Current WDM Systems in optical networks are commonly operated at 2.5 and 10 Gbit/s per wavelength. First transmission systems, offering the possibility of using 40 Gbit/s/λ transponders, are now commercially available. Transaction of various field trials over the last years indicates remarkable interest for this systems among the network operators. Potential further stages of systems at higher channel data rates e.g. 80, 100 or 160 Gbit/s/λ thus attracted more and more attention in the R&D community. Existing concepts like broadband dispersion compensation, distributed Raman amplification, bandwidth efficient and impairment tolerant modulation formats have been applied, but also new circumstances and physical impairments have to be considered, which are negligible at lower data rates. For data rates above 40 Gbit/s chromatic dispersion causes pulses to broaden extremely rapidly, so that transmission behaviour can be regarded as "quasi-linear". We will present a comprehensive overview to what extend system reach limits can be stressed for 160 Gbit/s/λ data rates when different parameters such as fiber type (standard single-mode fiber SMF and non-zero dispersion fiber NZDSF), modulation format (return to zero RZ, carrier suppressed return to zero CS-RZ, intensity modulated differential phase shift keying IM DPSK), different dispersion compensation schemes and signal power levels are optimized. Further, the benefit from using balanced instead of single ended receiver is investigated for IM DPSK.

This paper introduces practical and high-performance transmission systems that employ distributed Raman amplification (DRA) technologies. The systems incorporate DRA/EDFA hybrid amplifiers as inline amplifiers with limited DRA pump powers. These powers are determined with respect to the practical safety and reliability of the systems against intense pump light. The practical aspects and merits of our systems are described both in detail and qualitatively. It is shown that a hybrid amplifier system using DSF performs better than one using SMF with limited pump powers. The use of DRA means that the optical SNR of the former system is typically about 2-3 dB higher than that of the latter. Moreover, this paper reports successful results of long-haul transmission field trials using the hybrid amplifier scheme in the L-band over installed DSF with an aggregate capacity of 1.28 Tbit/s (32 x 43 Gbit/s).

Transport fiber deployed in metro and long-haul networks exhibit chromatic dispersion which if not addressed closes the receive eye. At 10Gbps line rates, dispersion management strategies are required to allow reaches to surpass their dispersion limited values. Several solutions exist which resolve this limitation. A survey of these is presented. Since dispersion is a linear operator acting upon the E-field of the transmitter, precompensation is one such dispersion management strategy. This paper presents the implementation of a transmitter which enables tunable electronic precompensation by E-field modulation. The key technology considerations brought to bear on the realization of this precompensation are discussed. The compensation of additional impairments and new capabilities of this transmitter are discussed. Experimental results, with standard single-mode fiber (G.652), without any optical compensators, demonstrate the capabilities of this new dispersion management strategy.

To date, the mainstream Ethernet Passive Optical Network (EPON) bandwidth allocation schemes as well as the new IEEE 802.3ah Ethernet in the First Mile (EFM) Task Force specifications have been centralized, relying on a component in the central office, Optical Line Termination (OLT), to provision upstream traffic. Hence, the OLT is the only device that can arbitrate time-division access to the shared channel. Since the OLT has global knowledge of the state of the entire network, this is a centralized control plane in which the OLT has centralized intelligence. One of the major problems associated with a centralized architecture is the "single-point of failure" problem that is the failure of the OLT software will bring down the whole access network. It is the purpose of this work to propose a distributed solution to this problem, and to devise and experimentally demonstrate the feasibility of implementing a novel Ethernet over Star Coupler-based PON architecture that uses a fully distributed time division multiple access arbitration schemes. Specifically, we assess the viability of implementing a distributed control plane architecture that facilitates internetworking among connected users. In addition to the added flexibility and reliability associated with distributed control plane architecture, as well as emulating shared LAN capability among different users, the distributed networking architecture and the associated bandwidth allocation algorithms have characteristics that make them far better suited for provisioning Quality of Service (QoS) schemes necessary for multimedia services over a single line.

