferred, for initial assessment of the influence of the phase error on the performance such active phase adjustment provides a versatile and viable option. Both versions of the hybrids where investigated for comparison. The 90° hybrids have a size of 1.3 mm x 10.5 mm thus fitting into the existing packages compliant with the multi source agreement (MSA) for compact coherent receivers (CCRx), which have dimensions of 40 mm × 27 mm × 6.6 mm. Prior to assembly these components have been characterized in a configuration comprising a delay-line interferometer to supply the homodyne inputs of the 90° hybrid. A wavelength scan allows evaluation of the filter functions of these hybrids, Fig. 3(b) and Fig. 4(b). From these filter characteristics the relative phase relations between the outputs of the hybrids can be extracted, as shown in Fig. 3(c) and Fig. 4(c). For the former hybrid employing a 2x4 MMI, phase deviations of ~12° and excess loss below 4.5 dB were obtained, as illustrated in Fig. 3. In Fig. 4 it is demonstrated that for the 2x2 MMI based hybrid phase errors below 3° and an excess loss of less than 4.5 dB were achievable over the entire C-Band. Furthermore, this approach shows a minimum extinction ratio of 17 dB among the outputs, which appears to be slightly worse than that of the 2x4 MMI hybrid. On the other hand, the peaks of its filter functions are much better balanced (imbalance below 0.5 dB) than with the latter design.
Fig. 3 (a) 90° hybrid based on a 2x4 MMI. (b) and (c) are the filter characteristic of all four output channels and the relative phase in-between, respectively.
Fig. 4 (a) 90° hybrid comprising four 2x2 MMIs. (b) and (c) are the filter characteristic of all four output channels and the relative phase difference in-between.
3. Receiver assembly and characterization
Two receivers were assembled in packages and their performances are investigated. These receiver modules contained two 90° hybrids for the two polarization states using 2x4 MMI, and two 90° hybrids using four 2x2 MMIs including a phase adjustment, respectively. In either case PD arrays were integrated as outlined above.
Figure 5(a) shows the overall responsivity for the 2x4 MMI based coherent receiver. The effective responsivities of the X-/ Y-channels were around 0.05 A/W. Comparing the results in Fig. 5(a) to the responsivity of the individual PD of 0.6 A/W, we determine an insertion loss of 10.7 dB. This loss can be attributed to the 6 dB intrinsic loss of the 90° hybrid and an excess loss of 4.7 dB, which includes fiber-to-PLC and PLC-to-PD coupling imperfections. RF-bandwidth measurement results given in Fig. 6(a) show a bandwidth of 24±1 GHz for all assembly output channels. The relative phases between the outputs were also measured. The maximum phase error of 13.5°±1.5° for both X-/Y- channel can be found from the lines with the legends “X-branch of 2x4 MMI” and “Y-branch of 2x4 MMI” in Fig. 6(b). Although the values exceed the target value of ≤ 7° and the Rx with 2x4 MMI does not comprise active phase adjustment element, the even evolution of the phase relation over the entire C-Band indicates usability of such hybrids once the correct design parameters are met.
Fig. 5 Assembly responsivity of the Rx channels vs. wavelength, with the Rx modules comprising two hybrids using (a) 2x4 MMIs and (b) four 2x2 MMIs, respectively.
Fig. 6 (a) Small-signal frequency responses Sij (Index ij: cf. Figure 1). (b) Phase errors of X and Y channels for Rx using 2x4 MMIs are given by the lines “X-/Y-branch of 2x4 MMI”, and the lines “X-/Y-branch of four 2x2 MMI” for Rx using 2x2 MMIs with phase adjustment; where the dashed area indicates the tolerance range with phase adjustment.
The second receiver module with four 2x2 MMIs shows similar properties, such as assembly responsivity given in Fig. 5(b), but with substantially lower phase errors of below ±3° (the lines with the legends “X-branch of four 2x2 MMI” and “Y-branch of four 2x2 MMI” in Fig. 6(c)) due to application of the active phase adjustment via thermal heaters. Typical adjustment currents needed for optimized phase relation is ~10 mA. Due to the small thermal conductivity of polymers [8], such a phase adjustment scheme will also be feasible for monolithic integration of more than one hybrid, e.g. for dual-polarization configurations.
