Optical ADD/DROP multiplexers (OADM) are incorporated into all-optical network structures that provide fixed access to a subset of the wavelengths in Wavelength Division Multiplexer (WDM) systems. The rapid growth of broadband data communications and the drive toward cost reduction have made optical MEMS (Micro- Electro-Mechanical Systems) an extremely attractive technology for applications in optical communications. This paper will present theoretical analysis, simulation and testing results of an ADD/DROP multiplexer based on the MEMS-based micro-actuators. The micro-actuator is a MEMS-based compound grating (MCG) with a reconfigurable surface that couples the mechanical motion with optical diffraction. The diffraction patterns depend on the wavelength, incident angle and the grating structural parameters. This property is used to design an OADM that can be applied to broad areas in optical communication. A theoretical analysis is presented to establish the relationship between diffraction beams and the structural parameters of the grating, the wavelength of incident light, incident angle. Prototypes of these micro-actuators have been fabricated. The initial testing demonstrated the feasibility of using the MCG as an OADM. New designs of the MCG for application to the 1.55um optical telecommunication standard will be discussed.

A tunable optical add/drop multiplexer (OADM) is demonstrated byu sing a micromachined 2 × 2 optical switch and a tunable fiber Bragg grating (FBG). The new hybrid 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 2.84 dB and 1.8 dB respectively. The transmission loss is 2.04 dB. These losses are all able to reduce greatlyi f the circulators are modified. The crosstalk between channels is less than -20 dB, and it can be further reduced byi mproving the reflectivityo f FBG. The tuning speed is on the order of millisecond. The tuning range is 3.0nm. The fabricated system is demonstrated by selectively adding/dropping one of the adjacent four channels with the spacing of 100GHz. Systems with multi channels being dropped and / or added can be achieved bycas cading the proposed structure.

For the past decade, the explosive growth of Internet data communication, together with the ever-increasing traffic demand for traditional voice service, has fueled the rapid development of optical communication systems. While the aggregate traffic demand is largely met by the deployment of DWDM technologies and the increasing data rates toward 40 Gb/s, the current optical communication system has also evolved into a complicated multi-wavelength system, which asks for many innovative technologies for realizing its vast capacities and promises. One such greatly desired network element is a large-scale optical cross-connect (OXC) in the core network to facilitate optical layer networking at the wavelength level. In recent years a variety of technologies have been studied as potential candidates. Many advances have been made, and many lessons have been learned. In this paper, we will discuss the architectural and performance requirements for OXC, and analyze various optical switching technologies. Particularly we will focus on the microelectro- mechanical-systems (MEMS) technology that has demonstrated vast promises for large-scale OXCs.

Micro-electromechanical Systems (MEMS) have been around since the 1960s. Early applications of MEMS were biomedical and automotive such as drug delivery system, disposable blood pressure sensors, accelerometer used in airbag sensor and antilock braking systems. Recently, this technology, known as micro-optical MEMS or MOEMS, is invading the fiber optic communication industry for its ability of moving and managing light directly without converting the optical data to electrical signal for processing, hence it is immune to bit rate and data protocol. This paper will discuss the 3D MEMS optical switch development program at Agere Systems from the design concept to volume production and the dedicated reliability program to qualify this technology for telecom applications.

The electromechanical characteristics of a novel micro torsion-mirror actuator and the optical properties about the light scattering and reflection of its micro-mirror under the influence of mirror surface roughness are investigated experimentally and theoretically. The electrostatic yielding voltages for driving the metal-coated polysilicon or single crystal silicon micro-mirrors suspended by elastic torsion beams to tilt from 0° to 90° spontaneously are in the range of 270~290 V and the minimum holding voltages for keeping the tilting angle of the mirrors to be in 90° are found 55 V or so when the thickness of the torsion-beams is about 1 ?m. The lifetime and the estimated shortest actuating time of the micro torsion-mirror actuators can reach 108 at least and less than 2 ms, respectively. The maximum value of the surface roughness of the micro-mirrors is 69.443 nm and the distribution of the surface roughness is smooth, which is acceptable for wavelength division multiplexing applications basically. In general, the novel torsion-mirror actuators can be applied as optical switches for optical networking.

