PreprintPDF Available

Silicon Photonics DWDM NLFT Soliton Transmitter Implementation and Link Budget Assessment

Authors:
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

We analyze the implementation and capabilities of a silicon photonics transmitter capable of modulating and multiplexing groups of four solitons with different frequency and time spacing based on the nonlinear Fourier transform.
Content may be subject to copyright.
Silicon Photonics DWDM NLFT Soliton Transmitter
Implementation and Link Budget Assessment
A. Moscoso-Mártir1*, J. Koch2, J. Müller1, S. Sharif Azadeh1,3, S. Pachnicke2, J. Witzens1
1 Institute of Integrated Photonics, RWTH Aachen University, Alois-Riedler-Str. 4, 52074 Aachen, Germany
2 Chair of Communications, Faculty of Engineering, Kiel University, Kaiserstraße 2, 24143 Kiel, Germany
3 Now at NINT Department, Max Planck Institute of Microstructure Physics, 06120 Halle, Germany
e-mail: amartir@iph.rwth-aachen.de
ABSTRACT
We analyze the implementation and capabilities of a silicon photonics transmitter capable of modulating and
multiplexing groups of four solitons with different frequency and time spacing based on the nonlinear Fourier
transform.
Keywords: Nonlinear optics, DWDM system, NLFT system, integrated optoelectronic circuits.
1. INTRODUCTION
The optical backbone infrastructure has to support the continuing demand for exponentially growing data
throughput. A cost-effective solution to increase the channel capacity avoiding the deployment of new fibers
consists in increasing the limits of useable power per fiber without incurring nonlinear penalties [1]. In
transmission systems based on the nonlinear Fourier transform (NLFT), the nonlinear Kerr effect simply results
in a change of the phase in the nonlinear spectrum that can be compensated at the receiver (Rx) end [2]. However,
current systems present some limitations in their ability to create higher-order solitons for more efficient data
transmission [3]. These limitations stem from the required computational complexity, high sampling rate and high
bandwidth (BW). To overcome these, it has been proposed to shift the soliton merging from the electrical to the
optical domain using a photonic integrated circuit (PIC) [4]. In this solution, first-order solitons are generated on
individual optical carriers and are multiplexed later using a programmable architecture that allows a precise control
over time and frequency spacing between the individual solitons with minimal spectral distortion [4]. Since only
first-order solitons have to be generated and detected in the electrical domain, BW requirements on the electrical
and optoelectronic components are significantly reduced. Moreover, it is possible to parallelize first-order soliton
generation and detection, further reducing the digital signal processing (DSP) speed requirements.
In this paper, we present the realization of a silicon photonics (SiP) transmitter (Tx) that can modulate four
first-order solitons and multiplex them with adjustable time-delays by means of a programmable delay network
implemented with coupled (ring-)resonator optical waveguide (CROW) optical add-drop multiplexers (OADMs)
[5]. Based on the modeled and measured characteristics of the fabricated PIC, and previous experience with fully
functional wavelength division multiplexed (WDM) transceivers [6], a complete model for a system setup
combining commercially available equipment and the proposed PIC is established and analyzed. A link budget is
derived to obtain realistic power values for the multiplexed soliton pulses at the output of the PIC. This information
is used to upgrade the simulation tool presented in [4] to model the link together with the fabricated chip, so that
the features and programmability of the PIC can be investigated. Using this simulation tool, we have run different
scenarios concerning pulse timing and explored the capabilities of the SiP DWDM NLFT soliton transmitter under
development. Experimental system-level results available by the time of the conference will also be presented.