4. System measurements
Both Rx module variants were tested in a non-return-to-zero (NRZ) DP-QPSK system as shown in Fig. 7 . The DP-QPSK transmitter consisted of a 56 Gbit/s bit error rate (BER) transmitter (SHF 12100B), an external-cavity laser (Agilent 86118A), and an IQ-modulator (Fujitsu FTM7960EX) for DQPSK transmission format. The IQ-modulator is made on Ti:LiNbO3 and consists of two dual-drive push–pull Mach–Zehnder modulators incorporated in a Mach–Zehnder superstructure. The symbol rate of the modulated QPSK signal was 28 Gbaud. The word length of the pseudo-random-binary sequence (PRBS) was 27-1, de-correlated between both polarizations by 95 bit. The transmission experiment was performed with and without a 100 km long single mode fiber (SMF). A polarization controller (PC) was inserted in front of the PBS at the receiver. By adjusting the PC and thus the state of polarization of the QPSK signal, the X and Y polarization-channels of the receiver assembly can be investigated separately. The optical power of the local oscillator is launched into the receiver with the QPSK signal. The results obtained with the receiver assembly involving the four 2x2 MMI design are as follows:
Figure 8 shows the BER results of the system measurements after offline processing (LeCroy SDA 8 Zi Serial Data Analyzers). Figure 8(a) shows the BER curve of the X- and Y-branch of the receiver operated solely by suitable polarization adjustment, and the BER of the dual polarization (XY-) operation both in back to back configuration and after transmission over 100 km SMF without applying optical dispersion compensation.
Fig. 8 (a) BER vs. OSNR, (b) corresponding X-/Y-constellation diagrams for BER of 10−3 and for error free operation (bottom), (c) BER versus launched signal power in back-to-back configuration and after transmission over 100 km of SMF without dispersion compensation.
First, the OSNR sensitivity was investigated in the back-to-back experiments. OSNR at BER of 10−3 was determined to be 15.1 dB, which compares with a theoretically predicted minimum of 13.8 dB [12]. The penalty of 1.3 dB can mainly be attributed to the limited bandwidth (~0.57 Hz/Symbol) and the low oversampling rate (1.43 Samples/Symbol). A discussion of the optimum oversampling rate and the offline signal processing can be found in Ref [13]. Small additional penalty contributions may be caused by tolerances in the demonstrator setup. For transmission over 100 km SMF without applying optical dispersion compensation, this OSNR penalty increases to ~2.1 dB only.
Furthermore, the dynamic range of the input power of the DP-QPSK signal to the receiver was investigated. In Fig. 8(c) the OSNR value was set to 17 dB in order to reach a mid-range BER value of 10−4. Then the dynamic range was measured versus the input power of the DP-QPSK signal at a fixed optical power of 15.5 dBm (at the beam splitter input) from the local oscillator. In the back-to-back configuration the performance for a signal power below −15 dBm is limited by the noise of the receiver electronics. For input powers above 0 dBm the saturation of the TIAs hampers the coherent detection. These values were compared with those obtained after transmission over the fiber link of 100 km without dispersion compensation. In the latter case, the input signal power level for BER of 10−4 is limited to between −15 and −5 dBm. The decrease of the dynamic range in the presence of the fiber link is regarded to be due to the fact that the additional recovery of chromatic dispersion requires operation within the linear range of the amplifier.
5. Conclusions
A coherent receiver has been developed for DP-QPSK transmission that involves polymer planar lightwave circuit based 90° hybrid chips integrated with surface-mount photodiode arrays. These chips were mounted into a package that is compliant with existing packaging standards (CCRx MSA). The receiver assembly was successfully tested in a DP-QPSK system in both back-to-back configuration and with transmission over 100 km of SMF without using optical dispersion compensation. The OSNR requirement for a BER of 10−3 was 15.1 dB, which suggests a penalty of 1.3 dB only by comparison to the theoretical limit.
To extend the integration the hitherto separate chips assigned to the two polarization states can be readily merged into one receiver chip to realize a polarization diversity design. Furthermore, the required polarization beam splitter (PBS) function can also be implemented on such a monolithic chip using a thin-film element approach [10] to yield a very compact and virtually temperature independent polarization handling structure. Respective developments are underway.
