Search

The design and characterization of a new photonic chip on TriPleX $^{TM}$ integrated platform technology for investigating all-optical Haar transform (HT) based compression in the visible spectrum is proposed. A building block (BB) based on a multimode interference (MMI) structure is designed for green light wavelength (532 nm), enabling the optical HT operation in the circuits of the optical chip. The designed TriPleX chip demonstrates two types of MMI based optical circuits, i.e., restricted interference (RI) and general interference (GI), allowing the implementation of the optical HT and compression on the input image pixels. An optical image compression rate of 75% is achieved for a linear 8-pixels array of an image by the designed circuit in the chip. An experimental coupling ratio (CR) of ~98:2 is obtained by the designed circuit in the chip with identical optical inputs, proving high capability of the proposed design for the realization of optical HT and image compression.

We investigate the flexibility of a LUT-based hierarchical DM for Probabilistic Constellation Shaping providing flexible net bit rates. Using a fixed parallelized input and output LUT-based architecture, including a flexible strategy to organize the input bits in each LUT, the proposed scheme shows an efficient performance in terms of rate loss vs. complexity, resulting in a DM solution to enable flexible bit rates with improved channel transmission. We use an 800G example design with soft-decision FEC to discuss the optimization strategy and to demonstrate the flexibility of the proposed method, enabling a high granularity bit rate operation in the range of 160 to 800Gb/s with a rate loss always below 0.14bits/symbol.

Some of the different aspects of the DSP enabled receivers in access networks will be addressed in this article. A revision of the main topics which may impact DSP design will be discussed through application examples. Modulation formats, KK receivers, and coherent receivers are analyzed at light of the DSP enabled receivers’ impact. Also, relevant are FEC and burst mode which are debated at the light of PON requirements. Simplified analog coherent receivers are presented as a use case to serve as simplification path for non-DSP high performance receivers. On the other end, aiming at discussing current PON limits, an ultra dense wavelength division multiplexing (UDWDM) use case is also presented where several advanced topics like pulse shaping, high density, and simplified DSP operations are considered.

We experimentally evaluate the performance of Kramers–Kronig (KK) receivers for free-space optical transmission. We transmit a dual-carrier signal, each carrier modulated as a 25 Gbaud 16/32 quadrature-amplitude modulation (QAM) signal. This dual-carrier signal, with overall symbol rate of 50 Gbaud, is detected in a single receiver with about 35 GHz bandwidth. Key parameters such as the carrier-to-signal power ratio, the frequency allocation of the pilot tone, and the frequency separation between both modulated carriers are optimized. In addition, we also study the effect of the receiver bandwidth limitations on the KK technique. We perform long-term measurements, beyond one hour, for both 16 and 32 QAM signals in a 55 m outdoor free-space link during daylight with sunny weather conditions. Our measurements demonstrate that the link operates below the BER threshold assuming 20% forward-error correction (FEC) for the 32QAM signal and 7% FEC overhead for the 16QAM signal. The resulting net data rates are 208 and 184 Gbit/s for the 32 and 16 QAM signals, respectively.

The exponential traffic growth, demand for high speed wireless data communications, as well as incessant deployment of innovative wireless technologies, services, and applications, have put considerable pressure on the mobile network operators (MNOs). Consequently, cellular access network performance in terms of capacity, quality of service, and network coverage needs further considerations. In order to address the challenges, MNOs, as well as equipment vendors, have given significant attention to the small-cell schemes based on cloud radio access network (C-RAN). This is due to its beneficial features in terms of performance optimization, cost-effectiveness, easier infrastructure deployment, and network management. Nevertheless, the C-RAN architecture imposes stringent requirements on the fronthaul link for seamless connectivity. Digital radio over fiber-based common public radio interface (CPRI) is the fundamental means of distributing baseband samples in the C-RAN fronthaul. However, optical links which are based on CPRI have bandwidth and flexibility limitations. Therefore, these limitations might constrain or make them impractical for the next generation mobile systems which are envisaged not only to support carrier aggregation and multi-band but also envisioned to integrate technologies like millimeter-wave (mm-wave) and massive multiple-input multiple-output antennas into the base stations. In this paper, we present comprehensive tutorial on technologies, requirements, architectures, challenges, and proffer potential solutions on means of achieving an efficient C-RAN optical fronthaul for the next-generation network such as the fifth generation network and beyond. A number of viable fronthauling technologies such as mm-wave and wireless fidelity are considered and this paper mainly focuses on optical technologies such as optical fiber and free-space optical. We also present feasible means of reducing the system complexity, cost, bandwidth requirement, and latency in the fronthaul. Furthermore, means of achieving the goal of green communication networks through reduction in the power consumption by the system are considered.