2. TRANSMITTER IMPLEMENTATION AND LINK BUDGET
The envisioned system as well as a micrograph of the recently delivered system chip fabricated in a multi-
project wafer run at Advanced Micro Foundries (AMF) are shown in Fig. 1 and Fig. 2(a). A comb source from
Pilot Photonics (Lyra-OCS-1000) is used to provide the carriers for the first-order solitons. The comb has an
adjustable free spectral range (FSR) of 10 ± 4 GHz that can be used to adjust the soliton frequency spacing. Each
line has a power of -6 dBm and a linewidth of 80 kHz. Four of these lines are filtered with an external 45 GHz
passband optical filter with 2 dB insertion loss (IL). The selected carriers are amplified to 10 dBm with an EDFA
with a noise figure (NF) of 5 dB, so that the total power injected in the PIC does not exceed 16 dBm to prevent
permanent damage to the grating coupler (GC, IL = 3 dB) used as a coupling device. Inside the chip, four 2nd-order
CROW OADMs (IL = 1.6 dB, BW = 6.5 GHz) route carriers to one of the two input ports of two IQ Mach-Zehnder
modulators (MZM), as shown in Fig. 2(a). The designed IQ-MZMs with slotted transmission lines and inter-phase-
shifter cross-talk suppression [7,8] have 4.4 mm long phase shifters, a BW of 14 GHz optimized for the present
application, a VL of 2.45 V∙cm and 4.5 dB IL. Operated with 1 Vpp signals and biased to achieve full extinction
during soliton shaping, they introduce an insertion loss and modulation penalty of 13.5 dB, defined as the
attenuation of the peak power after modulation. Afterwards, channels 1 and 2, that respectively carry the same
information as channels 3 and 4, are delayed (IL = 3 dB) as a means to emulate four independent channels. All
channels are multiplexed onto one out of two bus waveguides using four 4th-order CROW OADMs (IL = 2 dB,
BW = 17.5 GHz) that are tuned to achieve equal peak power for all channels.
Figure 1. Block diagram of investigated link.
Figure 2. (a) PIC block diagram and chip micrograph (inset). (b,c) Soliton pulse sequencing within the transmission window.
A multimode interferometer coupler (MMI) finally serves to combine the two bus waveguides into one (IL =
3 dB), that is then routed off chip via a second GC. Merging via an MMI allows for pulses of even and odd channels
to have partial spectral overlap without clipping of the spectra [4]. Including interconnection and monitor tap losses
(1.5 dB), we obtain a soliton peak power of -20.6 dBm per channel. These power levels have to be reamplified
with an EDFA to reach launch powers that fulfill the soliton condition, depending on soliton pulse width and the
length of the used recirculating, amplified non-zero dispersion shifted fiber (NZDSF) loops [1].
In the implemented transmitter [Fig. 2(a)], neighboring channels (single increment in channel number also
corresponds to closest spectral and time overlap inside the transmission window (TW), since λ1 < λ2 < λ3 < λ4 and
t1 < t2 < t3 < t4) are sent to different IQ-MZMs to ensure fully independent data streams. This also ensures that
channels multiplexed onto the same bus waveguide at the output of an IQ modulator also have sufficiently low
spectral overlap to be routable by the OADMs with minimal cross-talk. The four optical channels are modulated
with only a 4-channel arbitrary waveform generator (AWG). The delay τWG applied to channels 1 and 2 by on-
chip delay loops can be selected to be either 300 ps or 300 ps + 200 ps. Together with the pulse repetition time at
the output of the AWG (duration of a complete symbol, i.e., a 4-pulse TW) as well as the time delay between
independent pulse trains applied to the two IQ-MZMs AWG), this allows reprogramming of the pulse-to-pulse
time delays (Dt). For example, this allows to explore pulse spacings Dt of 250 ps or 150 ps with pulse trains fully
filling 1000 ps or 600 ps long TWs [Fig. 2(b)]. A 3rd configuration also modeled below consists in closer spaced
pulses (Dt = 100 ps) only partially filling a 500 ps long TW also including a 100 ps down time [Fig. 2(c)].