Acknowledgments
This work has been conducted in the framework of the 100x100 Optics project partly funded by the Future Fund of the Land Berlin co-sponsored by the European Fund for Regional Development (EFRE). The authors would like to thank Ruiyong Zhang for RF characterization of the photodiodes.
References and links
1. A. Gnauck and P. Winzer, “Optical phase-shift-keyed transmission,” J. Lightwave Technol. 23(1), 115–130 (2005). [CrossRef]
2. M. Seimetz and C.-M. Weinert, “Options, feasibility, and availability of 2×4 90° hybrids for coherent optical systems,” J. Lightwave Technol. 24(3), 1317–1322 (2006). [CrossRef]
3. OIF, “Implementation agreement for integrated dual polarization intradyne coherent receivers,” 2010. http://www.oiforum.com/public/documents/OIF_DPC_RX-01.0.pdf.
4. A. Beling, N. Ebel, A. Matiss, G. Unterbörsch, M. Nölle, J. K. Fischer, J. Hilt, L. Molle, C. Schubert, F. Verluise, and L. Fulop, “Fully-Integrated Polarization-Diversity Coherent Receiver Module for 100G DP-QPSK,” in Proc. OFC’11 (Los Angeles, CA, USA, 2011), OML5.
5. C. R. Doerr, L. Zhang, and P. J. Winzer, “Monolithic InP Multi-Wavelength Coherent Receiver,” in Proc. OFC’10 (San Diego, CA, USA, 2011), PDPB1.
6. C. R. Doerr, P. J. Winzer, S. Chandrasekhar, M. Rasras, M. P. Earnshaw, J. S. Weiner, D. M. Gill, and Y. Chen, “Monolithic Silicon Coherent Receiver,” in Proc. OFC’09 (San Diego, CA, USA, 2009), PDPB2.
7. T. Ohyama, I. Ogawa, H. Tanobe, R. Kasahara, S. Tsunashima, T. Yoshimatsu, H. Fukuyama, T. Itoh, Y. Sakamaki, Y. Muramoto, H. Kawakami, M. Ishikawa, S. Mino, and K. Murata, “All-in-one 100-Gbit/s DP-QPSK coherent receiver using novel PLC-based integration structure with low-loss and wide-tolerance multi-channel optical coupling,” in Proc. 15th OECC (Sapporo, Japan, 2010), PD6.
8. H. Ma, A. K.-Y. Jen, and L. R. Dalton, “Polymer-based optical waveguides: materials, processing, and devices,” Adv. Mater. (Deerfield Beach Fla.) 14(19), 1339–1365 (2002). [CrossRef]
9. G. Yu, J. Mallari, H. Shen, E. Miller, C. Wei, V. Shofman, D. Jin, B. Chen, H. Chen, and R. Dinu, “40GHz zero chirp single-ended EO polymer modulators with low half-wave voltage,” in Proc. CLEO 2011 (Baltimore, MD, USA, 2011).
10. N. Keil, C. Zawadzki, Z. Zhang, J. Wang, N. Mettbach, N. Grote, and M. Schell, “Polymer PLC as an Optical Integration Bench,” in Proc. OFC’11 (Los Angeles, CA, USA, 2011), paper OWM1.
11. J. Wang, C. Zawadzki, N. Mettbach, W. Brinker, Z. Zhang, D. Schmidt, N. Keil, N. Grote, and M. Schell, “Polarization insensitive 25-Gbaud direct D(Q)PSK receiver based on polymer planar lightwave hybrid integration platform,” Opt. Express 19(13), 12197–12207 (2011). [CrossRef] [PubMed]
12. F. Xiong, Digital Modulation Techniques (Artech House, Boston, 2006).
13. J. K. Fischer, R. Ludwig, L. Molle, C. Schmidt-Langhorst, C. C. Leonhardt, A. Matiss, and C. Schubert, “High-speed digital coherent receiver based on parallel optical sampling,” J. Lightwave Technol. 29(4), 378–385 (2011). [CrossRef]