DWDM optical communication systems that transport bandwidth exceeding Tbps per fiber require fast and cost-effective methods to measure the transmission performance parameters of the optical link and the receiver at each system port. The reason is that fiber non-linearity and component degradation affect the quality of signal and the overall channel performance. However, in many applications, it suffices to verify whether the receiving channel meets the minimum performance criteria. In such case, simple and conclusive methods should be used to measure quickly and in-situ the channel performance parameters. Such parameters in optical communication are the bit error rate (BER), signal to noise ration (SNR), Q-factor, noise factor (NF), received optical power level, and more. In this paper, we describe a statistical sampling methodology that estimates these channel performance parameters in a very short interval and for all-practical purposes in real-time, which also is protocol independent; that is, performance is not estimated according to error detecting codes embedded on the signal of asynchronous data payloads such as in GbE, synchronous such as SDH/SONET.

The Optical Transport Network (OTN) is the extremely important technology to create the next-generation core optical networks with the capacity of several Tbit/s, because it can contribute to the reliability and transparency. We discuss the OTN-based Optical Cross-Connect systems (OXCs), and verify the fast restoration using OTN functionality.

Recently, Optical WDM technology is introduced into backbone networks. On the other hand, as the future optical switching scheme, Optical Burst Switching (OBS) systems become a realistic solution. OBS systems do not consider buffering in intermediate nodes. Thus, it is an important issue to avoid overlapping wavelength reservation between partially interfered paths. To solve this problem, so far, the wavelength assignment scheme which has priority management tables has been proposed. This method achieves the reduction of burst blocking probability. However, this priority management table requires huge memory space. In this paper, we propose a wavelength assignment algorithm that reduces both the number of priority management tables and burst blocking probability. To reduce priority management tables, we allocate and manage them for each link. To reduce burst blocking probability, our method announces information about the change of their priorities to intermediate nodes. We evaluate its performance in terms of the burst blocking probability and the reduction rate of priority management tables.

Optical Telecommunications bandwidth, spurred by the growth of the internet, experienced unprecedented growth in the late 1990's. The creation of new enterprises was vast and the expansion of established component, system and services companies was also breathtaking. This period of speculative growth was followed in 2001-2004 by one of the most significant market crashes in history. While $20B of venture capital was invested in optical telecom in the last 10 years, the vast majority of that has been written off in the last four. Countless start-ups inaugurated with great fanfare at the end of the 20th century were unceremoniously shut down at the start of the 21st. (1) As in all speculative bubbles, innovative technologies were born and buried. Nonetheless, new capabilities emerge from the chaos and disruption; one such example is the advent of Optical MEMS (MOEMS). Its development was vigorously pursued in both academic and corporate laboratories during the boom and, in the author's view; MOEMS constitutes a powerful and versatile tool set that is an invaluable residual of the last few years. In Telecommunications, MOEMS has proven to be the technology of choice for many optical switching and wavelength management applications. (2) Variable Optical Attenuators (VOA), Wavelength Blockers (WB), Dynamic Gain Equalizers (DGE), and most recently Wavelength Selective Switches (WSS) are being used in the numerous recent network deployments. Moreover, agile networks of the future will have MOEMS at every node. This presentation will provide an overview of the history of MOEMS in Telecommunications, discuss its byproducts and project the future of the technology.

Recent development of ultra high-speed optical code-label processing and its application to packet switching are reported. 10 to 40Gbit/s interface, scalable, optical packet switch prototype based on optical code-label processing is developed. Optical packet switching with 200Gchip/s all-optical label processing, 10 to 40Gbit/s/port variable data rate packet switching, and optical buffering to avoid packet collisions is experimentally demonstrated. A novel packet bit error rate (BER) and loss real-time measurement method is proposed. The first 40Gbit/s packet BER and loss measurement system is shown. 40Gbit/s BER and loss evaluation with various conditions is experimentally demonstrated. In real time, only the payload part of packet and burst stream with fluctuated interval time is evaluated. BER and packet loss of randomly received packet sequence because of routing and buffering can be also evaluated by the proposed system. 10Gbit/s packet BER and loss measurement with optical label switching, buffering, and preamble free optical packet 3R are experimentally demonstrated.