We report a real-time experimental demonstration of a long-reach coherent ultra-dense wavelength-division multiplexing (WDM)-passive optical network (PON) system with quadratic-amplitude modulation (QAM)-based dual-polarization optical transceivers. Using field-programmable gate array-based digital signal processing, the transmission of 20-WDM channels in 2.5-GHz grid, employing both 2.5-Gb/s dual-polarization quadrature phase-shift keying (DP-QPSK) and 3.75-Gb/s DP-8PSK signals, is successfully demonstrated with bit-error rate in real-time. The performance is evaluated over 100-km standard single mode fiber in terms of receiver sensitivity and ODN power budget for both modulation formats, showing the feasibility of coherent PON scenarios using flexible dual-polarization strategies supported by software-defined transceivers.

We propose low complexity, adaptive multivariate precomputed B-spline and Barycentric Lagrange interpolation-based lookup table ( $ {B}^{2}$ LUT) models for the estimation of channel capacity of multiple-input-multiple-output free-space optical communication systems over generalized Malaga $( {\mathcal{ M}})$ -distributed atmospheric turbulence channels. The models require low memory for efficient operation and exploit precomputed weights to achieve fast and stable operation. The models can significantly reduce the size of lookup table, thereby reducing the computational cost. Similarly, the models can enhance the speed by up to 15 orders of magnitude and beyond without loss of fidelity compared with the current analytical models.

We report an optimized hardware scheme for carrier frequency and phase recovery algorithms based on Mth-order power schemes for flexible intradyne coherent receivers. The phase recovery is based on the traditional Viterbi & Viterbi algorithm with a feedforward architecture, which requires an Mth-order power nonlinear operation. The output of such operation is also used in the differential phase-based approach for frequency offset estimation by means of a feedback loop. Hence, by using a single Mth-order operation, the reported architecture saves approximately half of the hardware resources for quadrature phase shift keying (QPSK) in the frequency recovery stage compared with fully independent phase and frequency algorithms. The evaluation of the accuracy of the proposed method is performed with digital signal processing in real-time using a field-programmable gate array (FPGA) for differential QPSK and 8PSK modulation formats. No significant performance penalty is observed when compared with an implementation featuring independent algorithms.

We experimentally assess the performance versus complexity tradeoff of advanced split-step and Volterra-based digital backpropagation (DBP) techniques, applied to a WDM (75-GHz flexigrid) ultra-long-haul (ULH) transmission system composed of five dual-carrier PM-16 quadratic-amplitude modulation 400 G superchannels. Using the recently proposed weighted Volterra series nonlinear equalizer (W-VSNE), we demonstrate a maximum reach improvement of 600 km, obtained at the expense of only six DBP steps for an entire $>$ 5000-km optical link (1000 km/step), representing a $\sim$ 90 $\%$ reduction on the total number of DBP steps relatively to the widely used split-step Fourier method (SSFM) implementation. The W-VSNE technique is also shown to be more accurate than the correspondent weighted SSFM (W-SSFM) algorithm when applied in medium (10–50 DBP steps) and low complexity ( $\%$ ), together with the associated latency savings (up to 90 $\%$ ), demonstrates the feasibility of low complexity and efficient DBP for high-speed ULH transmission systems.

We experimentally characterize a bidirectional 2.5 Gb/s UDWDM-PON system based on Nyquist shaped DQPSK with digital signal processing in real time. The optical distribution network power budget of this system is evaluated and in coexistence of UDWDM channels with upstream TWDM-PON and RF video overlay signals, a set of optimum parameters and their impact on the network operation is driven. Additionally, we report the first field-trial of bidirectional coherent Nyquist UDWDM-PON using commercial real-time FPGA-based transceivers, coexisting with the deployed GPON, RF video overlay, and NG-PON2 technologies. A −44.5 dBm receiver sensitivity is achieved for 64 × 2.5 Gb/s DQPSK downstream channels and a 17 dB tolerance in dynamic power range of upstream channels is performed.