We propose a share-per-node hybrid OXC with MEMS that has constraints to the number of connections between the lamda-plane and the converter-pool. And we devise the worst case blocking model which can calculate the blocking probability in large size OXC architecture and can also shows that throughput for the constraint share-per-node model over the full share-per-node model. Using the results of analysis, we can optimize the size of MEMS without increasing the blocking ratio over the full share-node model.

Wavelength Cross-Connects (WXCs) are a class of optical switches designed to dynamically route the transport of wavelengths between fibers in DWDM systems. More generic optical cross-connects (OXCs), also called optical core switches (OCS), are designed to switch optical channels without regard to wavelength. WXCs are typically deployed to add/drop wavelengths from a DWDM stream, to route wavelengths in DWDM mesh networks and to provide path protection and restoration in DWDM rings and meshes.

Spanke switch architecture is predominantly used in telecommunication switching circuits. In this architecture the number of inlets is equal to the number of outlets (symmetric type). However, some applications require unequal number of inlets and outlets (asymmetric type). No model is available to characterize the parameters of asymmetric configuration. This paper introduces a generic model for the parameters of Spanke optical switch configuration. Modeling has been done for the number of switch elements, number of crossovers, maximum loss and minimum loss. The properties of Spanke optical switch configuration are analyzed and their corresponding curves are obtained. The proposed model can be used for both types of Spanke configurations so to get optimal number of switch elements and crossovers.

Ultra high-speed optical switches (< 10 ns) are more important in optical networks and optical computers. This paper analyzes the switching speed and insertion loss of (2 x 2) Total Internal Reflection (TIR) electro — optic switch in Multiple Quantum Well (MQW) and Quantum Dot (QD) structures using equivalent circuit model approach. Then extended to analyze the switching speed and insertion loss of switch fabric (Spanke Architecture). For the purpose of this study, AlGaAs/ GaA° multiple quantum well stmcture of length 150 micrometer and width of 40 micrometer and InAs quantum dot of 100 K diameter are considered. This work is carried out with SABER Sketch (ver.4.2.3) package. It is found that the obtained switching speed of both quantum dot switch and fabric are improved by a factor of 1000 (from the order of nano seconds to femto seconds) and insertion loss reduced by a factor of 1000 (from the order of dBm to dBµ) while comparing to quantum well switch and fabric. Also noticed that the operating speed of the switch fabric is relatively oscillatory when the number of switch fabric stage is less than four.

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.

Fully reconfigurable broadcast and select OADMs are compared to conventional designs for ULH networks. The feasibility of an 80 x10.7 Gb/s broadcast and select OADM chain with an unregenerated reach exceeding 4160 km is demonstrated. Key engineering issues for widespread commercial deployment of all-optical ULH networks include ASE noise accumulation, filter concatenation effects, dispersion, fiber non-linearity, and crosstalk impairments.

Wavelength Division Multiplexing (WDM) multiplexes a number of optical signals at different wavelengths in a single optical fiber. Each optical signal comprises a separate channel, and thus the information capacity per channel is bounded by the modulation speed of the transmitter. Therefore, the engineering parameters that determine the complexity of the fiber span and cost of transporting data depend on the number of channels, channel spectral separation, optical power per signal, bit rate, type of fiber and length of fiber link. Based on these, a determination is made if amplification and what type is needed, if dispersion (chromatic, polarization) is significant, if other non-linear effects are considerable, and in general if there are mitigating parameters that affect the integrity and quality of signal. In this paper we present a parallel transmission WDM method by which high-bandwidth throughput is achieved yet with low modulation bitrates and low power source lasers. This method minimizes dispersion and other nonlinear phenomena over a given span, it greatly enhances the quality of signal, it increases transmission reliability and reduces cost per transported bandwidth in short and mediumhaul, Metro and point-to-point, DWDM applications.

In the future Internet, di erent applications such as Voice over IP (VoIP) and Video-on-Demand (VoD) arise with di erent Quality of Service (QoS) parameters including e.g. guaranteed bandwidth, delay jitter, and latency. Different kinds of service classes (e.g. gold, silver, bronze) arise. The customers of di erent classes pay di erent prices to the service provider, who must share resources in a plausible way. In a router, packets are queued using a multi-queue system, where each queue corresponds to one service class. In this paper, an adaptive Weighted Fair Queue based algorithm for traÆc allocation is presented and studied. The weights in gradient type WFQ algorithm are adapted using revenue as a target function.

