Ultrafast time-domain technology and its application in all-optical signal processing

In this paper, we review recent advances in ultrafast optical time-domain technology with emphasis on the use in optical packet switching. In this respect, several key building blocks, including high-rate laser sources applicable to any time-division-multiplexing (TDM) application, optical logic circuits for bitwise processing, and clock-recovery circuits for timing synchronization with both synchronous and asynchronous data traffic, are described in detail. The circuits take advantage of the ultrafast nonlinear transfer function of semiconductor-based devices to operate successfully at rates beyond 10 Gb/s. We also demonstrate two more complex circuits-a header extraction unit and an exchange-bypass switch-operating at 10 Gb/s. These two units are key blocks for any general-purpose packet routing/switching application. Finally, we discuss the system perspective of all these modules and propose their possible incorporation in a packet switch architecture to provide low-level but high-speed functionalities. The goal is to perform as many operations as possible in the optical domain to increase node throughput and to alleviate the network from unwanted and expensive optical-electrical-optical conversions.     

Optical signal processing using integrated multi-element SOA-MZI switch arrays for packet switching

The use of discrete but interconnected SOA-MZI switches for performing logical and highly functional processing tasks, demonstrating the multi-functional potential of the photonic switching elements is discussed. An all-optical 3R burst-mode receiver consisting of four SOA-MZI switches and operating error-free with 40-Gb/s optical bursts, proving that interconnection of multiple switching units can lead to the realisation of key network node functionalities offering increased intelligence at the physical layer is presented. In order to allow for easier interconnectivity between the SOA-MZI switches and to provide compactness and cost effectiveness to the developed subsystems, the integration of multiple switches into the same platform is proposed. To this end, the implementation of the first integrated quadruple SOA-MZI switch array is reported, increasing the integration density level and reducing packaging and pigtailing costs. Finally, possible applications of integrated multiple switch arrays are discussed, indicating their suitability for producing compact circuits performing common processing tasks in a multi-wavelength environment, as well as their potential to lead to the development of an all-optical high-speed packet switched node by implementing critical packet switching functionalities in a compact and efficient way.     

Recipe for intensity modulation reduction in SOA-based interferometric switches

This paper presents a theoretical and experimental analysis of saturated semiconductor optical amplifier (SOA)-based interferometric switching arrangements. For the first time, it is shown that such devices can provide enhanced intensity modulation reduction to return-to-zero (RZ) formatted input pulse trains, when the SOA is saturated with a strong continuous-wave (CW) input signal. A novel theoretical platform has been developed in the frequency domain, which reveals that the intensity modulation of the input pulse train can be suppressed by more than 10 dB at the output. This stems from the presence of the strong CW signal that transforms the sinusoidal transfer function of the interferometric switch into an almost flat, strongly nonlinear curve. This behavior has also been verified experimentally for both periodically and randomly degraded, in terms of intensity modulation, signals at 10 Gb/s using the ultrafast nonlinear interferometer as the switching device. Performance analysis both in the time and frequency domains is demonstrated, verifying the concept and its theoretical analysis.     

An SOA-MZI NRZ Wavelength Conversion Scheme With Enhanced 2R Regeneration Characteristics

We present a novel scheme for all-optical nonreturn-to-zero (NRZ) wavelength conversion with enhanced 2R regenerative characteristics. It employs a hybridly integrated semiconductor optical amplifier Mach-Zehnder interferometer using bidirectional data injection and an additional external continuous-wave signal for differentially biasing the interferometer arms, optimizing gain and phase conditions during the switching functionality. Experimental verification of its superior 2R regenerative capabilities are demonstrated for 10-Gb/s NRZ data signals.     

Optical Static RAM Cell

We demonstrate an optical static random access memory cell that provides read and write functionality at 5 Gb/s. The circuit comprises a hybridly integrated semiconductor optical amplifier-Mach-Zehnder interferometer (SOA-MZI) flip-flop serving as the memory unit and two additional SOA-based cross-gain modulation switches for controlling access to the memory cell.     

40-Gb/s 3R Burst Mode Regenerator Using Four Integrated MZI Switches

We demonstrate an all-optical 3R burst mode regenerator operating error-free with 40-Gb/s data packets. It is comprised of a sequence of four hybridly integrated semiconductor optical amplifier Mach-Zehnder interferometric switches and is shown to operate with short, variable length and asynchronous data packets, with a dynamic range of 9.3 dB     

Pulse repetition frequency multiplication with spectral selection in Fabry-Perot filters

We present methods for obtaining high-repetition-rate full duty-cycle RZ optical pulse trains from lower rate laser sources. These methods exploit the memory properties of the Fabry-Perot filter for rate multiplication, while amplitude equalization in the output pulse train is achieved with a semiconductor optical amplifier or with a second transit through the Fabry-Perot filter. We apply these concepts to experimentally demonstrate rate quadruplication from 10 to 40 GHz and discuss the possibility of taking advantage of the proposed methods to achieve repetition rates up to 160 GHz.     

Cascadability Performance Evaluation of a New NRZ SOA-MZI Wavelength Converter

We evaluate the cascadability performance of a new semiconductor optical amplifier (SOA) Mach-Zehnder interferometer-based nonreturn-to-zero wavelength converter in a loop experiment. We use the bidirectional data injection control scheme with an additional continuous-wave signal to optimize the gains and phases imparted by the SOAs. The scheme has been shown to be capable of eight cascaded, error-free wavelength conversions at 10 Gb/s.     

Additive noise and jitter performance analysis of passive optical interferometers operated at ultrahigh rates

In this paper, we theoretically associate the additive noise, the amplitude jitter and the timing jitter at the input and output of passive optical interferometers. We make use of the theoretical results to assess the noise and jitter performance of interferometer based applications such as pulse repetition frequency multiplication and clock recovery. We show that, for both applications, interferometers may successfully reduce the noise and the jitter existing in the input signals, and thus yield very high quality output signals. Furthermore, we focus on the practical aspects of deploying Fabry-Pe/spl acute/rot interferometers in rate multipliers and clock recoveries, and provide rules for selecting the characteristics of the Fabry-Pe/spl acute/rot interferometer to meet specific quality requirements for the output signal.     

All-Optical 3R Burst-Mode Reception at 40 Gb/s Using Four Integrated MZI Switches

We demonstrate an all-optical retime, reshape, reamplify (3R) burst-mode receiver (BMR) operating error-free with a 40-Gb/s variable-length asynchronous optical data packets that exhibit up to 9-dB packet-to-packet power variation. The circuit is completely based upon hybrid integrated Mach-Zehnder interferometric (MZI) switches as it employs four cascaded MZIs, each one performing a different functionality. The 3R burst-mode reception is achieved with the combination of two discrete all-optical subsystems. A reshape, reamplify BMR employing a single MZI is used first to perform power equalization of the incoming bursts and provide error-free data reception. This novel approach is experimentally demonstrated to operate error-free, even for a 9-dB dynamic range of power variation between bursty data packets and for a wide range of average input power. The obtained power-equalized data packets are then fed into a 3R regenerator to improve the signal quality by reducing the phase and amplitude jitter of the incoming data. This packet-mode 3R regenerator employs three MZIs that perform wavelength conversion, clock extraction, and data regeneration for every packet separately and operates at 40 Gb/s, exhibiting rms timing jitter reduction from 4 ps at the input to 1 ps at the output and a power penalty improvement of 2.5 dB