Nyquist shaping UDWDM-PON combined with digital coherent optical systems has been recently proposed as a promising technology for future optical access networks. In this paper, we report real-time transmitter and receiver for a future coherent PON with high spectral efficiency and low-cost DSP implementation. Using field-programmable gate array (FPGA)-based 8-bit DSP, we experimentally demonstrate a high performance system based on 16 × 2.5 Gb/s QPSK signals in terms of receiver sensitivity and fiber transmission. An optimal transmitted power of –7 dBm per channel is achieved for 50-km standard single-mode fiber in a 2.5-GHz grid. The experimental results show the feasibility of coherent PON scenarios based on DSP supported by FPGAs.

We experimentally analyze an adaptive polarization-demultiplexing (PolDemux) technique based on the representation of the state of polarization of the signal in the Stokes space. The performance and convergence speed of the Stokes PolDemux are compared with the constant modulus algorithm, for dual polarization QPSK and 16QAM modulation formats in a system with a maximum polarization rotation of 6 kHz. Moreover, the applicability of the adaptive Stokes algorithm is tested in an ultradense wavelength-division multiplexing scenario over 80 km of standard single-mode fiber. The BER results show a very small power penalty confirming the suitability of the adaptive Stokes method in systems where the accumulated chromatic and polarization-mode dispersion are small, such as high capacity optical metro and access networks scenarios.

We investigate the power budget of a fully coherent reflective passive optical network, based on self-coherent upstream (US) and self-homodyne downstream (DS) detection schemes. We consider, for US, both binary phase-shift keyed (BPSK) and quadrature phase-shift keyed (QPSK) at 1 GBd generated by direct modulation of a low-cost reflective semiconductor optical amplifier having 1-GHz electrical bandwidth and using simplified digital signal processing (DSP). We find that the penalty of increasing the order of US modulation format from BPSK to QPSK for higher capacity can be reduced to 1 dB by means of DS digital spectral shifting and US static postequalization.

We analyze the performance and the complexity of a real-time digital signal processing (DSP)-based optical network unit architecture for the digital detection of Nyquist-shaped signals in ultra-dense wavelength-division multiplexing passive optical network (UDWDM-PON) scenarios. Using field-programmable gate array-based 8-bit DSP, we experimentally demonstrate in a real-time mode operation the receiver performance in terms of Nyquist filter roll-off factor for $8\times 2.5$ Gb/s quadrature phase-shift keying (QPSK) with 2.5 GHz of channel spacing. We analyze the performance of the linear and cubic interpolation for the clock recovery as well as the Gardner algorithm for the timing-error estimation, considering the amplitude and power formulations. The experimental evaluation shows that the linear interpolation along with the power formulation is an attractive solution for the QPSK-based UDWDM-PON, considering the tradeoff between the performance and the complexity.

We demonstrate fully blind processing and reduced-complexity nonlinear equalization (NLE) of a 100G PM-64QAM optical channel in a 50 GHz WDM grid, achieving a maximum reach of 1524 km over pure silica core fiber at a bit error rate of $2.7\times 10^{-2}$ . The equalization of linear polarization-dependent effects is performed by a radius-directed constant modulus algorithm (RD-CMA), enabled by a multiradii training stage, yielding a very small penalty ( $Q$ 2 factor) relatively to data-aided CMA. Applying a simplified Volterra series nonlinear equalizer (simVSNE), operating in the frequency domain, we demonstrate a reach extension of $\sim$ 27% relatively to linear equalization. Due to its lower spatial resolution requirements, the simVSNE technique is shown to provide a more efficient NLE than the well-known back-propagation split-step Fourier method, both in terms of latency and number of complex multiplications per sample. The potential benefit of NLE for fully blind processing of high-order QAM optical signals is demonstrated by an incremental reduction of the RD-CMA penalty to

We demonstrate submicrosecond polarization alignment of nonreturn to zero 10-Gbit/s optical packets using an all-optical polarization attraction effect in a low polarization-mode dispersion-shifted fiber. We experimentally demonstrate the alignment of optical packets having durations of 205 and 409 ns using a counter-propagating four-wave mixing process. To evaluate this technique, consecutive packets were aligned with distinctly different polarization states using a fiber interferometer. We experimentally demonstrate the recovery of optical packets that were lost due to their state of polarization (SOP) misalignment with the receiver. Polarization stability of the proposed system was assessed over a period of 30 min and it was observed that the scheme was able to correct the SOP and equalize the received optical power.

