Generally speaking, the impairments caused by imperfections in an all-optical network will accumulate along the optical path through the network, and the signal quality will become unacceptable after some distance. Therefore, there will be transmission limitations in all-optical networks. Optical impairments which degrade the signal quality can be classified into three broad categories [1].
(a) Noise: random signal fluctuations that are often treated as a Gaussian process and can be signal level dependent. Examples are optical amplifier noise and laser noise.
(b) Distortion: modification of the average signal waveform, for example, the average waveform of the marks and spaces taken separately.
Distortion is caused by nonlinearities or fiber dispersion effects and may be signal level and pattern dependent and can lead to burst errors and BER floors. Examples include laser, optical amplifier and fiber nonlinearities, laser diode bit pattern dependent response, receiver bit pattern dependent response, chromatic and polarization mode dispersion, and phase induced intensity noise.
(c) Timing: fluctuations in the time registration of the bits. Timing jitter can occur as quickly as bit-to-bit or accumulate over many bit periods.
In order to achieve long link lengths, many well-known and perhaps countable deleterious effects of optical transmission must be minimized or controlled carefully, and all of these effects are controlled through the network design. The transmission margin is ensured to be enough, while the worst-case impacts of transmission impairments are taken into consideration in the network design. Notable transmission impairments include [1]
(a) Amplifier noise
(b) Amplifier distortion and transients (c) Chromatic dispersion (CD)
(d) Polarization–mode dispersion (PMD)
(e) Fiber nonlinearity induced distortion and crosstalk (self-phase modulation (SPM), cross-phase modulation (XPM), four-wave mixing (FWM), stimulated Raman scattering (SRS), and stimulated Brillouin scattering (SBS).
(f) Timing jitter
(g) Polarization effects (h) Interference effects (MPI) (i) Pump laser RIN transfer (j) Optical filter distortion (k) Linear crosstalk
As the network capacity increases, the above factors reduce the window of operability further. Therefore, new technologies have been developed to keep the window open, and optical performance monitoring (OPM) and impairment mitigation techniques are potential means of widening this window. These techniques can improve the control of transmission and physical layer fault management, and they are essential in building a high capacity and reliable all-optical network, which are expected to be transparent and dynamically reconfigurable. In future optical networks, OEO conversion will be mostly eliminated, and therefore the bit-rate-, modulation-format-, and protocol- independent performance monitoring and impairment mitigation techniques are required. Moreover, wavelength channels are added to and dropped from optical nodes dynamically, and these techniques have to be without knowledge of the origin or transport history of data stream. Using maintenance calls to manually adjust optical components or network elements is not practical, so the monitored signals are needed for impairment compensation or fault correction. These reconfigurable properties in all-optical networks will drive the need for adaptive compensation techniques, such as chromatic dispersion and PMD equalization.
Historically, performance monitoring refers to monitoring at SONET/SDH layer, and each data stream will be recovered at every network node. The optical
performance monitoring started to be paid attention, while people think more about optical network rather than optical transmission. The purpose of OPM is to examine the signal quality in the optical domain, and the broad spectrum of OPM includes a plethora of parameters to be monitored which can be classified into three categories as shown in the Fig. 1.2 [2].
(a) Signal loss monitoring: refers to the monitoring of in-line component failures and fiber cuts that cause a change in opacity in optical transmission. It is particularly important to monitor the active components such as EDFA and optical crossconnects due to their high failure probability.
(b) Signal alignment monitoring: on the other hand, concerns with the alignment of signal wavelength, filter position, and pulse carver to ensure proper operation.
(c) Signal quality monitoring: pertains to the monitoring of a multitude of disparaging effects that must be minimized or controlled. These impairments include noise, distortion, and timing jitter.
