Optical Communications and Networks

The following sections are included:

• Message from Conference Chairs

• Message from the Technical Program Chairs

• List of Committee Members

• Towards an All-Optical Internet

• Perspectives on Education and Training in Optical Communications

• Micro Photonic Integrated Circuits, MicPIC

• Contents

‘Ethernet over WDM’ is very promising as a next-generation network technology. After some prospective Ethernet-based services are introduced, a novel WDM-based wide-area access network, combined with a regional transport network, is introduced. This network enables the efficient deployment of such services. We show a detailed system configuration of the network and OAM information channels that are effective for it. We then present its management information model in reference to the architectural transport model. This model will be useful in building network management systems.

The recent explosive growth in the Internet technology has made it imperative to have a network which provides a very high bandwidth. Optical networks employing wavelength division multiplexing (WDM) technology are becoming a viable solution to meet this ever-increasing bandwidth demand. Optical burst switching (OBS) networks provide a good solution for a very high traffic influx which is generally bursty in nature. In this, a single control packet is sent for a group of packets which constitute a data burst with a time gap between them to ensure pure optical switching of bursts. Usually bursts that traverse longer paths are more likely to be dropped than those that traverse shorter paths resulting in fairness problem. Fairness means that for all ingress-egress node pairs in a network a burst must have equal likelihood to get through independent of the hop length involved. In this paper, we develop an efficient fairness method to achieve fairness among bursts with different hop counts. The key idea is to collect scheduling statistics by observing the bursts, and based on this information assign different offset time to bursts with different hop counts. Apart from ensuring fairness our method has several attractive features. By using the online statistics, our methods inherently captures the traffic loading patterns and network topological connectivity. It also ensures that the delay experienced by a burst is low and shorter-hop bursts are not over-penalized while improving the performance of longer-hop bursts. We demonstrate the effectiveness of our fairness method through extensive simulation experiments for a number of random networks which differ in the number of nodes and topological connectivity.

This study focuses on the edge router of an optical network. This device works as an interface between the electronic and optical domains : it takes packets coming from client layers and converts them into optical packets to be sent into the optical network.

Arriving variable length client packets fill the optical packet until the latter is full or cannot accommodate the arriving block. The aim of this work is to measure the efficiency with which electronic packets can be carried into optical packets.

Performance measures such as the packetisation efficiency and the mean time of packetisation give good results.

In this paper, we propose an efficient and simple fair queueing algorithm, called Minimum Possible Virtual Start Time Fair Queueing (MPSFQ), which has O(1) complexity for the virtual time computation while it has good delay and fairness properties. The key idea of MPSFQ is that it has an easy system virtual time recalibration method while it follows a rate-proportional property. MPSFQ algorithm recalibrates the system virtual time to the minimum virtual start time among all possible virtual start times of HOL packets in backlogged sessions. We will show our algorithm has good delay and fairness properties by analysis.

In WDM passive star networks(PSN), Two critical collision problems are a data channel collision during packet transmission and a receiver collision during receiving at a receiver node. We propose a new transmission scheduling scheme that avoids two collision problems and provides the differentiated services scheme with guaranteed reservation time. The proposed protocol also introduces the self collision prediction mechanism with a fixed roundtrip delay by using the FDL and uses the multiple control slots mechanism for real time transmission of high priority traffic.

In this paper, we discuss the three GVD compensation schemes, including the post-compensation, pre-compensation and hybrid-compensation schemes with considering PMD. Our analysis is based on the 10Gbit/s and 40Gbit/s NRZ OOK systems, in which we only consider three physical factors, Kerr effect, GVD and PMD. By comparison, we find that, taking the impact of PMD into account, the GVD pre-compensation scheme performs best when the bit rate is 10Gbit/s, but the GVD post-compensation scheme performs best for the case of 40Gbit/s. The results are different from those already reported under the assumption that PMD is neglected. Therefore our researches provide a different insight into the system optimization when the impact of PMD, Kerr effect and GVD are considered altogether.

A simple chromatic dispersion measurement method based on photonic microwave technique is presented in the paper. The dispersion is directly determined from the measured frequency response at a specified modulation frequency. The dispersion measurement range and sensitivity can be flexibly adjusted, without reconfiguring the measurement system. The validity of our method is demonstrated experimentally, and the results show that the dispersion measured by the proposed method agrees well with that measured by the conventional measurement technique.

A simple method, by using an electrooptic intensity modulator as both a mode-locker and a pulse-amplitude-equalizer, for the amplitude equalization of high repetition rate pulses generated from a rational harmonic mode-locked σ-cavity fiber laser was proposed and experimentally demonstrated in this letter.

The well-known modulation phase-shift technique is adapted for field use. This modification uses measurements of rf optical power-modulation amplitudes, from which relative group delays (RGD) are computed. It doesn't require any control or calibration of relative phases in the electrical domain. On account of this simple implementation, one can determine precise values of RGD over a wide range of link losses and in field conditions. A broadband emitting LED transmitter - tunable filter receiver combination is used. With this simple set-up, accurate RGD measurements are possible over un-amplified spans of more than 180 km of non-zero-dispersion-shifted fiber from 1530 to 1560 nm. Measurements through optical amplifiers are also presented.

We demonstrate a new electrically tunable dispersion compensator by electrically tuning the chirp of a uniform fiber Bragg grating (FBG) without center wavelength shift. A distributed on-fiber thin film heater was deposited on the surface of a taper FBG by means of controllable electrical plating to adjust the chirp and hence dispersion of the FBG. A negative thermal-expansion coefficient (NTEC) ceramic was used to fix the central wavelength in the tuning. With an applied electrical power smaller than 0.68 W, the dispersion tuning range from−178 ps/nm to −302 ps/nm with a dispersion ripple of less than 30 ps was achieved. The corresponding central wavelength shift was as small as 0.173 nm.

Transmission performance of 40Gb/s CRZ format with low-duty-cycle was studied numerically. It was found that, pulse prechirping will enhance the effect of second-order dispersion in 40Gb/s longhaul transmission. So dispersion slope compensation is very important even for single channel ultra-high-speed transmission.