An integrated IP (Internet Protocol) network that can handle leased-line, voice, and data traffic, offers great advantages in simplifying the networks of communication carriers and reducing network costs compared with separate servicespecific networks. A key issue of such integrated IP networks will be to design a router that can efficiently handle each type of traffic strictly according to its QoS (Quality of Service). Diffserv (Differentiated Services) is a popular solution to handling multiple QoSs, so most routers tend to support it. In this paper, we propose a new DiffServ queue architecture featuring multiple EF (Expedited Forwarding) queues with traffic shaping functions. Because each EF queue has a traffic shaper, its QoS can be controlled without affecting other EF traffic. For example, when the leased line traffic is assigned to the first EF class and VoIP (Voice over IP) traffic to the second EF class, the VoIP traffic is transferred with high QoS even if the leased line traffic has bursty characteristics. This queue architecture can provide a low-latency, assured bandwidth for both leased-line and VoIP traffic. We evaluated the traffic characteristics of VoIP traffic by simulating this queue architecture using various bandwidths and voice codec types. The results indicate that this queuing mechanism can achieve lower latency and better QoS control than the usual DiffServ queue architecture without traffic shaping. The benefits of traffic shaping are more noticeable with low-bandwidth traffic.

Multiservice networks will carry di erent kinds of applications in the near future. Bandwidth requirements change rapidly, and the network resource management will play an important role to guarantee the use of the limited resources in the most eÆcient way. We approach the channel capacity allocation problem by developing an SLA (Service Level Agreement) based channel allocation method. In our model, the channel may be wired or wireless, so this method can be adapted in multi-technique networks. The algorithm allocates resources to several di erent service classes via several di erent capacity routes. Service provider perfroms optimization by allocating data rate in such a way that the satisfactory of the customers as well as the revenue is maximized.

In this paper, we propose a novel optical packet switching mechanism incorporating void filling that makes some packets switch in optical domain, others in electronic domain. The optical packet switching is solved in time, space, wavelength and electronic domain. We also present numerical simulations under Poisson traffic and variable IP packet length condition. Simulation results have shown improvement in packet loss probability by using our switching mechanism.

The paper describes our activities on optical packet switching by means of our out-of-band optical label switching technique. It also discusses our recent progress in some key components for optical packet/burst switching, including an optical packet synchronizer and a burst-mode optical receiver.

The realization technology of a new kind of multi-wavelengths label optical packet switching technology is discussed. In this switching, optical header is labeled by several optical pulses at different wavelengths in the same optical communication channel band as optical payload. The optical transmitter with header generation and packet formation, receiver with data restoration and switching node with route processing are introduced. The switching principle is verified by a simplified experiment.

In this paper, we focus on data channel scheduling algorithms for offset-time based QoS scheme in OBS networks and their performance evaluation. We propose two new data channel scheduling algorithms, namely, the Earliest Available Unscheduled Channel (EAUC) algorithm and the Least Gap with Void Filling (LGVF) algorithm. Simulation results show that the proposed burst scheduling algorithms can not only provide the better fairness for scheduling data channels and fair QoS supporting of different class service, but also achieve high link utilization while keeping the burst dropping probability low.

Address/label lookup and data forwarding are the most fundamental functions of a router. Much research work has been concentrated on minimizing the lookup and forwarding time in a high-throughput Terabits router. Though novel algorithms and data structures have been developed to overcome this potential bottleneck, all of these schemes are still processed in the electronic domain. This implies major throughput limitations on electro-optic conversion and processing of packets one at a time (per forwarding engine). In this paper, we attempt to overcome these bottlenecks by introducing multiwavelength optical packet-based processing for table lookup. The distinctive features of this technique include, (1) feasibility of processing (table lookup) packets in their native form, namely optics; (2) enhancing the parallelism to provide a constant time comparison; (3) (capable of) computing in tens of terabits per second. Such processing technique is highly desirable in packet-based networks.

The application of transfer function matrix (TFM) synthesis concepts of linear space-invariant systems to lightwave circuits (LWCs) is reviewed and briefly discussed. The applied decoupling concepts are 1-D though the systems are 2-D where the application of 2-D decoupling concepts is left for forthcoming papers. Although the approach is applicable to nonlinear (NL) systems we restrict presently to linear systems. Distributed feedbacks (FBs) are established by means of a flexible grating concept which includes the electro-optical generation of reflections as well as their tuning. In this way reflections may be switched off and on, enabling novel devices and operations (presented in forthcoming papers) , and the gain of FB controller matrices may be adjusted according to the algebraic results of control theory. The discussion of further components (optical isolators and frequency converters), needed throughout the proposed concept, is left for forthcoming papers.

