A new linearization method for optical transmitters based on directly modulated lasers (DMLs), named the Stretched A method, was proposed in Parts I and II of this work. Parts I and II presented the theoretical framework of the method for non-return-to-zero (NRZ) modulation and related detailed simulation and experimental results. Here, we extend the method to pulse amplitude modulation schemes (PAM). Focussing on 4-level PAM (PAM-4), we present the theoretical background and discuss implementation options. A simplified variation of the method for the generation of PAM signals with a significantly lower number of sub-currents is proposed. Simulation studies for PAM-4 transmission at 50 GBaud (100 Gb/s) and an experimental proof-of–principle demonstration at 16 GBaud (32 Gb/s) are reported based on 850 nm vertical-cavity surface-emitting lasers (VCSELs). For PAM-4, products of effective eye diagram areas ( ${\boldsymbol{PS}}$ ) of 0.6 × 10 −5 and 43 × 10 −5 before and after applying the Stretched A method were measured, demonstrating an improvement ratio of ∼72. The sensitivity and tolerances of the method are analyzed using simulation and experiment.

In this paper, we report the first experimental demonstration of the Stretched A linearisation method. The theoretical framework of the method and related simulation studies have been presented in Part I of this work. Here, we apply the method on an 850 nm multimode vertical-cavity surface-emitting laser (VCSEL) specified for operation at 10 Gb/s and demonstrate good quality optical output waveforms with little non-linear distortion at 16 Gb/s using non-return-to-zero (NRZ) modulation. An experimental sensitivity analysis on the generation of the required modulating current components demonstrates large tolerance to parameter mismatch. In addition, it is shown that an adaptive gain scheme can improve the tolerance to the timing offset between the current components. Potential hardware implementations using either analog or digital electronics are also discussed.

The performance of directly-modulated lasers (DMLs) is severely impaired by nonlinear behaviour when operating at high symbol rates. We propose a new linearization method for DML-based transmitters which can significantly reduce nonlinearity. This method, named the “Stretched A” (StrA) method, relies on the generation of an approximation to the ideal modulating current that generates a linear optical output waveform. In Part I of this work, the theoretical framework of the proposed method is presented and detailed simulation studies illustrate its implementation and demonstrate the benefits it offers. Although the method is applicable to any type of DML, the simulation studies presented herein focus on optical links based on vertical-cavity surface-emitting lasers (VCSELs) as these comprise the vast majority of short-reach optical links. Part II of this work presents the proof-of-principle experimental demonstration of this new linearization method and discusses its possible implementations using either analog or digital electronics.

A new practical method to correct the non-linearity of directly-modulated lasers is presented and an experimental proof-of-principle demonstration is reported. High-quality NRZ transmission at 16 Gb/s is achieved using a VCSEL specified for 10 Gb/s.

The first signal model for a single photon avalanche diode (SPAD)-array communication receiver for multilevel modulation schemes is reported. This paper proposes a novel, generalised SPAD array signal and noise model for both digital and analogue, synchronous and asynchronous SPAD readout arrays, which includes the competition between the input photons, dark counts and after-pulsing counts. With this contention signal and noise model, multilevel signals including the signal variation after distortion or equalisation can be evaluated. Also, we report the first numerical investigation for SPAD-based, high data rate, free space, visible light communication using higher order pulse amplitude modulation (PAM) with matched filter, linear and non-linear Volterra post-equalization. Simulations have been carried out to analyze and compare the bit error rate (BER) performances under a variety of conditions. The model is verified by comparison with published experimental results.

In this paper, we review recent work on the development of a novel low-complexity equalizer that can enable single-lane <100 Gb/s short-reach optical links based on carrierless amplitude and phase modulation. This equalizer, named the CAP equalizer, can mitigate the transmission impairments in the link due to a non-ideal channel frequency response, providing significant performance advantage over conventional FFE and DFE equalizers and enabling higher data rates and longer reach. Its use is demonstrated in a VCSEL-based MMF link achieving data transmission of 112 and 124 Gb/s over 100 m OM4 MMF.

This paper reviews optical fibre technology for local area optical communications systems. Technologies used in local systems include single and multimode fibre, single and multimode lasers, optical modulators, photodetectors, wavelength division multiplexing, multilevel modulation formats, electronic packet switching, electronic equalization and error correction. These methods have enabled the local area optical link data rate to increase from 0.1 Gb s-1 in 1990 to nearly a Tb s-1 in 2019. The challenges to increasing link data rates further, while reducing the transmitted power per bit, at reduced cost are discussed. Potential technical solutions and newly proposed methods which might address these challenges are highlighted.

Spectral efficient modulation formats can enable the transmission of higher data rates than conventional on-off keying (OOK). Carrierless amplitude and phase modulation (CAP) is such an attractive modulation scheme that has been widely considered for use in different types of optical links. The scheme however can suffer from intersymbol interference (ISI) and channel crosstalk (CCI) when the frequency response of the channel is not ideal. Conventional equalizers based on feedforward (FFE) and decision feedback (DFE) equalizers are easy to implement in practice and can mitigate some of the induced ISI. However, they fail to suppress the induced CCI in the link as each channel is equalized independently. As a result, we have recently proposed the use of a new equalizer structure for use in CAP-based optical links to mitigate these transmission impairments. This new equalizer, named CAP equalizer, can be formed with conventional FFEs and DFEs with minimal additional complexity whilst providing significant performance advantages. In this paper therefore, we review the equalizer structure and report recent demonstrations of its use in short-reach optical links. We present experimental studies on a 112 Gb/s CAP-16 VCSEL-based OM4 MMF link and a 4 Gb/s CAP-16 LED-based POF link and compare the performance of the links when both a conventional FFE and DFE equalizer and the newly proposed CAP equalizer are used. The results clearly demonstrate that the CAP equalizer offers improved receiver sensitivity and enables successful data transmission over longer fibre reaches.