This paper presents a new algorithm that dynamically allocate restorable bandwidth guaranteed paths in IP over DWDM network. This paper tackles the bandwidth efficiency by considering sharing of the bandwidth on the backup lightpaths. The amount of sharing that can be achieved is related to the information available to the algorithm. Three information scenarios are considered here: complete information scenario, partial information scenario and no information scenario. Simulation results show that the developed algorithm with a partial information scenario which uses only aggregated and not per-path information performs very well compared with the complete information scenario.

This paper provides an analytical basis for comparison of reservation policies in Optical Burst Switching (OBS) networks. In particular, the Just-Enough-Time (JET) and Burst Segmentation policies are analysed. We first compare these two policies using single link models and then we extend these models to a network scenario using a reduced load fixed point approximation model to evaluate blocking probabilities. We use this analysis framework to provide insights into the significant advantages provided by the Burst Segmentation policy for OBS networks, taking into account network related effects which cannot be studied by single link models alone.

Distributed wavelength reservation protocols are employed to dynamically set up and tear down lightpaths in WDM networks. In this paper, we develop an analytical model, which can analyze the blocking performance of backward reservation protocols (BRP) under both single and multi-fiber networks. The effects of different wavelength selection polices are also investigated.

This paper describes the subtle design and implementation issues of a reliable and integrated GMPLS-based control plane. Although there are currently standard-track drafts that specify the protocols for controlling IP/optical network, there is limited work taking into account the applicability of these protocols to seamlessly and reliably control such a network. This paper is to study and provide a seamless and robust architecture for the different protocols to inter-operate, and to provide a tightly coupled, integrated and reliable control in IP/optical networks.

This investigation evaluated the MB810 and the 8B/10B line codes used in optical Ethernet systems. Compared with the 8B/10B, the MB810 has the advantage of a sharp shape of the optical carrier frequency component and Nyquist bandwidth. These advantages make it possible to compensate for optical transmission impairments such as dispersion and distortion, and to improve the signal-detection capability at the optical receiver. We compared the performance of the MB810 and 8B/10B in an optical transmission system based on a application of 10Gigabit Ethernet. The experimental result showed that the power budget of the MB810 1.1dB better than that of the 8B/10B.

Superstructure fiber Bragg gratings (SSFBG), in which the amplitude and phase in grating corrugation are controlled, can realize versatile functions for DWDM systems. We review our technique to fabricate densely-spaced SSFBG, multiple phase-shift (MPS) technique.

We propose a new design of a gain enhanced EDFA for WDM applications by inserting a wavelength division multiplexing (WDM) filter (with periodic multiple pass band wavelengths coinciding with the ITU channel grid) and an isolator within the EDFA. It is shown that by placing the filter and isolator combination at a distance from the input end of about 30 % of the total length of the EDF, it is possible to achieve about 10 dB improvement in gain and a reduction of about 2 dB in noise figure compared to the case without the filter and isolator. Such an improved EDFA design should be of interest in fiber optic communication systems employing wavelength division multiplexing.

A double-pass erbium-doped fiber amplifier (DP-EDFA) configuration for 980 nm pump is demonstrated. This amplifier provided gain up to 40 dB in the conventional band C-band. In this configuration the signals are looped back into the EDF with the aid of optical circulators and amplified twice by the EDF. Double-passing the signal through the EDF produces a net enhancement. The DP-EDFA gain is 6.5 dB higher than the standard single-pass fiber amplifier gain but at the cost of noise figure (NF) increasing by 1.4 dB.

Amplification characteristics of regenerative erbium-doped fibre amplifier operating just below the oscillation threshold are presented. Resonant amplification is preserved only for small drive powers. At the higher drive powers, resonant amplification characteristics vanish and the amplification bandwidth increases.

A new gain improvement technique for long wavelength band erbium-doped fiber amplifier (L-band EDFA) is presented. It uses an unpumped erbium-doped fiber (EDF) in a double pass configuration in order to effectively suppress and completely exploit the backward amplified spontaneous emission (ASE). Compared to a conventional single pass amplifier, the gain coefficient can be improved from 1.5dB/mW to 3.1dB/mW without much noise figure penalties. The L-band EDFA can also reduce the pump power necessary to achieved a specified gain by about half however at the cost of noise figure penalty of 1.6dB. The optimum length for an unpumped EDF is experimentally measured to be 3m for both single pass and double pass system. By adding 3m of unpumped EDF, the maximum gain enhancements of 2.7dB and 5.7dB were obtained for single pass and double pass system at wavelengths of 1576nm and 1600nm, respectively.

This paper describes multi-layer traffic engineering and signaling technologies in the photonic-GMPLS-router (Hikari-Router) network. Multi-layer traffic engineering, which yields the dynamic cooperation of IP and photonic layers, is described with the goal of providing IP services cost-effectively. The establishment of photonic cut-through paths is triggered when the Layer 3 traffic loads become excessive which better uses the optical layer (L1) resources. To realize multilayer traffic engineering, we propose an OSPF extension that advertises both the number of total wavelengths and the number of unreserved wavelengths, and an RSVP-TE extension that minimizes the number of wavelength conversions needed. Networks are reconfigured by the proposed scheme so as to utilize the network resources cost-effectively.

We investigate the issues related to the merging of a WDM ring fiber-radio backbone incorporating wavelength interleaving where each base station drives a sectorized antenna interface and a standard WDM infrastructure. The investigations show that by re-allocating the interleaved WDM channels to fit within a 100 GHz block, it enables the mm-wave fiber-radio system with sectorized antenna interface to integrate easily within standard WDM systems. In addition, such channel allocation also improves the overall network capacity.