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. In addition, deep -etched structures such rotary gratings, circular mirror, microlens, and grooves for diode laser and fiber are illustrated to form widely tunable lasers.

As silicon LSIs become more densely integrated and the bit rates between them increase, traditional printed circuit boards present problems similar to electrical cables, which are already a bottleneck to cabinet interconnection. These problems include limited line length, poor flexibility in layout design, power dissipation, and crosstalk (EMI). One promising solution to these on-board electrical interconnection bottlenecks is to use optical interconnection technology at the chip level. Making chip-level optical interconnections practical requires revolutionary changes in optoelectronics packaging; the packaging must not only keep pace with silicon LSI capabilities but should also inherit the advantages of today’s mature electronics packaging and manufacturability technologies. To meet these requirements, we developed novel optical-I/O packages for low-cost chip-to-chip optical interconnections. Unlike conventional optical packages, our packages are fully compatible with surface-mount technology (SMT) because they do not need optical connectors. The packages free users from all on-board connection work and fiber management. Using the developed optical-I/O packages, in which a VCSEL/PD array is mounted at the bottom, chip-to-chip parallel optical interconnection through a polymer waveguide array is demonstrated.

As the Internet traffic exponentially growing, the next generation IP network is forced to scale far beyond its present performances. The more and more mature optical switching technology, such as optical burst switching, is expected to provide an ideal infrastructure for meeting the demands. However in optical switching, there is one critical issue, namely contention, which roots from multiple optical data requesting the same output port How to resolve contention in optical domain will have a significant effect on the performance (in terms of the burst-loss rate, average delay time and network throughput) of optical switching network. The paper proposes a contention resolution scheme based on FDL, AWG and TWC. Here FDL is used as two functions, i.e. forwarding and feedback for smaller or longer buffering time requirements respectively. In the paper we incorporate the scheme into the design of optical switch. We descript the optical data buffering strategy when contention occurs. We also study the performance of the scheme in a Markov process model under the assumption of uniform Bernoulli traffic, and validate the analysis through numerical simulation. The computer simulation results show that the scheme can efficiently use FDL buffering and AWG switching capacities, hence can obviously reduce the contentions.

This paper reports the design, modeling, fabrication and testing of a novel variable optical attenuator for multi-channel power equalizers to be used in dense wavelength division multiplexed (DWDM) 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 ?m2 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. It has 1.5 dB insertion loss and 45 dB attenuation range, and requires only 8 V driving voltage, showing that it is promising for DWDM applications. Optical and mechanical models of the attenuator have also been established. Although the models are developed with the initial intention of modeling the MEMS attenuator, they are also available to the other types of devices in which the preconditions of the models are satisfied.

In spatial multiplexing method, the steering devices for reference wave and object wave are necessary. A new scheme applying computer-generated holograms(CGHs) to the steering device is proposed. The beam steering device using CGHs can be performed simultaneously the coarse address function directing the reading or writing beams to the suitable layer and the fine address function corresponding to the particular holographic page within the chosen common volume unit. This new scheme is compared with both the beam steering method using spatial/angular multiplex AODs and the mechanically steered method in terms of access time, cost, and efficiency.

We study the design of planar micro-grating and propose a new approach to fabricate planar micro-grating. By the basic grating diffractive theory (rigorous coupled wave theory), suitable and optimized structure (within our manufacturing capabilities) of grating for our application is found. The fabrication of planar micro-grating is demonstrated by Laser LIGA technology, using the instrument of KrF (248 nm) excimer laser system. The micro-grating with the continuous smooth profile and the pitch 2.5 ?m is shown.

To analyze pulse switching in two-nonidentical-core couplers, we derive a pair of new generalized coupled-mode equations that describe the switching dynamics of optical pulses in two nonidentical parallel waveguides by considering dispersive properties of the coupling coefficient. We study the influence of the first-order coupling coefficient dispersion (i.e. intermodal dispersion) on pulse propagation in a directional coupler which consists of two nonidentical cores. It is found that the intermodal dispersion can cause severe pulse distortion and affect the coupler functioning as an optical switch. To show how the pulse switching dynamics are affected by the nonlinearity of the coupler, we study the pulse propagation under different input conditions. It is studied that an increase in the threshold amplitude is obtained for the case of nonidentical cores.