The OPM techniques can either be digital or analog. High-speed processor is used in digital techniques to recover bits embedded on optical waveforms, and high correlation with bit error rate (BER) can be obtained. However, digital methods can not isolate the effects of individual impairments effectively. Analog waveforms are measured and analyzed directly in analog methods, so analog measurement are typically protocol-independent. Analog measurements can be subdivided further into time domain and spectrum methods. The time domain methods includes eye diagram and auto-correlation measurements, and the spectrum methods includes optical spectrum and RF spectrum measurements.
Compared to time domain methods, the spectrum techniques can be implemented by narrowband electronics, so it is less complicated and cost effective, even high frequency subcarriers are involved. In addition, the monitoring sensitivity may be increased as well, because narrow detection bandwidth is used. High sensitivity is of importance for optical monitoring, because the optical power levels for monitoring is usually quite small.
Indeed, performance monitors that provide feedback for closed loop control of compensation elements will be desirable for management systems to mitigate or compensate the degrading effects, and this OPM combination might be implemented with little additional hardware. Real time performance monitors can measure the signal quality, and the monitored signal will be used to trigger alarms and feedback to active compensators adaptively. As a result, the impairments caused by component aging, fiber degradation or environmental changes will be controlled or mitigated, and therefore system reliability is guaranteed.
In this dissertation, we will focus on performance monitoring and impairment mitigation techniques related to (1) chromatic dispersion monitoring and equalization, (2) pulse carver and data alignment monitoring, and (3) electro-optical pre-distortion technique. For high-speed wavelength division multiplexed (WDM) systems, the system performance is strongly affected by fiber chromatic dispersion and fiber nonlinearity. In a re-configurable WDM network, the accumulated chromatic dispersion of each channel may change frequently. To achieve full dispersion compensation, a variable dispersion compensator (VDC) is needed to optimize the residual dispersion for each channel. Moreover, the chromatic dispersion is sensitive to ambient temperature, and this variation would harm the high-speed signals greatly.
So, the online automatic chromatic dispersion equalization techniques become indispensable in high-speed WDM systems.
The combination of the two modulators for the pulse carving and data modulation poses the problem of maintaining the correct timing of two devices.
Since the optical path length in between depending on the ambient temperature, the bit-synchronous modulation is easily asynchronized. Accordingly, to drive phase modulator synchronously with the pulse sequence, the clock timing alignment is essential for bit-synchronous modulation, such as RZ (return-to-zero) and CS-RZ (carrier-suppressed RZ), which use two optical modulators for data and clock, respectively.
Duo to the stringent carrier-to-noise ratio (CNR) requirement of analog AM-VSB signals, semiconductor optical amplifiers (SOAs) have long been considered not suitable for CATV systems which transmit multi-channel AM-VSB signals. This is because the amplifier’s nonlinear distortion (NLD), which is caused by signal-induced carrier density modulation, can potentially degrade system performance seriously, and can not be accepted by AM-VSB signals. As a result, linearization techniques to improve SOA saturation characteristic and thereby to
expand the SOA input power dynamic range are attractive for CATV systems.
This dissertation is organized as follows. Chapter 2 describes the dispersion monitoring that we propose for high-speed transmission systems at 40 Gb/s or beyond, and it is a novel post-detection method based on an optical delay-and-add filter (DAF).
The proposed method can be used with or without a pilot tone, and works well even when there exists a residual chirp due to the finite DC extinction ratio of a Mach Zehnder modulator or self-phase modulation. In Chapter 3, a novel technique for monitoring the timing alignment between a pulse carver and a phase modulator in RZ-DPSK systems is proposed and demonstrated. An optical frequency discriminator is used to monitor the spectrum broadening caused by timing misalignment. Chapter 4 describes a novel linearization technique for CATV external-modulated systems, and it can simultaneously suppress the nonlinear distortions (NLDs) generated in a transmission system and increase the received carrier-to-noise ratio (CNR). Both analytical and experimental results are presented.
Finally, conclusions and future works are drawn in Chapter 5.
Fig. 1.1 The broad spectrum of OPM [2]