We investigate the role of a reflective semiconductor optical amplifier (RSOA) in a fully coherent wavelength-shared passive optical network architecture, based on self-coherent upstream (US) and self-homodyne downstream (DS) detection schemes. We present a proof of concept experimental study, demonstrating 32-dB loss budget over a 30-km fiber link, operating at 9.4-Gb/s DS and 750-Mb/s US effective rates with 7% forward error correction and 8b/10b US overheads, respectively. This was achieved using simplified digital signal processing and a low-cost, off-the-shelf RSOA with 18-dB gain and 1-GHz modulation bandwidth.

Coherent optical communication has been well established as the technology of choice for long haul and high bit rate communication systems since a decade ago. Recent technology advances and ongoing price erosion further open the window of opportunity for the application of coherent optical transmission technology in other domains. This paper describes in detail the capabilities, design and implementation of a coherent ultra dense WDM technology for optical metro and access networks. Its capabilities enable a number of attractive options, such as variable downstream bit rates from 150 Mbit/s up to 10 Gbit/s per user, embedded OTDR and the coexistence with legacy systems such as GPON, EPON, XGPON or RF-Video in optical distribution networks. Due to its flexibility and capacity, it is also suitable for deployments in metropolitan networks, as well as for mobile front-haul and back-haul applications.

In this letter, we demonstrate a proof of concept fully loaded bidirectional ultrahigh capacity coherent passive optical network (2 × 1008 × 8.3 Gb/s). This was achieved using partial spectrum overlap, Nyquist shaping, digital frequency shifting, self-homodyne detection, and pilot tone remodulation. Upstream burst mode operation is also demonstrated.

In this paper, we numerically and experimentally demonstrate a bidirectional Terabit+ ultradense wavelength-division multiplexing (UDWDM) coherent passive optical network with Nyquist shaped 16-ary quadrature amplitude modulation, offering up to 10 Gb/s service capabilities per user/wavelength in a total spectrum of 12.8 nm over 40 km of standard single-mode fiber. This paper first demonstrates the capability of Nyquist pulse shaping to mitigate crosstalk arising from back-reflections and nonlinear effects in UDWDM networks with coherent transceivers. The latter part of the paper experimentally investigates the bidirectional transmission in terms of receiver sensitivity and nonlinear tolerance under different network transmission capacity conditions, e.g., number of users and bit rate per users.

“Copyright © [2013] IEEE. Reprinted from IEEE Journal of Selected Topics in Quantum Electronics. ISSN: 1932-4553. This material is posted here with permission of the IEEE. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by writing to pubs-permissions@ieee.org. By choosing to view this document, you agree to all provisions of the copyright laws protecting it.” RSOA is considered a strong candidate to play an important role in realizing the next generation WDM PON, based on the wavelength reuse concept. Therefore, accurate and efficient modeling of RSOA is of significant importance. We present a time-domain wideband model for simulation of spatial and temporal distribution of photons and carriers in a bulk RSOA. A trade-off between the accuracy and the computational efficiency of the model is studied. Multi-objective genetic algorithm is utilized for parameter extraction. Experimental validation has been performed for continuous wave input, NRZ and QPSK signaling pulses up to 40 Gb/s of bit rate, in both amplification and remodulation regimes. Saturation, noise, chirp and signal broadening are successfully predicted, while reducing the computational time compared to other wideband models.

We propose and experimentally demonstrate a flexible wavelength division multiplexing/time division multiplexing network architecture for converged metro-access environment. Entire passiveness in the fiber plant is achieved with remote amplification in the signal distribution nodes along the metro ring and in the power splitters of the local access tree. We assist a traditional remote pumping scheme with a distributed pump provided by the optical network units and demonstrate that loss budgets beyond 30 dB can be supported. Data transmission of up to 10 Gb/s is evaluated in different deployment scenarios, reaching from a 78 km long reach rural to a dense 1:128 split/ urban configuration with field installed fibers, including also worst case resilience configurations.