This paper proposes a segmented network architecture that is very effective in constructing a large scale network that uses transparent OXC's, and clarifies the network design criteria. Network costs with different network design scenarios are investigated. In particular, IP plus optical layer networks with multi-layer traffic engineering and one plus one optical path protection are examined. Numerical experiments show the effects of the segment size on the network cost and that the proposed designs are effective in designing large-scale networks.

Increasing the data rates by a factor of four has historically reduced the cost for the transmission of a unit bandwidth by approximately 40%. As the next natural increase in data rate, 40 Gb/s has been the focus of extensively investigation in research and development labs around the world for several years. However, despite the obvious potentials, 40 Gb/s systems have not yet been commercially deployed, in part because 40 Gb/s is associated with a number of misconceptions, e.g., that 40 Gb/s technology is not mature and that the transmission distance is severely limited by fiber dispersion. In this paper we address the practical requirements that 40 Gb/s systems must meet to become commercially deployed. We show that seamless migration of a 10 Gb/s system to 40 Gb/s per channel is possible with correctly designed line cards. Moreover, we discuss the technologies needed to implement different modulation formats, and the corresponding trade-off between complexity/cost of line cards and the achievable fiber transmission distance.

Recently, further high-speed and larger capacity communication is requested in the IP network. And multilayer network that constructs the IP network on an optical network such as IP over WDM is focused on. Under such a network environment, the multipath routing for setting two or more paths between a starting and ending nodes in each network is researched. The multipaths are taken advantage of load balance and path recovery and damaged traffic is switched to other paths at the starting node if a failure happens on a path. However, the traditional routing often finds multipaths overlapping on some links and nodes because of its low "disjoint property". Hence, path recovery is difficult because the multipaths may be interrupted at the same time. In order to resolve the problem in multilayer network, we propose the local path configuration method considering global path optimality. Concretely, we propose a network architecture where networks are connected by multiple edge nodes and multipath routing method with new link cost and parallel calculation of multipaths. As a result, the multipaths are prevented from overlapping and the disjointed multipaths enable efficient path recovery in case of a failure. Moreover, we verify the effectiveness of the disjoint property.

High speed optical networks require optical transmission to be robust to noise as well as linear and non-linear distortions. Robustness to the time evolution and reconfiguration of the optical channel is also required. In this tutorial we discuss traditional optical modem and line technologies which have been deployed and, their relation to the above considerations. In addition, the significance of emerging digital optical-modem technology is explored.

40 Gbit/s or higher bit-rate DWDM has been seen as a promising candidate for next generation optical transport networks for some years because it reduces the number of wavelengths, resulting in easier operation and maintenance. It can also potentially reduce the cost and footprint of equipment. However, the reality is that even 40 Gbit/s DWDM has not yet appeared in commercial long-haul systems. This seems to be due not only to financial depression in the market but also to technical difficulties with 40 Gbit/s. The most significant difficulty in 40 Gbit/s systems is that an additional 6 dB of optical signal-to-noise ratio (OSNR) is required compared to 10 Gbit/s systems. One way to gain the 6 dB needed for 40 Gbit/s is by combining (1) the 3 dB provided by the DPSK, with (2) the 3 dB available from the stronger forward error correction technique (FEC). Recent advance in opt-electrical devices and high-speed LSI technologies have made it possible to employ advanced modulation formats and powerful and complex FEC techniques. Another challenge in higher bit-rate transport systems is to extend tolerance against the unwanted waveform distortion due to the wavelength chromatic dispersion and polarization mode dispersion. Various kinds of adaptive delay equalizing technologies have also been proposed to mitigate this issue. In this paper, these enabling technologies will be reviewed and discussed for the near future high-speed and flexible transport networks.