This paper describes an experimental study of the Resilient Packet Ring (RPR) media access control (MAC) technology which is optimized for IP traffic in a Metropolitan Area Network (MAN). The study involved the deployment and trials of an RPR test-bed encompassing a public optical fiber infrastructure using Cisco Systems' Dynamic Packet Transport (DPT) Ring Technology – a pre-standard RPR implementation. The work described here focused on a number of important RPR protocol features that are vital to its future success as a MAN network technology. Related investigation works on RPR/DPT have so far been done through simulation studies only.

In this paper we propose the concept of optical subnets in WDM rings. By strategically placing few WDM optical add-drop multiplexers (OADMs), interleaved with many passive nodes, this hybrid ring can perform as well as a conventional OADM ring. The section of the ring between two given OADMs is called an optical subnet, while the OADM itself is called a wavelength gateway. We show analytically as well as through simulation that the capacity is still almost comparable to the case with only WDM OADMs, while the cost involved is greatly reduced. The architecture termed as optical subnets has salient features of low cost and good throughput.

We propose a novel bidirectional optical crossconnect (BOXC) using fiber Bragg gratings (FBGs) and optical circulators. Dynamic control for wavelength routing is achieved by employing tunable FBGs. Because only a few optical circulators are used, the proposed BOXC is more cost-effective than previously reported BOXCs. Our preliminary study shows that the proposed BOXC has potentially small insertion loss and low crosstalk. It requires a small tuning range equal to wavelength channel spacing for tunable FBGs. We show that a large BOXC can be built based on the Benes network structure and present a method to reduce the complexity.

Microstructured optical fibres have increased degrees of flexibility that allow tailoring of optical properties for various applications. Three different algorithms for calculating mode properties and structural or confinement loss in these fibres are compared. A number of biologically-inspired approaches to optimising the fibre design (hole size, shape and arrangement) within the context and constraints of a particular fabrication technology are presented.

A modified design of segmented cladding fiber is presented. The fiber can provide effective single-mode operation in the wavelength range 1000 – 1700 nm with a core diameter as large as 20 μm. The mode area of the fiber is 4 times larger than that of a conventional single-mode fiber and can thus suppress the nonlinear effects effectively. The fiber should find applications in DWDM systems, fiber amplifiers, and lasers.

We have developed a new coupled wave equation to analyze the pulse switching dynamics of a two-unequal-core fiber coupler under different radii ratio. The analysis is based on a flexible and simple method called Fourier series analysis technique (FSAT).

Enhanced photosensitivity has been observed in hydrogen-loaded tin-phosphosilicate fibres by using a 248 nm excimer laser. Isothermal measurements up to 860 K demonstrated significant advantages over fibre gratings written in conventional H-loaded fibres. Gratings written in this fibre require a considerably shorter post-fabrication thermal annealing in order to satisfy the stability requirements of telecom components.

By changing the construction of NOLM, it can possess many different functions. In this paper we analyzed the effects of input power on the transfer characteristic, and the effect of SMF or DSF length on transfer characteristic.

This paper presents innovative concepts related to the realization of ultra-fast optical switches, which are to be used in next-generation optical networks. Experimental measurements, which characterize the performance and reliability of such switches, are presented.

This paper introduces a prearranged delay asymmetric structure of optical burst switching. In this paper, system logical structure, design of concomitant spare exports and contention-resolving mechanism are investigated in detail.

An arrayed waveguide grating (AWG) router is optimised for use in an AWG-based optical add-drop multiplexer (OADM) of a WDM transparent optical network with 10 Gbit/s per channel and channel spacing of 25 GHz. Two different techniques to flatten the AWG frequency response are investigated. It is shown, by simulation, that the optimised AWG router of an OADM with fold-back configuration should have bandwidth similar to the channel spacing, good isolation to adjacent channels and low or moderate additional losses. Numerical results reveal that AWG routers with frequency response between the flat and conventional AWG responses are the best candidates.

Optical packet switching is a promising technology for future all-optical networks. The contention problem in an optical packet switch can not be easily handled because the optical random access memories are not available with present technology. Fiber delay lines and tunable wavelength converters have been investigated to emulate the optical buffering, however, the hardware volume of the optical components limits the buffer capacity. In this paper, we have proposed an efficient method for contention resolution with very limited amount of wavelength converters and fiber delay lines. The potential of this method is analyzed by simulation.

We propose a new reconfigurable add-drop multiplexer using tunable fiber Bragg gratings (FBGs) and optical circulators. The new OADM allows dropping several fixed wavelength channels and at least one tunable wavelength channel. By properly assigning wavelengths channels that are add-dropped at an OADM node, the tuning range required for tunable FBGs can be made as small as the channel spacing. The proposed OADMs have potentially low insertion loss. The operation and performance of this OADM are experimentally investigated.

We investigate the inband crosstalk generated at the synchronization module of photonic packet switches and found that the inband crosstalk increases in proportion to the number of cascaded optical switches and delay lines. We show that the impact of this crosstalk on the BER is significant when the number of cascaded switches and delay lines is greater than 10.

We report recent results on all-optical wavelength conversion using broad-band orthogonal pumps four-wave mixing. The different approaches are developed based on a semiconductor optical amplifier placed in a fiber-ring cavity. Using a single external pump laser, a conversion efficiency with a variation smaller than 3 dB has been achieved over a 40-nm operating range. We further demonstrate polarization-insensitive widely tunable wavelength conversion with a polarization diversity scheme. The power penalty is determined to be 3 dB in a 10 Gb/s bit-error-rate measurement. In a modified approach employing a double-ring SOA fiber cavity, all external pumps have been successfully eliminated.

A review of the characterisation of optoelectronic guided-wave devices using the modal solution, junction analysis and beam propagation methods based on the numerically efficient finite element method is presented. Numerically simulated results for various photonic devices such as optical modulators, spot-size converters, compact bends, power splitters, and Bragg gratings are presented.

A technique to implement an optical multichannel flattop filter with simultaneous complementary outputs is proposed and demonstrated. Using a bipolar delay line structure based on cascaded high birefringence (Hi-Bi) fibres, flattened optical passbands with ripples <0.1dB and sidelobe suppression >30dB is achieved. Due to the single-line cascaded Hi-Bi fibre configuration, the filter exhibits stable wavelength characteristics with low insertion loss. The configuration is not limited to optical domain applications and can be easily adapted as a microwave transversal filter that overcomes limitations imposed by the coherence of the source. Microwave flattop bandpass filtering is further demonstrated based on this technique.