In digital image processing image edge detection, enhancement and extraction have very important role. There are many methods and systems, already accepted very widely, which show the above operations with their own advantages and limitations. Here in this paper we propose a completely new scheme of getting the edge of any digital image by proper use of optical non-linear material. Some intensity modulated switching behaviors of optical non-linear materials are also coped up in this method. Here we can obtain the required image edges as necessary. Images with variation of transparencies are used here suitably.

Bandwidth demand in Internet is growing at a very fast rate. While links can provide the requested bandwidth, the situation is much more critical for routers. The design of high-speed switching fabric is a crucial part in current routers and switches. This paper presents a scalable and high speed switching fabrics for Internet routers— —torus network, which was mainly used in high performance computers before. Using wornihole routing and virtual channel in torus network can achieve a high-speed and scalable switching fabric for Internet router. In this paper, the performance of 4 X 3 X 2 3D torus network was analyzed as a core switching fabric for Internet routers. A novel dynamic and distribute routing algorithm is employed at switching system. We achieved an analysis model in OPNET simulation tools. Throughput and latency characteristic can be analyzed through this model. Simulation results show that maximal throughput is close to 100% under uniform traffic with low latency. It was found that using this architecture can reach Tera-bit capacity switching fabric and can be expanded from 36OGbit/s to 1.6 Tbit/s without increasing the link speed between nodes. At last, this paper presents some exciting directions in which the work can be extended.

In order to reduce the complexity of the optical switching nodes at a reasonable level and increase its capacity at the same time, the concept of multi-granularity was proposed. With the utilization of multi-granularity switching, the number of ports for the optical switching matrices can be decreased. In this paper, we consider the two-stage multiplexing that includes the wavelength multiplexing and waveband multiplexing. A simple algorithm that can realize the grouping from wavelengths into wavebands efficiently is proposed. As the decreased number of the optical switching matrices' ports is determined partially by the number of the nodes passing the waveband path, the length of the waveband path is considered in this algorithm. To group as more wavelengths as possible into wavebands, all the wavelength paths will be evaluated in the algorithm as long as they have common sub-path. This algorithm is also tested in two networks with six nodes, and the saving number of the switching matrices' ports is compared with which using only the one-stage wavelength multiplexing.

Wavelength-division multiplexing appears to be the solution of choice for providing a faster networking infrastructure that can meet the explosive growth of the Internet. Future networks are expected to utilize the WDM technology for optical functionalities capable of handling multi-Teragabit signals and thus the cost, therefore it is very important that the Optical cross-connects (OXCs) and optical Add-drop Multiplexers (OADMs) have performed routing and switching capacities. Optical Packet Switch (OPS) technology allows us to rapidly deliver the enormous network bandwidth. OPS offers high-speed data rate and format transparency. But there are still many technologies and key components need to solve. In this paper we propose a novel architecture for OPS using Tunable Wavelength Converters (TWC) and Fiber Delay-Lines (FDLs). We use wavelength converters and demuxes connected with several FDLs that provide different delay periods. Input wavelengths can be tuned to different ones in order to acquire different delay periods. Wavelength converters are shown to improve the traffic performance of the switch blocks for both random and burst traffic. It can also provide contention resolution solution in wavelength and time domains. This architecture improves packet switching speed and significantly decreases the use of optical switches comparing with general architecture. Finally, the simulation results show that the packet blocking can be decreased.

Packet contention is an important problem in optical packet switching system. The contention resolution techniques greatly influence the design of optical packet switch. There are several techniques for contention resolution and their mechanisms are illustrated in this paper. In order to reduce the hardware volume and cost of fiber delay lines and tunable optical wavelength converters in optical packet switch, sharing use of optical component becomes an efficient candidature. A new optical packet switch model with sharing resources has been proposed and its performance is evaluated by simulation. It is shown that by sharing use of wavelength converters and fiber delay lines, the number of optical components is greatly reduced and the performance with respect to packet loss probability is acceptable.