3. SIMULATION SETUP AND RESULTS
For the simulations, blocks of 40,000 bits are QPSK modulated and split up into two channels. Using these
symbols, soliton pulses are created with a characteristic pulse width (T0) of 38 ps, a BW of 8.8 GHz (99% power)
and a peak amplitude of 1 Vpp. These soliton pulses are then applied to the IQ-MZMs to modulate and multiplex
the four channels. The relative carrier frequencies are set to Δf = [-15 GHz, -5 GHz, 5 GHz, 15 GHz]. As mentioned
above, soliton pulses have a peak power of -20.6 dBm at the output of the PIC, which needs to be amplified by an
EDFA (NF = 5 dB) up to -0.3 dBm fulfilling the soliton condition. After amplification, the combined solitons are
launched into N spans of 50 km NZDSF combined with an EDFA (NF = 5 dB), resulting in a 28 dB optical signal-
to-noise ratio (OSNR) at the Rx after 3,750 km (OSNR accounting for 4 soliton pulses and noise in the entire TW).
On the Rx side, the 4-pulse trains are wavelength demultiplexed and sent to four parallel coherent Rx. After
demodulation, each soliton pulse is low-pass filtered, sampled and digitized. Finally, solitons are normalized and
analyzed using the forward-backward method, followed by a blind phase search for phase recovery [4].
We have simulated the system to understand how the time spacing affects soliton integrity and demodulation.
As seen in Fig. 3(a), it is possible to transmit groups of 4 solitons with Dt = 250 ps (8 Gbps) up to 2,000 km with
Figure 3. (a) BER as a function of distance for different Dt. Soliton pulse eye diagrams for (b) Dt = 150 ps at 1,950 km and
(c) Dt = 150 ps at 2,350 km.
a BER below the hard-decision forward error correction (HD FEC) limit and up to 2,750 km below the soft-
decision (SD) FEC limit. Similar results are obtained for a tighter soliton sequencing with Dt =150 ps (13.3 Gbps).
Packing the solitons even closer clearly affects their integrity as can be seen for Dt = 100 ps. This shows that the
four solitons can be packed together in a TW of width 4T0 with small penalty, but that higher compactification is
detrimental. Ripples are seen in the BER curves at 1950 km for Dt = 150 ps and at 3500 km for Dt = 250 ps. In
soliton merging, soliton pulses propagate together with different group velocities, changing their initial position
within their TW. At specific distances they can partially or even completely overlap, as seen in Fig. 3(b). This
makes it more difficult to recover them at the Rx even when compared with longer distances where they are again
equally spaced [Fig. 3(c)], as a consequence of the partial spectral overlap of the soliton pulses [4] making
wavelength demultiplexing imperfect. Changing the frequency spacing or the initial time domain separation as
enabled by the Tx reconfigurability allows shifting these ripples away from a targeted transmission length. Hence,
this problem can be avoided in a real transmission scenario. While data rates that are currently considered are still
low, they will be increased once the system has been experimentally proven by increasing the number of solitons,
using more complex modulation formats and reducing the width T0 of the solitons. With the current chip, we could
generate QPSK modulated solitons up to 20 Gbps. It is also possible to redesign the IQ-MZM to achieve higher
BWs [8] and higher values of Vpp could be used to reduce the modulation penalty. Furthermore, data dependent
time shifts on a small scale can be included to introduce amplitude modulation in the NLFT domain [1].
4. CONCLUSIONS
We have presented a SiP DWDM soliton Tx that has been characterized and is currently being used in
experimental system tests. It can modulate four first-order solitons and multiplex them with different frequency
and time spacing. Nonlinear link simulations have been run for different pulse-to-pulse spacing and fiber lengths.
Simulation results show that solitons can be compactified together in a transmission window 4T0 and still allow to
reach up to 2,500 km with a BER below the SD FEC limit. Furthermore, we have explored the evolution of soliton
merging over distance, providing a solution to soliton overlap by shifting it out of the region of interest via Tx
reconfiguration. There are multiple paths towards increasing the data rates we are presently working with that will
be taken once the current configuration has been experimentally proven.
REFERENCES
[1] S. K. Turitsyn, et al.: Nonlinear Fourier transform for optical data processing and transmission: advances
and perspective, Optica, vol. 4, issue 3, pp. 307-322, 2017.