Injection locking technique has been commonly studied and can be used to achieve high power stable laser. This work is done to investigate such a phenomenon in an erbium-doped fiber laser (EDFL). The injection locking phenomenon occurred at high injected power. A stable output power has been achieved under locking condition.

Twin cavity multiwavelength Brillouin/erbium fiber laser (BEFL) demonstrates multiwavelength output in long wavelength band (L-band) region. The system consists of two identical EDFLs that share a length of single-mode fiber to produce interdependently alternate bi-directional multiple Stokes-shift frequencies. The twin cavity BEFL generates a 22GHz comb of 2 × 11 output channels at 1600nm that has potential application in future wavelength division multiplexing (WDM) system.

The Management Information Base (MIB) provides network-level configuration information for the management and control. Numerous MIBs have been defined by IETF for the IP over WDM development. However, there is no MIB specified for the management and control of optical devices such as Optical Add-drop Multiplexer (OADM) and Optical CrossConnect (OXC). This paper intends to fill in the gap by presenting the development of a MIB for OADM/OXC optical devices. The logical basis of the MIB is provided by the development of a network-level information model using TINA-C NRIM specifications. The information model is made to portray managed object classes relevant to network resource management, leading to the construction of the MIB for OADM/OXC using the SNMP SMIv2 standard. The MIBs have been used to develop a network management system prototype for a WDM Ring network with appropriate GUIs using Java and the AdventNet SNMP APIs. The usefulness of the MIBs is demonstrated by developing the capabilities for controlling the devices and managing the faults.

As control plane and signaling seems to be a critical part in next generation networks, this paper proposes an integrated control plane (ICP) for IP enabled optical networks. Signaling procedure for ensuring dynamic end-to-end lightpath setup has also been presented. Path provisioning comprises a string of operations like service and neighbor discovery, route computation, signaling requests, and path setup.

A novel heuristic algorithm for multi-layer traffic engineering is proposed in this paper, with the objective of minimizing network congestion. The basic idea of this algorithm is to optimize network resource utilization by lowering the cost of setting up connections for fulfilling the traffic demands. This algorithm can be applied to arbitrary network topologies with various OXC structures. Another advantage of this algorithm is that it is designed for nonbifurcated traffic routing, making it more suitable for practical implementation. Computational results show that it can achieve congestion close to the LP lower bound.

This work investigates the behavior of a reservation mechanism known as JET, Just Enough Time, highly promising in WDM networks adopting the optical burst switching (OBS) paradigm. In this article we assess the performance, in terms of burst blocking probability, that an optical cross-connect achieves when employing JET in conjunction with the Multiprotocol Lambda Switching in the control plane on top of the optical layer, and compare it with the Labeled Optical Burst Switching, LOBS, solution. The burst blocking probabilities of three service classes with different QoS requirements are determined, demonstrating that the LOBS alternative guarantees a remarkable improvement with respect to the Multiprotocol Lambda Switching - OBS combination.

In this paper we examine a scheduling algorithm for burst switched networks, based on a policy to aggregate bursts at the edge routers of a network. We also consider aggregation as well as scheduling of bursts from a QoS perspective. We guarantee delay based QoS to individual packets for end-to-end flows. We also describe a new aggregation process that assembles bursts from packets and determines both the appropriate size of the burst as well as the optimum time at which a control packet needs to be sent to reserve bandwidth.

Analytical formulas are presented to describe the transmission characteristics of a long-period fibre grating (LPFG) in response to changes in temperature, strain, pressure, external refractive index, etc. The effects that arise from the modal dispersion in the fibre are highlighted. The results are useful for the evaluation and the design of LPFG-based devices and sensors.

Theoretical investigation of fiber Bragg grating (FBG) under two cases of lateral compression has been carried out. Lateral pressure is applied over two different angles of 20° and 100° along the circumference of the fiber. The shift in the center wavelength of FBG with respect to the radial pressure is analysed. Pressure sensitivity is more than 4 times bigger if the pressure is distributed over a large angle. The strain-induced birefringence created due to the refractive index difference in the radial directions has also been presented. The simulation results show that distributing the pressure can reduce the birefringence effects.

A cantilever beam and fiber Bragg grating is used to measure acceleration. The cantilever induces strain on the grating resulting in a Bragg wavelength modification that is subsequently detected. The output signal is insensitive to temperature variations and for a temperature change from −20°C to 40 °C, the output signal fluctuated less than 5% without any temperature compensation schemes. Because the sensor does not utilize expensive and complex demodulation techniques it is potentially inexpensive. For the experimental system a linear output range of 8 g could be detected.

Strain and temperature response of polymer fiber gratings have been measured and analyzed. A rather large tuning range with excellent reversibility is obtained by simple tensile strain on polymer fiber Bragg gratings. Furthermore, polymer fiber Bragg gratings show much higher temperature sensitivity compared with silica fiber Bragg gratings. Very high linear response of Bragg wavelength to temperature with no hysteresis effect is also revealed. Because of these two linear responses of polymer fiber gratings, its future application in strain and temperature sensing has also been proposed.

We show that solitons can be trapped in a fibre Bragg grating by a local defect provided that the soliton energy and defect strength are within a certain range of values.

The use of fibre-fed millimetre-wave radio systems seems to attract the attention for future broadband services, such as Intelligent Vehicle Highway Systems, high speed Wireless Local Area Networks, Interactive Multimedia Services, and Mobile Broadband Services. Optical fibre and millimetre-wave radio are capable of supporting the large bandwidth requirements of these systems, and on the other hand the low attenuation of the optical fibre provides the best method for the distribution of millimetre-waves at present…

A scheme for subcarrier-multiplexed fiber- optic mm-wave radio is proposed in this paper. Using an overdriven phase modulator, higher order optical sidebands can be generated. Three coherent lightwaves are produced from three laser diodes injection-locked to the optical carrier and the two optical modulation sidebands. The two laser diodes locked to the upper optical sidebands are bias current modulated by a group of subcarriers. When these waves are mixed in a PD, two mm-waves are obtained which contain the same amplitude modulations that are present in two of the lightwaves.