In this paper, we compare performance of three classes of forward error correction schemes for 40 Gb/s optical transmission systems. The first class is based on the concatenation of RS codes and this is employed in state-of-the-art fiber-optics communication systems. The second class of codes is the turbo product codes with BCH component codes. The application of these codes in optical communication systems was extensively studied by Sab et al., and Mizuochi et al., and they are being considered for the next generation of long-haul systems. The third class of codes is the low-density parity-check (LDPC) codes that have gained great significance over the past decade. Structured binary LDPC codes, nonbinary LDPC codes, and generalized LDPC codes are considered in this paper. A sophisticated simulator that models all major transmission impairments is used to assess the performance of the error correction schemes. The simulation results show that LDPC codes outperform significantly two other classes of codes.

A bandpass optical channel contains four degrees of freedom (DOF) corresponding to two quadrature phase and two quadrature polarization components. If independent information is sent on each of these four components, the spectral efficiency is nearly four times that achieved when utilizing a single component. For a given spectral efficiency, the minimum number of required photons per bit N_p diminishes with increasing number of degrees of freedom (DOF), but the lower limit on N_p, obtained when the spectral efficiency goes to zero, is 0.693 independent of DOF. For the case of an incoherent channel with square law detection and without polarization filters an upper bound to spectral efficiency and a lower bound to the number of photons per bit is obtained by assuming parallel coherent channels, but in which the same signal is transmitted in each channel. An even tighter bound on performance of the incoherent channel is obtained by computing the mutual information for a specific choice of input distribution utilizing a Markov chain approach.

Optical coherent techniques are used to eliminate power fading found in optical subcarrier multiplexed systems. In a double-side band optical subcarrier system with direct detection the signal experiences a periodic power fading that is dependent on the fiber dispersion and subcarrier frequency. This power fading results from interference between the two side-bands following the square-law photodetector. It is shown that the use of an appropriately modulated optical local oscillator to coherently detect the subcarrier channel can eliminate this power fading as well as phase error that gives rise to eye distortion. For homodyne detection an optical local oscillator, centered at the optical carrier, is double-sideband suppressed-carrier (DSB-SC) amplitude modulated by the subcarrier frequency of interest. By independently controlling the phases of the optical local oscillator and the DSB-SC modulation both the phase error and power fading of the detected subcarrier channel can be eliminated. This technique also allows the subcarrier to be selected optically, before the optical-to-electrical conversion.

The opportunity to address special future client interfaces, i.e. IP router interfaces, and to reduce CapEx and OpEx in network domains with highly aggregated traffic are arguments for network operators to insist on the principal option to employ 40G in their backbone network. The fiber infrastructure of most network operators is adequate for a 40G introduction if parameters such as chromatic dispersion, fiber attenuation, nonlinear fiber effects are considered. Already the transition from 2.5Gbit/s to 10Gbit/s per channel the Polarization Modem Dispersion (PMD) for many operators proved to be a limiting factor. The heterogeneous distribution of PMD of cable and fiber segments enabled the operators to install 10 G systems by measuring and selecting the fibers. The migration towards 40G is limited mainly by the PMD of the fiber infrastructure. Again the heterogeneous distribution of PMD values means that only fraction of the possible links are feasible for 40G transmission. To extend the usable part of the infrastructure it is very important to define accurately the PMD limit which is acceptable for 40 G transmission.

Recommendations for Broadband Passive Optical Networks (B-PON) and Gigabit Capable Passive Optical Networks (G-PON) are now reaching maturity in ITU-T. Many vendors are offering conformance with the B-PON series of standards and products with G-PON conformance are emerging. These systems typically operate over a maximum 20 km, 32-way split optical distribution network. The capacity of B-PON ranges from 155/155 Mbit/s symmetrical to 1240/620 Mbit/s asymmetrical transmission. G-PON extends this capacity to 2.4/2.4 Gbit/s and allows more efficient transmission of Ethernet. This paper presents an update on the status of standards on passive optical networks systems with emphasis on recent developments in Q.2/15 of ITU-T.