Next-generation passive optical networks (NG-PONs) promise significant bandwidth increase to end users. Research and standardization efforts are underway to enable enhancements from contemporary PONs to NG-PONs. An area that has not been much investigated from the context of NG-PON2 (as defined by the Full Service Access Networking [FSAN] body) is the medium access control. We showcase for the first time the open problem of MAC implementation beyond WDM PONs, in NG-PON2 with multiple-line-rate channels. We propose a solution called OLIMAC, or the Open Lambda Initiative Medium Access Control, as a mechanism to provision bandwidth and guarantee a services framework in flexible wavelength spaced systems. The article describes the implementation of OLIMAC. The OLIMAC solution is shown to work with different protocols showcasing backward compatibility with existing end-user technologies while facilitating voluminous bandwidth to end users. The OLIMAC concept is simulated, and results show the benefit of the proposal as well as how OLIMAC is instructive in the implementation of NGPON2, especially for multirate flexible wavelength-spacing solutions.

This paper addresses transmission aspects on very high aggregate wavelength-division multiplexing based passive optical networks (WDM-PON). The ultradense WDM-PON system evaluated in this paper transports high-order modulation formats such as M-ary phase-shift keying (PSK) and M-ary quadrature amplitude modulation (QAM) in which coherent detection is performed at the optical network unit (ONU), after transmission over 25 km, 60 km, and 100 km of standard single-mode fiber (SSMF). The first part of this work covers the impact on the system's performance of the most relevant fiber nonlinearities such as self-phase modulation (SPM), cross-phase modulation (XPM) and four-wave mixing (FWM), and their interplay between transmission distance and modulation format. Using a Volterra series method allows estimating the error vector magnitude (EVM) of the received constellation related to different fiber nonlinearities. In a 32 × 625 Mbaud system spaced by 3 GHz (0.025 nm), the FWM to XPM ratio (F/X) varied from 25 up to 30 dB for phase-modulated signals (1.25 Gb/s-QPSK and 1.875 Gb/s-8PSK). On the other hand, this ratio ranged from 2.6 to 7.4 dB for amplitude-modulated signals (2.5 Gb/s-16QAM, 3.75 Gb/s-64QAM and 5 Gb/s-256QAM). After applying a frequency-allocation scheme at both transmitter and local oscillator lasers, some of the FWM crosstalk is mitigated by 2.4 to 3.5 dB. This EVM reduction confirms that increasing the transmission distance and the order of the constellation, the system's performance becomes limited by both interchannel FWM and XPM.

This paper presents and experimentally demonstrates a novel method for the estimation of optical signal-to-noise ratio (OSNR) based on the comparison of an asynchronous histogram of the signal under analysis with a reference asynchronous histogram. The latter is acquired from the signal under analysis at a calibration stage. The proposed method allows the use of optical amplification to increase the sensitivity of the optical monitoring system (OMS) by a factor 20 dB, when using an erbium doped fiber preamplifier. In addition, the use of a semiconductor optical preamplifier, initially designed for nonlinear operation at 2.5 Gb/s is used in the OMS to preamplify 40-Gb/s signals, achieving a sensitivity gain of 10 dB. It will be experimentally demonstrated that the proposed method is applicable to 40-Gb/s nonreturn to zero (NRZ) signals arbitrarily degraded by group velocity dispersion (GVD). Furthermore, accurate monitoring of the OSNR of return-to-zero (RZ) signals will also be possible using a simple RZ-to-NRZ converter based on narrow-band optical filtering within the OMS. The proposed method also allows estimating of the GVD-induced OSNR penalty between the signal under analysis and the signal at the calibration stage.

In this work we perform a study about the temperature implications on the optical fibers chromatic dispersion and the implications on the quality of transported information. To solve this constrain we suggest two approaches to dynamically compensate chromatic dispersion, one based on chirped fiber Bragg gratings and other based on dispersion compensation fibers. With the dispersion compensation fibers, we obtain a first and second order tuning coefficients of 3.472 ps nm-1 oC-1 and -2.7 × 10 -3 ps nm-2 oC-1, respectively. With the compensator based on chirped fiber Bragg gratings we obtain dispersion values between -113.2 ps nm-1 and 143.1 ps nm-1, for temperature gradients of 40oC.