A co-located dual DFB fibre laser configuration is realized for optical generation of a millimetre wave signal. Dual polarization mode operation is obtained and the heterodyne signal measured at ~32.5GHz has a linewidth of < 1kHz. The two operating laser modes, originate from co-located cavities, exhibit high common-mode suppression of phase noise to produce a beat signal of high spectral purity. No complex feedback scheme is required and the beat frequency achievable is determined by grating parameters only. Dual mode operation with wavelength spacing measured ~0.75nm is also demonstrated.

The performance of an optical wavelength division multiplexed system is analyzed including the effect of four wave mixing (FWM) in presence of chromatic dispersion. FWM process in the single mode fiber as well as in the fiber amplifier is taken into account. It is observed that the efficiency of FWM generation can be greatly reduced by simultaneously making the spacing between channels unequal and by employing nonuniform chromatic dispersion along the fibers in the communication link. The power limitations are evaluated under various conditions of system parameters. The power depletion is compared under different channel spacing schemes.

A gain-clamped fiber Raman amplifier is proposed and builds up. The result shows that gain fluctuation will be champed to a lower level with the dynamic gain control. And the dynamic gain control method is also fit for the Raman amplifiers with multi-wavelength pump.

This paper investigates launch power management for DWDM transmission systems that use two stage lumped amplification surrounding dispersion compensating fiber (DCF). The impact of modulation format and fiber type on optical launch power is investigated. The results show that for all modulation formats and transmission fiber types, launch power into DCF should be much lower than into transmission fiber, to maintain high optical signal to noise ratio (OSNR). However, the values of the optimum launch powers vary with the modulation format and transmission fiber type.

The dramatic reduction of transmission loss in optical fibers coupled with equally important developments in the area of light sources and detectors has brought about a phenomenal growth of the fiber optic industry during the past twenty five years. Recent advances in the area of zero dispersion operation, dispersion compensation, rare earth doped fiber amplifiers and soliton communication systems have provided communication engineers with almost limitless bandwidth operating over millions of kilometers. Some of the recent developments in this area, including devices like Fiber Gratings will be briefly described in this lecture.

Note from Publisher: This article contains the abstract only.

A novel structure of external cavity semiconductor lasers with suppressed thermally induced wavelength shift and mode hopping is proposed in this paper. The calculation results have shown that the wavelength stability with temperature is dramatically improved and the side mode suppression ratio (SMSR) remains a higher value (>35dB) when temperature change from 20°C to 50°C.

In this paper, we propose a novel dynamic transmission-line-laser-based model of the multiple-phase-shift distributed feedback (DFB) lasers. The dynamic and static characteristics of three phase shift (3PS) DFB structures are analyzed based on this model. The longitudinal carrier density distribution of quarterly wavelength shifted (QWS) DFB LDs and 3PS DFB LDs are compared. Our results show that 3PS DFB structure has a more uniform longitudinal carrier density distribution. Therefore, such a structure reduces the spatial hole burning effect and hence strengthens the single-longitudinal-mode stability.

An optimization design for broadband gain-flattened multi-wavelength backward-pumped fiber Raman amplifiers (MW-FRAs) was presented. We proposed a new crossover method (Bounded Gaussian Distributed Crossover) in real-coded Genetic Algorithm for the optimization design of broadband gain-flattened MW-FRAs. By applied it to the simplified inverse design problem, we achieved optimal designs with gain ripple ranged from 0.72% to 12.55% (which corresponds to 15 pumps∼3 pumps). Furthermore, we also gave the partial optimization design when the wavelengths of pumps are not ideal or fixed.

The authors have demonstrated a dual-pumped tunable Raman fiber laser. Tuning of the laser is done by adjusting a tunable bandpass filter that yields a range of about 45 nm from 1520 to 1565 nm and with peak power output of over 11 mW. Two 1.2 W fiber lasers at 1455 and 1480 nm pump the laser.

In recent years, optical wireless communications has emerged as a suitable alternative to radio frequency transmission for short-range indoor data communications since infrared offers several advantages over radio. To utilize the advantages of diffuse IR wireless system such as very high data rate and security over the RF system in an indoor environment we need to exploit the ubiquitous computing and smart sensing mechanism which could be used to mitigate some of the drawbacks of diffuse IR such as different background light strengths at different places, in which case at some places BER gets deteriorated or receiver fails to detect the signal due to its limited sensitivity. In this paper a scheme for tackling the ambient or background lights of different strengths in different locations in a typical indoor space using smart and aware environment has been discussed.

Multimedia applications present new challenges to the current networking technology. One of them is Quality of Service (QoS) requirement. Due to the increasing dynamics of traffic introduced by multimedia applications, dynamic and flexible QoS (Quality of Service) control is needed to ensure both QoS satisfaction and resource efficiency. Because integrated service networks are designed to support a wide range of traffic classes, which have different service requirements, developing metrics to evaluate them is complex. In this paper there are presented the results of some tests performed on Cisco routers in a quality of service environment.

The self-phase modulation (SPM) effects of nonreturn-to-zero (NRZ), return-to-zero (RZ) and carrier-suppressed return-to-zero (CS-RZ) signal formats are experimentally compared for 10Gbit/s single-channel transmission in a SMF dispersion managed system at nearly zero residual dispersion. The results show that NRZ format is seriously affected by SPM, while RZ and CS-RZ formats have excellent tolerance. With the input power increasing further and due to the SPM effects, RZ and CS-RZ experience a recovery after the deterioration.