A novel artificial neural network equalizer for use in short-reach optical links is proposed. 112Gb/s and 56Gb/s CAP-16 data transmission are demonstrated by simulation and experiment respectively with receiver sensitivities of −4 and −7 dBm. © 2019 The Author(s)

112 Gb/s data transmission over 150 m MMF with a BER within the HD-FEC threshold is experimentally demonstrated using a novel equalizer type and CAP modulation. A ∼2 dB receiver sensitivity improvement over an adaptive DMT scheme is achieved for a 100 m MMF link.

A novel equaliser that mitigates phase non-linearity in short-reach optical links is proposed. It is demonstrated via simulation that the equaliser enables CAP-based 100Gb/s data transmission using a single VCSEL over 100m OM4 MMF providing improved link performance over OFDM.

Inter-symbol-interference (ISI) noise in Continuous Variable Quantum Key Distribution (CVQKD) is caused by the overlap between the consecutive detected signals. In practice, this problem limits the repetition rate of the quantum signals for a given bandwidth detector and hence reduces the rate of secure key generation. We propose a method of using equalization to mitigate ISI noise in CVQKD systems. Its feasibility is studied using an analytical model with practical parameters. The improved performance of CVQKD detection is investigated using secure key analysis. The simulation results show an increase of 12.8 Mbits/s in secure key rate for a 1GHz bandwidth CVQKD system operating over a 20 km link.

Optical feed-forward equalization (FFE) of a single-mode fiber link has been studied theoretically. It is demonstrated that a simple short optical FFE having three to five taps can provide significant performance improvement of bandwidth-limited transmitters for both short and long reach links compared with the nonequalized scenario. Since optical FFEs can simultaneously operate on multiple wavelength division multiplexing channels, they can provide a simple, low power, and cost-effective solution for high-speed optical communication links.

Optical equalizers and electrical equalizers for a fiber link has been explicitly studied and compared. The results demonstrate that both can improve optical links significantly but optical equalizers can perform better in terms of noise enhancement.

We propose the first combined 4×100Gb/s hybrid multiband CAP-16 transmitter and QAM-16 receiver system and simulations show that it has 0.7 dBo (2.2 dBo) more power margin than 8×50Gb/s (4×100Gb/s) PAM-4 over DML (EML) SMF link.

We review possible architectures for 400 Gigabit Ethernet links based on advanced modulation formats for the first time. Their optical link power budget, digital complexity, and power dissipation are compared via simulations. The challenges of implementing the physical layer are discussed.

Mode partition noise (MPN) can become the dominant limitation in 850 nm VCSEL-based multi-mode fiber (MMF) links at high data rates. Fluctuations in the partition of energy between the transverse modes of the VCSEL combined with the chromatic dispersion in the fiber leads to intensity noise at the receiver. The impact of MPN on non-equalized and equalized links has been studied with a numerical model of the VCSEL and MMF. The MPN in 25 Gb/s VCSELs has been investigated by examining noise in individual mode groups isolated using a thin film Fabry-Perot filter. The measured k factor below 0.15 should enable links significantly longer than 100 m at 25 Gb/s and higher data rates.

Hybrid multiband (HMB) CAP/QAM transmitter/receiver systems are proposed for the first time. Simulation results are provided to show the feasibility of 100 Gb Ethernet links employing a single-laser source transmitting HMB CAP-16/QAM-16, CAP-32/QAM-32, and CAP-64/QAM-64 signals. The proposed hybrid scheme has low sensitivity to directly modulated laser non-linearities. We found that QAM receivers bring about identical jitter tolerance to ideally phase compensated CAP receivers, and QAM receivers are more practical since no phase tracking and compensation are required. Compared with the case of using a standard non phase compensated CAP receiver, the use of the modified QAM-16/32/64 receiver significantly lowers system timing jitter sensitivity in the multiband, as well as single-band case. Results also show that the use of increasing number of bands causes increased system power margin. For practical jitter conditions of +/- 6 ps, three HMB CAP/QAM systems with optimum band counts are identified to be capable of supporting single-laser 100 Gb/s transmission over 15-km SMF.

Layered asymmetrically clipped optical orthogonal frequency division multiplexing (ACO-OFDM) with high spectral efficiency is proposed in this paper for optical wireless transmission employing intensity modulation with direct detection. In contrast to the conventional ACO-OFDM, which only utilizes odd subcarriers for modulation, leading to an obvious spectral efficiency loss, in layered ACO-OFDM, the subcarriers are divided into different layers and modulated by different kinds of ACO-OFDM, which are combined for simultaneous transmission. In this way, more subcarriers are used for data transmission and the spectral efficiency is improved. An iterative receiver is also proposed for layered ACO-OFDM, where the negative clipping distortion of each layer is subtracted once it is detected so that the signals from different layers can be recovered. Theoretical analysis shows that the proposed scheme can improve the spectral efficiency by up to 2 times compared with conventional ACO-OFDM approaches with the same modulation order. Meanwhile, simulation results confirm a considerable signal-to-noise ratio gain over ACO-OFDM at the same spectral efficiency.

100 Gb/s PAM4-CAP2 modulation is demonstrated for next-generation datacommunication links. Simulation studies indicate a power budget advantage of 2.5 dBo relative to PAM8 modulation. A real-time experimental demonstration is performed.