This paper presents a study on the influence of nonideal sampling in asynchronous Q-factor monitoring. An overestimation of the Q-factor of signals impaired with optical noise may occur if the averaging effects due to nonideal sampling affect only the noise component of the monitored signal. It is shown that this problem can be rectified using a correction factor, dependent only on the receiver parameters. This allows reasonable Q-factor estimates using sampling systems with aperture times close to the monitored signal bit period. Nonideal sampling may also affect the signal component of the monitored signal. For such cases, a novel and accurate technique is presented to estimate the Q factor using asynchronous histograms, which takes into account changes in the shape of the monitored signal due to averaging effects. Numerical simulations are used to evaluate the proposed technique achieving Q-factor estimates within a 14% error margin, independently of the sampling aperture for the considered range. Distortion affecting the signal due to the transmission system can also be handled by the new technique presented in this paper. It is verified that the impact of intersymbol interference due to chromatic dispersion or electrical filtering in the receiver on the performance of the system is well reflected by the estimated Q factor.

Black-box models (BBMs) for erbium-doped amplification have been demonstrated to be a powerful tool. A new extended BBM including variations of the erbium-doped fiber length is presented for fiber length optimization in amplified systems. The comparisons with experimental results show a very high accuracy, with maximum discrepancies of about 0.5 dB.

An all-optical converter from nonreturn-to-zero (NRZ) to carrier-suppressed return-to-zero modulation format is proposed and experimentally demonstrated. The converter is based on cross gain and phase modulation in a semiconductor optical amplifier. Single- and multichannel operation is experimentally assessed at 40 Gb/s. In single-channel operation, the required optical signal-to-noise ratio for a bit-error rate of $10^{-9}$ is improved by 3 dB, in comparison to the input NRZ. Considering multichannel operation with two and four channels, this improvement decreases by only 0.6 and 1.5 dB, respectively.

This letter analyzes the degradation induced by cross-phase modulation (XPM) in multisection links using only dispersion precompensation. A previously developed analytical model is used to assess the XPM induced by a pre-compensated nonreturn-to-zero 10-Gb/s signal. It is shown that the XPM degradation is conditioned by the relation between the XPM resonance frequency and the relative minimums and maximums of the power and phase spectral densities of the pre-compensated pump. This observation provides a simple rule valid for uniform links, to relate the precompensation with the XPM degradation.

An alternative approach is presented to minimize the impact of fiber nonlinearities in optical single-sideband systems at 10 Gb/s using electrical dispersion precompensation by adding optical postcompensation while maintaining low accumulated dispersion. Numerical simulation is used to show that the dependence of the system performance on the dispersion compensation scheme resembles a pseudolinear transmission regime, although it is not considered as such in the strictest sense. A launched power increase of 3 dB is achieved when compared to full electrically precompensated systems. A feasible implementation of the optical transmitter is considered, which imposes a maximum transmitted modulation depth. It is shown that low modulation depths result in an asymmetric optimum dispersion compensation map.

An all-optical converter from return-to-zero (RZ) pulses to the nonreturn-to-zero (NRZ) format is presented. The converter operates in two stages: the laser generated in a gain-clamped semiconductor optical amplifier (SOA) is modulated with the data signal; afterwards this signal is wavelength-converted by cross-gain modulation in a common SOA. The setup is noninverting and can feature wavelength conversion. Experimental error-free conversion from 5- and 40-ps RZ pulses to NRZ format is presented at 10 Gb/s using a 211-1 bit sequence

An all-optical setup to generate vestigial sideband signals based on self-phase modulation in a semiconductor optical amplifier is experimentally demonstrated at 10 Gb/s. Sideband suppressions higher than 15 dB are reported with improved eye opening. Wavelength-independent operation over 26 nm is demonstrated. Increased chromatic dispersion tolerance is verified: a receiver sensitivity penalty of 5.3 dB, relative to back-to-back, is obtained after transmission over 2720 ps/nm; whereas conventional double sideband is penalized by 4.0 dB after 1360 ps/nm

A new polarization-assisted method for the implementation of multiwavelength optical orthogonal codes for optical code-division multiple-access is presented. It is based on fiber Bragg gratings written in highly birefringent fibers. The advantages of the technique include the reduction of the interference between users and the reduction of the number of coders and decoders needed.