Optical Burst Switching combines the advantages of circuit switching and package switching, and it can improve the efficiency of system resource with high speed transmission and Qos. Now the rationale of OBS has been discussed a lot, and many schemes have been given. While there are still some key problems need to be resolved, such as in the model of IP over WDM, how to assemble the IP packages into burst that will be transmitted in the optical networks; and how to set the offset time between payload and its control package. In this paper, we provide a part randomized offset-time method to calculate the offset time upon the basis of: the network condition of the optical path that the payload will pass, the shaping scheme which is introduced to reduce the drop probability. At the same time, although there are many factors should be considered when setting the offset time, too sophisticated algorithm will be discarded. Because our ultimate end is to get equilibrium between the control complexity and bandwidth efficiency even some redundancy will be introduced in this method. We also consider a balance between end-to-end delay and blocking probability introduced by the new offset-time determination algorithm. At last, we will analyze the improvement in the network performance if the offset time has been set properly, especially when associated with delayed reservation technology.

This article focuses on the problem of switch control strategy for optical burst switch with optical buffer supporting variable length packets and under asynchronous mode. Two novel control algorithms are proposed, and their performances as well as that of the well-known reserve-switching mode, are evaluated and compared by computer simulations. Issues on the optimal configuration of optical buffers are also discussed. The numerical results show that the time delay resolution of fiber delay line has significant influence on the system performance in all three switching modes.

A new scheme of variable optical buffer for IP packets is reported. It may be used in access control of HORNET (Hybrid Optoelectronic Ring NETwork), to avoid collision of added packet and the packet already on the ring and improve the loss ratio of the packets. In this scheme, a new multi-wavelength fiber loop memory technique is employed. This architecture uses the wavelength converter (WC) to specify the packets delay and wavelength conversion is accomplished by the technique of four-wave-mixing (FWM) with a semiconductor optical amplifier (SOA). The range of delay is 10 to 9990 bytes periods. First, architecture, operation principle, characteristics and applications of this scheme of variable optical buffer are introduced. Next, a new unslotted CSMA/CA MAC scheme based on the variable optical buffer is briefly introduced too. Finally, the simulation results are presented.

Optical label switching is a promising optical packet switching technology to ease the process of getting the route information from the packet label in optical domain. In this paper, a promising kind of multi-wavelength label is adopted and to which a simple label extraction and recognition technology through a array of tunable Fiber Bragg Gratings (FBGs) is presented.

We propose and simulate a novel optical label swapping scheme using a frequency-doubled subcarrier multiplexed (SCM) signal processing technique. This method can support a dispersion-insensitive and polarization-independent transmission and extraction of an SCM optical label using a standard fiber Bragg Grating (FBG) filter and an optical circulator without a conventional SCM coherent detection circuitry. Simulation results show that our method is applicable to an FBG filter with wide reflection bandwidth (BW), and relieves the restriction of an optical FBG filter, satisfying the wavelength tolerance of the dense WDM sources and FBG’s center wavelength . Driving an external Mach-Zehnder (MZ) modulator over the full switching voltage improves the receiver sensitivity, compared to the previous standard double-sideband (DSB) SCM header processing technique.

Spatial. input/output encoding technique has established itself as very sucessful candidate in all-optical digital information processing and computing. This technique was used in several all-optical parallel operations, which were proposed by many scientists. Here in this paper we 'propose a completely new mechanism by which we propose a completely new mechanism by which we can encode any number of input variables. It is known that there are many correctional technique to encode two binary variables. For example, if A and B are two binary variables where each of which can take either 1 or 0 value, we can encode A and B very easily. But the problem comes if the number of input variables increase. Let it is A, B and C (three variable case) or A, B, C and D (four variable case) or any higher number of variables. In this communication we find a new technique by which we can encode three variables of input, four variables of input or any higher variable of inputs. This encoding is done in such a way that we can proceed to conduct all-optical operations with proper uses of input variables.

R-OADMs are a key network element in next generation all-optical networks with applications in both metro and longhaul systems. R-OADMs enable dynamic connectivity in optical networks, enhance network efficiency, make automatic provisioning possible and increase network survivability. They also reduce the network management and operation costs. Switching R-OADMs utilize field-proven technology and components, offering a reliable, low cost solution.

A multi-pipeline optical interconnection network for distributed computing system has been designed. Each sub-layer network is connected to the ring with an access node (AN), which can transmit data at every wavelength with a tunable laser diode. The data transmission speed at each wavelength is 1.25 Gbit/s. With 8 wavelengths, a total bandwidth of 10 Gbit/s can be obtained. Each AN only receives a certain wavelength. With a band pass filter, the desired optical signal can be dropped down. Pipelining data transmission is achieved among different wavelengths. This network is a multi-pipeline structure. So the communication latency and communication overheads can be decreased. Meanwhile, the ring topology has good scalability. The scale of the network can be expanded adopting more wavelengths at each access node.