Small form factor pluggable (SFPs) transceivers have been used to demonstrate that legacy optical networks operating in the metro-core-edge region of the network can be upgraded cost effectively through optical-electrical-optical (OEO) wavelength translation (WT). The WT technique involves detection of the original communication signal, re-timing and retransmission at a new laser wavelength and linewidth. The change is a deliberate one and gives rise to better optical transmission characteristics, resulting in longer system reach. In this paper we present experimental results that demonstrate how WT can be used to upgrade existing fiber systems through improvements in system dispersion or attenuation or by making optical amplification possible.

The need for broadband access is well appreciated by the telecommunications industry, and many types of broadband-access networks have been built. It is widely accepted that wired optical networks is a most promising solution to the broadband-access problem. This paper focuses on passive optical networks (PONs) because PON is a major technology today. We concentrate on the components of PONs. After brief survey of the principle of PON operation, architectures, topologies, protocols and standards, we turn our attention to the PON physical layer. We analyze the structure and elements of PON components and discuss their technical characteristics.

Linearized optical intensity modulator is recognized as one of the building blocks in any analog fiber-optics links systems such as subcarrier multiplexing (SCM) systems, ultra-dense CATV, Radio-over-Fiber (RoF) communications, and other platform access systems. For more than 30 years, the quest for highly linearized optical modulator with SFDR > 130 dB-Hz^2/3 represents a major, on-going technology goal. This invited paper has three-fold objective namely: (1) provide comprehensive overview of the numerous existing linearized optical intensity modulators, (2) introduce a classification of these linearized modulators, and (3) present recent development of new, super-linear (SFDR = 130-140 dB-Hz^2/3) modulator which the authors pioneered. Other features of this new modulator are simple setup, high tolerance and low-cost. Performance results are presented via numerical simulation, its potential applications and limitations are also discussed.

Photonic delay lines (PDLs) are a powerful and important technology for microwave photonics applications and in particular for the implementation of wideband phased array antenna controllers since they solve many of the limitations of electronic controllers, such as large instantaneous bandwidth and tunable signal processing bandwidth. They also offer immunity to electromagnetic interference and electromagnetic pulses. Two of the most critical properties of PDL for advanced phased array antenna control are the fast switching speed and the accurate control of the time delays. In this paper, we propose a novel combined use of semiconductor optical amplifier (SOA) and Optiflex^TM fiber circuit technology for the implementation of PDL architectures. Both technologies are currently quite mature and advanced. SOA technology has been used extensively on optical networks. On the other hand, Optiflex technology has been invented and used for optical backplanes and interconnects. In these PDL architectures, the SOAs act both as ultra-fast switching optical elements and as insertion loss compensators. On the other hand, accurate control of the time delays is achieved through the proper design of OptiFlex circuits. The integration of SOA and OptiFlex to build novel photonic delay line architectures provides small size, low power consumption, and ease in assembly, offering at the same time very good RF performance.

Stimulated Brillouin scattering in fiber is used to realize optical single sideband modulation of the 11GHz radio over fiber (RoF) system. It amplifies the lower sideband of modulated signals and attenuates the upper sideband. Optical single sideband modulation is then achieved which reduces the dispersion caused by traditional double sideband modulation in RoF system. The performance of 11GHz RoF system is then improved. In the experiment, about 2dB SNR gain is attained at 10Mbit/s signal with QPSK modulated format.

Wireless Free Space Optics (FSO) is one of the most promising candidates for future broadband communications, offering transmission rates far beyond possible by RF technology. However, free space wireless optical channel presents far more challenging conditions for signals than typical RF channels, making system availability and throughput a critical issue. Multirate fractal modulation has been considered to address adverse channel conditions in FSO systems. In this paper, channel coding for such a multirate communications scenario is discussed. Specifically, we have used Fountain Codes, a new class of erasure correction codes, in concatenation with an inner Convolutional code. We argue that for a parallel multirate system Fountain coding is a flexible method for receiving data from multiple streams.