We present RWA algorithms for a new class of traffic model where, besides the source, destination and size (number of lightpaths) of the traffic demands, their set-up and tear-down dates are known. They are called Scheduled Lightpath Demands (SLDs). We model the RWA problem as a spatio-temporal combinatorial optimization problem and provide two solution algorithms. The time disjointness that could exist among SLDs is taken into account in order to maximize the utilization of resources and hence, minimize the amount of globally required resources. We compare our algorithms to an online RWA algorithm and show that taking into account the time disjointness of demands can lead to a gain of resources of 20% in average.

Optical multicast is usually modeled as a two-step process: multicast routing and wavelength assignment. This paper proposes a new multi-layer model, in which the WDM network is represented as a multilayer wavelength structure, each of which has finite capabilities, to integrate routing and wavelength assignment. The algorithm makes use of the light splitting and wavelength conversion capabilities simultaneously in a deterministic manner, avoiding the re-routing procedure. It handles the finite light splitting capability, thus providing a general solution to the problem. Some simulations results are presented in the paper.

To provide the high capacity transport network facilities, with the development of the WDM technology the optical networks containing multiple OXCs are deployed in carrier networks. Furthermore, a consensus is emerging in the industry on utilizing an IP-centric control plane within optical networks to support dynamic provisioning and restoration of optical layer connection [1]. So, in this paper, we propose the fast incremental line restoration method: how to discover fast the secondary path through the incremental examination of the available resource. The proposed scheme is able to use the resource efficiently. And it is more appropriate for the strict route specified all the exact nodes explicitly rather than the loose or abstract route since the traffic information of the previous path can be mostly kept.

We discuss an all-optical flip-flop memory and a wavelength converter based on nonlinear polarization rotation in a semiconductor optical amplifier. The flip-flop is realized from two coupled polarization switches. The concept is explained and experimental results are presented that demonstrate that a contrast ratio of over 20 dB and a switching power of less than −3 dBm can be obtained. The all-optical wavelength converter employs a single polarization switch. We show that error-free inverted and non-inverted wavelength conversion can be realized at a data-rate of 10 Gbit/s. Both optical signal-processing functionalities can be utilized in all-optical packet switches.

The widespread use of DWDM technology has presented service providers with a new challenge: how to manage increasing optical channels (OChs) to provide fast, reliable services to their end customers cost effectively. OCh-level operation, administration and maintenance (OAM) function in optical are important to effective management, and digital wrapper can provide these functions.

In this paper, we propose an ITU-T Rec. Draft G.709 and G.798 based overhead processing model for OCh-level OAM.

We generated millimeter-wave (MMW) signals for fiber-radio systems. The frequency separation between two optical signals was approximately 30-50 GHz. The generated millimeter-wave signals were transmitted through a millimeter waveguide and detected. The spectral linewidth of the millimeter- wave was 2 kHz, and power fluctuation of the millimeter-wave was less than 2 dB over a 30-50 GHz range. This method can reduce the loss of millimeter- wave power and obtain a wider bandwidth compared with a detection method using a photodiode.

Two failure monitoring methods, leaking lightwave monitoring and extra traffic monitoring, are proposed for monitoring failure of protection line in WDM ring networks in this paper, and which can be used both in two fiber unidirectional ring networks and four fiber bidirectional ring networks. The performance of the two monitoring methods are analyzed and compared in detail. Moreover application condition of the two monitoring methods is analyzed and advised.

We propose a new method for monitoring optical signal quality. The new method is based on sub-harmonic sampling technique and is experimentally verified. With the new method, the amplitude histogram of the optical signal is acquired by sub-harmonically synchronous sampling. We demonstrate experimentally that the Q-factor obtained from the amplitude histogram coincides very well with the Q-factor deduced from the bit-error-rate that is directly measured by an bit-error-rate tester. The new method does not require high-speed sampling circuits, and hence it is very cost effective for monitoring high-speed optical signals that have different bit rates and coding formats.

This paper explains the important conceptual and technical differences between the method of p-cycles and two other recent advances involving a cyclic orientation to protection. These are enhanced rings and cycle double covers. The most fundamental difference that is unique to p-cycles is the aspect of straddling span failure protection. This enables mesh-like efficiency levels at well under 100% redundancy. In contrast enhanced rings and advanced cycle cover methods are both seeking to reduce span overlaps in what is otherwise a purely ring-like logical paradigm in which 100% redundancy remains the best that can possibly be achieved.

A single fiber cut can potentially influence a total of 1.6 Tb/s of traffic [1] and is surprisingly frequent and results in significant loss of traffic [2]. The restoration time in the optical networks consisted of optical cross connector OXCs is slower than that of the SONET/SDH networks.

For the fast link-based protection several mechanisms for the existing MPLS network are currently under discussion in IETF. However, there are some problems to apply the mechanisms to the optical network.

So, we describe the deficiencies of the existing fast re-route mechanism and then propose the fast link restoration mechanism using the extensions of LMP for the optical network and simulate the restoration time performance. The key characteristics of the proposed mechanism are the transmission of the main LSP label for the back up path by the LMP ChannelStatus message.

OXC (Optical Cross-Connect) system is on indispensable element to realize the all-optical mesh networks by supporting the increase of link channels and expansion of an OTN (optical transport network) capacity at low expense after investing initial equipment. This paper proposes a design of network restoration and connection management for OXC system based on distributed object-oriented technology. We describe the functions of NPRS, which handles UNI/NNI signaling of high layer client and manages network connection. This design of network connection and restoration can be applied in not only the layered-centralized network management but also the distributed network management.

Wavelength division multiplexing (WDM) networks have become a viable solution to meet the increasing bandwidth demand. These networks carry messages in the optical domain at data rates of several Gb/s. Such high data rates make it necessary for using efficient restoration protocols in an event of a failure. This paper addresses the problem of dynamic path-based restoration in WDM optical networks. Our objective is to develop a protocol that ensures rapid recovery while keeping the restoration ratio high. We propose two distributed dynamic restoration protocols called disjoint path-wavelength grouping protocol and disjoint weighted path-wavelength grouping protocol. Both these protocols group all the candidate backup paths of the failed paths into path-disjoint groups before assigning wavelengths to them. This allows the failed paths to search different sets of wavelengths on various candidate backup paths in parallel without any reservation conflicts. Also, it does not require any reservation retries, thus resulting in rapid restoration. The grouping method attempts to restore increased number of failed paths by allowing increased number of wavelengths to be searched on a path. The performance of the proposed protocol is verified through extensive simulation experiments.