[2] S. Li, et al.: Optical signal processing in the discrete nonlinear frequency domain, Proc. Opt. Fiber Commun.
Conf. & Expo., San Diego, CA, USA, March 2018, paper W2A.40.
[3] J. Koch, et al.: Transmission of Higher Order Solitons created by Optical Multiplexing, IEEE J. Lightw.
Technol., vol. 37, issue 3, pp. 933-941, Feb. 2019.
[4] J. Koch, et al.: DWDM Soliton Transmission enabled by Integrated Photonics and Nonlinear Fourier
Transform, Proc. 20th ITG-Symposium on Photonic Networks, pp. 52-54, May 2019.
[5] F. Xia et al.: Ultra-compact high order ring resonator filters using submicron silicon photonic wires for on-
chip optical interconnects, Opt. Expr., vol. 15, issue 19, pp. 11934-11941, 2007.
[6] A. Moscoso-Mártir, et al.: 8-channel WDM silicon photonics transceiver with SOA and semiconductor
mode-locked laser, Opt. Expr., vol. 26, pp. 25446-25459, 2018.
[7] F. Merget, et al.: Silicon photonics plasma-modulators with advanced transmission line design, Opt. Expr.,
vol. 21, issue 17, pp. 19593-19607, 2013.
[8] S. S. Azadeh, et al.: Advances in silicon photonics segmented electrode Mach-Zehnder modulators and
peaking enhanced resonant devices, Photonics North, Proc. SPIE 9288, Sep. 2014.
ResearchGate has not been able to resolve any citations for this publication.
Conference Paper
Full-text available
We demonstrate the possibility to densely multiplex solitons in time and frequency domain utilizing integrated silicon photonics and the Nonlinear Fourier Transform to transmit a fourth order soliton over multiple spans using lumped amplification.
Article
Full-text available
We demonstrate an integrated 8 by 14 Gbps dense wavelength division multiplexed silicon photonics transceiver that makes use of an external mode-locked laser as a light source and a single semiconductor optical amplifier for post-modulation signal amplification. Remaining components necessary for modulation, filtering and (de‑)multiplexing are monolithically integrated in a single chip. In all system experiments, all eight channels are jointly operated with independent data streams in order to include impairments arising out of nonlinear effects inside the SOA while benchmarking the system performance. The transmitter, measured with a commercial reference receiver, supports on-off keying data transmission with an uncorrected BER ranging between 1e-5 and 5e-7 for all channels in back-to-back configuration and between 8e-4 and 1e-5 after 10 km transmission (both PRBS 2³¹-1). The three best channels of the full link consisting in the silicon photonics transmitter operated with the silicon photonics receiver in back-to-back configuration maintain a BER better than the targeted 5e-5. Based on link budget modeling, we expect this target to be reached for all 8 channels pending improvement of the receiver offset compensation loop.
Conference Paper
Full-text available
We investigate the possibility of optical signal processing (OSP) for the nonlinear discrete spectrum. OSP is used to generate and separate a fifth-order soliton optically by a novel transmitter and receiver setup with highly-nonlinear fiber.