This paper investigates the performance of using various non-quadratic adaptive algorithms in the adaptation of a non-linear receiver, coupled with a second-order phase tracking subsystem, for asynchronous DS-CDMA communication system impaired by double-spread multipath channel and Gaussian mixture impulsive noise. These algorithms are the lower order (where the power of the cost function is lower than 2), the least-mean mixed norm (where a mixed-norm function is introduced, which combines the LMS and the LMF functions), and the least mean square-fourth switching (where this algorithm switches between LMS and LMF depending on the value of the error). The non-linear receiver comprises feed-forward filter (FFF), feedback filter (FBF), and an equalizer/second order phase locked loop (PLL). The investigations study the effect of using the proposed algorithms on the performance of the non-linear receiver in terms of the mean-square error (MSE) and bit-error-rate (BER). Computer simulation results indicate that the least-mean mixed proposed receiver's algorithm gives the fastest convergence rate and similar BER performance, in comparison with the NLMS adaptive receiver. Furthermore, extensive computer simulation tests have been carried out to determine the optimum values of the step-size, the power of the cost function, and the adaptation parameter of the proposed algorithms. Results show that the optimum values of the step-size for the lower-order, least-mean square fourth, least-mean mixed norm, and the NLMS algorithms are 5x10^-4, 10^-6, 5x10^-4, and 0.01, respectively. The optimum value of the power of the lower-order algorithm is 1.9 and the optimum value of the adaptation parameter of the least-mean mixed algorithm is 0.9.

In this paper, we proposed two waveband switching algorithms: Minimal Hop Routing (MHP) and Maximal Overlapped Routing (MOP). The differences between them are the weight of link in the layered graph at the waveband and wavelength planes. In MHP algorithm, the weight of links at waveband and wavelength planes are the same, but MOP assigns lower weight to them at the waveband plane than that at the wavelength plane. We conducted extensive simulations with dynamic traffic patterns in the mesh network topology. We evaluated the performance of the proposed algorithms in terms of blocking probability and the number of OXCs port with waveband algorithms of MOP, MHP and RWA algorithm of shortest path routing (SPR). Simulation result shows that waveband algorithm has low blocking probability, and less number of ports is used.

Dispersion compensating discrete Raman amplifier are known to open up new wavelength bands. However there is also the issue of wastage of Raman pump power. The length of dispersion compensating discrete Raman amplifier is decided to minimize the dispersion. Hence significant pump power is wasted. In this paper, a novel design of dispersion compensating Raman/ Two stage EDFA hybrid is reported which recycles the residual pump power from the dispersion compensating Raman and feeds this pump to the second stage of two stage EDFA. The first stage is remotely pumped from another laser source. Using this configuration, we have achieved an extremely large gain bandwidth of 117.5nm from 1582.5-1700nm with a 3dB ripple, has been achieved This amplifier topology while using minimum no of pump sources solves the twin problem of wastage of Raman pump power and providing amplification in U-band. We also performed the simulations of another topology in which a 5m long unpumped EDF was inserted between the first stage and second stage of the EDFA. The backward traveling Amplified Spontaneous Emission (ASE) from the second stage caused the pumping in the unpumped EDF thereby causing signal gain instead of loss. This topology further showed an enhancement in gain of 1-2 dB in the wavelength band of interest (1600-1700nm). The design issue in these topologies is the length of the EDF's. By suitably modeling these lengths, we can obtain appropriate gain profile.

In this paper, a traffic-grooming problem for multi-granularity traffic of SDH/SONET in WDM grooming mesh networks is investigated. Our objective is to improve the throughput of SDH/SONET WDM mesh networks. We propose a heuristics algorithm to solve this problem. The performances of this traffic grooming heuristics algorithm are evaluated in WDM grooming networks. Finally, we presented and compared the simulation results of this methodology in dynamic traffic grooming WDM mesh networks with that of other methodologies.