In the modern business environment for optical mesh networking, there is considerable interest in being able to support a range of service classes, and to charge accordingly. In this work we consider the capacity design problem for different combinations of four protection priority classes in a span-restorable mesh network. Most significantly we have test-case findings that 15% to 30% of all demand can be in the gold class and enjoy 100% restorability, solely through preemption of economy class service capacity. This indicates the potential to design and operate mesh-based re-storable networks which have no spare capacity at all in the conventional sense: all capacity is bearing service of some paying type. The methods provided could be the basis for future mesh-based optical transport networks that simultaneously support multiple QoP services with optimal capacity.

In this presentation, we report on our recent development of erasable three-dimensional bit optical data storage, the formation of micro-cavities and micro-fabrication of waveguide structures in photorefractive polymers. These devices are important for optical communications and networks.

One of the most critical elements within the framework of wavelength division multiplexing optical network is the optical cross-connect. The core of an optical cross-connect is the optical switch capable of switching a large number of input and output fibers. In this paper, we report the design of a polarization independent liquid crystal optical switch. It is pointed out that liquid crystal spatial light modulator technology is a very competitive technology to realize a large-scale reconfigurable optical switch.

Chalcogenide optical glass is an interesting material, for both fibre and planar technologies, as it offers possibilities for a wide array of devices suitable for use in both non-linear applications and as IR sources. Channel waveguide structures were directly written into bulk gallium lanthanum sulphide glass via above-bandgap (λ=244nm) illumination. Focused fluence of 1.5 – 150 J/cm² from a UV-laser was applied, inducing giant-photocompaction and photochemical changes. A typical channel waveguide had attenuation < 0.3 dB cm⁻¹ and was spatially single-mode at 1 μm. The damage threshold for fabricating such devices in this glass system is also discussed.

Optical photodiode module structure has been studied for high coupling efficiency and easy packaging. This paper proposes the optical photodiode module structure with focusing grating outcoupler to accomplish high coupling efficiency and easy packaging. A focusing grating outcoupler makes a small spot size beam at free space, so the light that passes through the optical fiber is gone to a optical photodiode chip with a high coupling efficiency by focusing grating outcoupler. Because the optical photodiode module structure proposed this paper has less sensitivity about the fiber position change that maybe happened in the packaging process, we can have a easy packaging process.

Numerical simulation about pedestal reduction and chirp elimination of high-order compressed solitons by using SOA and tunable filter was demonstrated. Under certain conditions, the pedestals and chirp can be successfully eliminated without broadenning the pulse.

This paper describes the architecture and operation of a Very high-Speed Traffic Aggregator (VISTA) for an Optical Network. The unique feature of the traffic aggregator is that it works in conjunction with the GMPLS control plane of the integrated electro-optical network. The emphasis of our work is on the access to the optical network, at the level of the edge router.

Novel distributed medium access protocols for a time-slotted WDM metro ring employing optical packet switching and supporting variable-length data packets such as IP data-grams directly in the optical layer are proposed in this paper. A scalable control-channel based multi-hop WDM ring is considered in which each node is equipped with three transmitters and two £xed-tuned receivers. The new access control scheme relies on a frame-based slot reservation strategy including reservation of successive slots for data packets longer than the given slot size, and immediate access for packets shorter than the slot length.

This paper presents the performance of distributed-controlled dynamic wavelength-conversion (DDWC) optical networks based on our simple route and wavelength assignment (RWA) policy. A source node first chooses an optical-path route by using aggregated optical link-state information, then sends a signaling message on the selected route, after which wavelength assignment is performed. If wavelength conversion is needed at a transit node, the wavelength is adaptively converted by using trunk-type wavelength converters. We show via simulations that the performance of the DDWC network with our simple RWA policy emulates that of the WC network with small number of trunk-type wavelength converters.

Fast lightpath creation is very important for ASON (automatically switched optical network), but the traditional protocols such as RSVP-TE or CR-LDP work in a serial style and can't support that service well. This paper proposals a new simple parallel signaling protocol for fast lightpath creation in ASON. The most time-consuming step in lightpath creation is processed in a parallel style and the other operations still work serially as the traditional protocols.The signaling procedure is discussed in detail and the formal description of the protocol is given. An experiment is also setup to test the protocol and the result is compared with hop by hop signaling protocols.

In this article, an overview of IP network development is given. Architectures of Ethernet switch, software-based router, switch router, LSR, and GMPLS-based OXC are presented and compared. The discussion is made from the implementation perspective of network equipment, especially the GMPLS-based OXC system rather than the detail of algorithms or protocols. Arising from the discussion, guidelines are presented for the implementation of GMPLS based OXC system. In particular, it is recommended that GMPLS-based system should support Label switching and IP Layer forwarding for its data traffic over control channels.

We describe our recent activities for developing a large-scale 3-D MEMS optical switching fabric. The switches, consisting of MEMS 2-axis tilt mirror arrays and free-space optical interconnections, have great potential in the construction of large-scale switching fabrics with more than 1000 ports.

This paper mainly describes the newly proposed MEMS mirror arrays, which consist of single-crystal silicon mirrors supported by high-aspect ratio torsion springs and low-cost, highly accurate two-dimensional optical collimator arrays. A prototype switch having 64ch optical fiber I/O provides a low coupling loss of 3.2 dB and switching time of 14 msec.

A new process has been presented to fabricate single crystal silicon (silicon) micro mirror switching devices. Silicon torsion and flexural springs have been integrated with a mirror plate for angular displacement amplification. Mirror switching is achieved with a bimorph thermal actuator, which is insensitive to environmental temperature changes and fabricated using surface micro machining layers. The release process combines silicon etching in KOH with sacrificial etching of oxide from backside. The design offers protection for the structure layers from both front and backside. Device demonstrates an angular amplification concept. An angular displacement of 15 degree was achieved.