Article
Full-text available
Fiber-optic communication systems are nowadays facing serious challenges due to the fast growing demand on capacity from various new applications and services. It is now well recognized that nonlinear effects limit the spectral efficiency and transmission reach of modern fiber-optic communications. Nonlinearity compensation is therefore widely believed to be of paramount importance for increasing the capacity of future optical networks. Recently, there has been steadily growing interest in the application of a powerful mathematical tool—the nonlinear Fourier transform (NFT)—in the development of fundamentally novel nonlinearity mitigation tools for fiber-optic channels. It has been recognized that, within this paradigm, the nonlinear crosstalk due to the Kerr effect is effectively absent, and fiber nonlinearity due to the Kerr effect can enter as a constructive element rather than a degrading factor. The novelty and the mathematical complexity of the NFT, the versatility of the proposed system designs, and the lack of a unified vision of an optimal NFT-type communication system, however, constitute significant difficulties for communication researchers. In this paper, we therefore survey the existing approaches in a common framework and review the progress in this area with a focus on practical implementation aspects. First, an overview of existing key algorithms for the efficacious computation of the direct and inverse NFT is given, and the issues of accuracy and numerical complexity are elucidated. We then describe different approaches for the utilization of the NFT in practical transmission schemes. After that we discuss the differences, advantages, and challenges of various recently emerged system designs employing the NFT, as well as the spectral efficiency estimates available up-to-date. With many practical implementation aspects still being open, our mini-review is aimed at helping researchers assess the perspectives, understand the bottlenecks, and envision the development paths in the upcoming NFT-based transmission technologies.
Conference Paper
Full-text available
We report recent progress made in our laboratory on travelling wave Mach-Zehnder Interferometer based Silicon Photonics modulators with segmented transmission lines, as well as on resonant ring modulators and add-drop multiplexers with peaking enhanced bandwidth extended beyond the photon lifetime limit. In our segmented transmission lines, microstructuring of the electrodes results in radio-frequency modes significantly deviating from the transverse electromagnetic (TEM) condition and allows for additional design freedom to jointly achieve phase matching, impedance matching and minimizing resistive losses. This technique was found to be particularly useful to achieve the aforementioned objectives in simple back-end processes with one or two metallization layers. Peaking results from intrinsic time dynamics in ring resonator based modulators and add-drop multiplexers and allows extending the bandwidth of the devices beyond the limit predicted from the photon lifetime. Simple closed form expressions allow incorporating peaking into system level modeling.
Article
Full-text available
We have investigated two novel concepts for the design of transmission lines in travelling wave Mach-Zehnder interferometer based Silicon Photonics depletion modulators overcoming the analog bandwidth limitations arising from cross-talk between signal lines in push-pull modulators and reducing the linear losses of the transmission lines. We experimentally validate the concepts and demonstrate an E/O −3 dBe bandwidth of 16 GHz with a 4V drive voltage (in dual drive configuration) and 8.8 dB on-chip insertion losses. Significant bandwidth improvements result from suppression of cross-talk. An additional bandwidth enhancement of ~11% results from a reduction of resistive transmission line losses. Frequency dependent loss models for loaded transmission lines and E/O bandwidth modeling are fully verified.
Article
The nonlinear Fourier transform (NFT) is a promising tool to linearize the inherently nonlinear optical fiber channel. The NFT transforms a time domain signal into the continuous and the discrete spectrum. The discrete spectrum is composed of an arbitrary number of complex valued discrete eigenvalues and their associated amplitudes. These discrete eigenvalues relate to solitons, which maintain their shape or return to it in an oscillating manner, while passing through the optical channel. Higher order solitons consisting of multiple eigenvalues are complex pulses, which are created and demodulated by sophisticated digital signal processing (DSP) leading to demanding hardware requirements. This paper shows a way to work with higher order solitons in a WDM like fashion by using opticalelectrical signal processing and presents boundaries of this method. Optical-electrical signal processing decreases the required electrical and electro-optical hardware specifications substantially and enables to use a simplified DSP. The proposed creation method is subsequently employed to transmit higher order solitons consisting of five QPSK modulated eigenvalues. Furthermore, the optical-electrical processing is benchmarked against the Darboux transformation, which creates higher order solitons purely numerically. The results show that for a 5th order soliton transmission the proposed method can significantly reduce the hardware requirements and DSP complexity.
Article
Ultra-compact 5(th) order ring resonator optical filters based on submicron silicon photonic wires are demonstrated. Out-of-band rejection ratio of 40dB, 1dB flat-top pass band of 310GHz with ripples smaller than 0.4dB, and insertion loss of only (1.8+/-0.5)dB at the center of the pass band are realized simultaneously, all within a footprint of 0.0007mm(2) on a silicon chip.