The temperature-dependent insertion-loss of volume phase holographic wavelength-division demultiplexers(WDDM) is analyzed using Bragg diffraction theory in this paper. We demonstrate that most of the temperature-induced insertion-loss results from the diffraction deflection. Finally, we propose a new way to improve the temperature characteristics of volume phase holographic demultiplexers if we have a suitable choice between the thermal expansion coefficients of the gratings and the thermal coefficients of the gratings index. This result is useful to design a self-temperature compensation WDDM.

Optical Burst Switching (OBS) functioning as an IP traffic aggregator at the edge of the Optical Transport Network is provided to enhance network flexibility and efficiency. This paper gives an overview of OBS node and the benefits it will bring about. It explains how an OBS node is modeled and simulated under different traffic conditions and gives the effect of Fiber Delay Line (FDL) in OBS node. Finally, the simulation results are presented and the performances of OBS node are evaluated.

A novel 2 X 2 free-space polarization-independent bi-directional fiber optical switch is presented in this paper. A general formula for coupling-loss of fiber collimators is derived. The insertion loss and crosstalk are analyzed in detail. The experiment results indicated that the insertion loss was less than 1. 1 dB, the interchannel crosstalk was about 18.5 dB, and the switching time was about ims. The new kind of configuration of the optical switch grants the features of few optical components, high compactness, low optical interchannel crosstalk, fast switching speed, polarization insensitivity, and easiness of optical assembly.

First, the relationship between effective index change and thickness of cladding and substrate in thermo-optic switch is studied. Then, rectangle heater is simulated using the Beam Propagation Method. And a novel heater, arc heater, is designed. The thermo-optic switch using arc heater has better properties than using rectangle heater.

The 4X4 LiNbo3 optical wave-guide switch matrix , which is composed of five directional couplers, is designed. Alternative ? ? electrode structure is adopted for easier fabrication. The guides of switch matrix fabricated successfully are shown in this paper. Light testing is proceeded and the result is also shown .Some important switch matrix's performance are calculated theoretically as following: switch time 2.9ns; switch voltage 25.6v; insertion loss 7.0 17 dB.

Optical Label Switching (OLS) is one of the practical manners of optical packet switching. It is a main research domain in optical networks. The advantage of OLS is its bandwidth efficiency and ability to support various network devices. There are a lot of practical problems need to be resolved. Today, research into OLS technology has focused on key networking and hardware issues such as high-speed optical switch, optical storage component, and network nodes source distribution optimization algorithm. The optical Label Header contains header synchronizing bits, routing address bits, etc. Header reading and rewriting includes synchronizing, routing and checking technologies. So it is the key issue in the OLS, and has very important research and application values. In this paper, a novel proposed fabric scheme was discussed to achieve the function of the header reading and rewriting. It can be applied to the high-speed header reading and rewriting to satisfy the requirements of practical applications. The fabric scheme adopts a high-speed tunable laser and wavelength converter to rewrite the header. The high-speed tunable laser is based on Grated Coupler Sampled Reflector (GCSR) structure and the wavelength converter is based on Cross-Phase Modulation (XPM) in a Semiconductor Optical Amplifier (SOA) and a Mach-Zehnder Interferometer (MZI) structure. The synchronization of the packet in the node also be discussed.

In this paper, we overview the research and development of optical interconnection network from 1993 to 2001 in Tianjin University. These works are divided into two categories. One is optical interconnection network for massively parallel processing systems. Another is optical interconnection technique for computer clusters. In the former, we have designed a massively parallel processing system that consists of 64 processors, which are interconnected by optical fiber links. In the later, we have developed giga-bitls optical links, hardware-routing optical rings and wavelength-routing dual rings. The optical layer data transmission speed has been improved from 20 Mbit/s in the massively parallel processing system to 1.25 Gbit/s in the computer clusters. Meanwhile, the communication latency has been reduced to 200 ns adopting hardware-routing technique. To further reduce the data packet routing-latency, wavelength routing technique has been implemented in a dual-wavelength ring. Finally, we have designed some novel network structures for tera-bit/s optical interconnection network.