A low-voltage variable optical attenuator has been developed for applications in dense wavelength division multiplexing systems. The attenuator is fabricated by silicon surface micromachining technology and is then manually assembled and integrated with two single mode optical fibers that act as optical input and output. A 40 × 40 μm² mirror coated with gold is driven by a proprietary drawbridge structure to cut partially into the light path between two fibers, enabling the attenuation. The attenuator has a dimension of 0.6 × 1 mm² excluding the fibers.

The attenuator has achieved 1.5 dB insertion loss, 45 dB dynamic range and 37 ms response time, and requires only 8 V driving voltage. The wavelength dependence loss is 0.9 dB over C-band (1528 – 1561 nm). The low driving voltage makes it possible to integrate IC control circuits onto the same chip. An array of the VOA can be fabricated to equalize multiple wavelength channels of a DWDM system.

Tunable lasers have wide applications in DWDM systems to save inventory cost and to improve the optical network functionalities. The Microelectromechanical Systems (MEMS) technology has shown strong promise to miniaturize the conventional mechanical tunable lasers with adding merits of high compactness, high speed, batch production and so on. In this paper, external cavity tunable diode lasers using MEMS movable mirrors and rotary gratings as the external reflectors are presented. One tunable laser of 2 mm × 1.5 mm is formed by integration of a surface-micromachined 3D mirror with a diode laser and an optical fiber. A wavelength tuning range of 16 nm is obtained.

We demonstrate a tunable optical add/drop multiplexer (OADM) that uses a micromachined 2 × 2 optical switch and a tunable fiber Bragg grating (FBG). The new OADM can be tuned to add/drop one of the multi input channels dynamically. The insertion loss of the dropped channel and added channel are both 6.04 dB. The transmission loss is 2.08 dB. These losses can be reduced significantly if the loss arising from the circulators and MEMS switch is reduced. The crosstalk between channels is less than -20 dB, and can be further reduced by improving the reflectivity of the FBG used. The tuning speed is in the order of milliseconds. The tuning range is 3.0 nm. The fabricated system is experimentally demonstrated by selectively adding/dropping one of the four adjacent channels with the spacing of 100 GHz. Systems with multi channels being dropped and/or added can be achieved by cascading the proposed structure.

Recent advances of densed wavelength multiplexed photonic transmission systems are investigated for ultra-long reach (ULR) transport networks, including (i) Optical modulation formats including return-to-zero (RZ), carrier suppressed RZ (CS-RZ), vestigial side-band RZ (VSB-RZ)^[3] and their impacts on design and implementation of integrated optic transmitters and associated coding and modulation techniques for dispersion tolerance; thence extending the system span and optical node interconnecting distance (ii) dispersion management techniques chromatic and polarization mode dispersion and compensation techniques (iii) Optical amplification sub-systems including EDFA and Raman distributed and lumped amplifiers and gain equalization and (iv) photonic multiplexing and de-multiplexing sub-systems for the photonic transport networks.

Synchronous Optical NETwork, or SONET as it is popularly called, is one of the most advanced telecommunications technology available on the eve of the 21st century. The SONET standard provides a layer of compatibility, which allows for the exchange of bit streams with one common format. SONET equipment can accept country specific bit streams, convert them to the SONET standard, and then pass them across the SONET network with other bit streams. The paper centers on the design of the framer and deframer for OC –192 frames. They operate at data rate of 9.953 Gbps. The core parameters are the A1 and the A2 bytes, which are added during the framing process and then removed during the deframing process. . Since the channel is noise prone, errors can peek in at any place. This added error should not delay or spoil the deframing process. Furthermore, since a 64-bit bus is considered, the framing bits may start from any bit position. So a robust design that should take care of the bit errors and skews is needed. The architecture put forth is analyzed using various test conditions and is eventually coded using Verilog – HDL.

We review the progress and problems in the development of soliton transmission systems. Dispersion managed soliton could be a future candidate as the information carrier. Future research prospect is presented.

An improved model is proposed to approximate the actual Raman gain profile of silica, which in turn is used to calculate the power depletion of the shortest-wavelength channel due to all other channels in the system. In this improved model, the stimulated Raman scattering (SRS) effects from channels at the tail of Raman gain profile are taken into account. A statistical analysis is done for the SRS crosstalk in a wavelength division multiplexed system including modulation statistics. The bit error rate and power penalty are estimated, where the improved model for the Raman gain spectrum is used.

We present a novel simulation model based on the proposed segmented Volterra series transfer function (S-VSTF) for advanced optical communication systems. The S-VSTF model is used to obtain analytic solution of the nonlinear Schröndinger (NLS) wave equation. Simulated results are compared with those using the well-known split-step Fourier method and it is confirmed that the model is efficient and accurate for optical pulse transmission systems.

Progress in planar rare earth doped glass devices has been significant over the past few years, and commercial products are now available. In this paper, erbium doped waveguide amplifiers for telecommunication applications are reviewed.

Optical pulse with the repetition rate of 40GHz and pulsewidth of 7ps have been obtained by using of actively mode-locked fiber ring laser with 20GHz bandwidth electro-optical modulator. The performance of the output has been analyzed.

The self-frequency shift effect is examined in an actively mode-locked fiber laser. We observed, with the insertion of a filter, this effect can be greatly suppressed and the range of gain for stable operation is wider.

An actively harmonic mode-locked fiber ring laser was constructed and its stability was analyzed. A chaotic behavior in the actively harmonic mode locked fiber ring laser was observed.

The highly nonlinear interactions in multiple-wavelength pumped fiber Raman amplifiers (MW-FRAs) make the performance simulation extremely time-exhaustive. So the efficient models are necessary. Here we carried out simulation with Average Power Method and discussed the numerical results.

The following sectios are included:

• AUTHOR INDEX