Search

A new scheme for optical burst switching (OBS) networks under self-similar traffic is proposed in this paper. Self-similar traffic is classified into two types of payload by partial pre-deflection edge node router, and the two types of payload are transferred using different routing algorithms - least hop first routing and shortest path first routing respectively, in order to properly utilize network resources and avoid burst contention at the core node switch. Simulation results indicate that the new scheme performs better than fixed routing and alternative routing for OBS networks under self-similar traffic. Furthermore, the implement of this scheme is feasible and easy, which makes OBS technology more applicable in network engineering (NE).

Many recent studies have convincingly demonstrated that network traffic exhibits a noticeable self-similar nature, and most studies of optical burst switching (OBS) networks are under a fundamental assumption that full wavelength conversion is available throughout the network. In practice, however, economic and technical considerations are likely to dictate a more limited and sparse deployment of wavelength converters in the optical network. Therefore, we present a novel scheme for OBS Networks, called logical cascaded private subnet with start wavelength assignment policy. We define the concept of canoe relative to cluster in self-similar traffic, and introduce a new device named payload segregator at the edge node as a gateway to the OBS core node. According to the changes in the edge node framework, we put forward the concept of cluster private subnet and canoe private subnet in OBS core node correspondingly. Due to the absence of wavelength conversion capabilities, we assign start wavelengths to both private subnets before network start operation. A new start wavelength assignment policy is proposed for the absence of wavelength conversion capabilities in the core node of OBS Networks. The analytical results clearly show that logical cascaded private subnet scheme can yield better performance in terms of block probability than traditional OBS Networks, and logical cascaded private subnet with start wavelength assignment policy has the potential to improve the block probability by more than one order of magnitude compared with traditional OBS Networks, helping reduce the performance gap with respect to full wavelength conversion.

With fiber being cheaper and cheaper, the core fiber bandwidth utilization is not the key factor of the network, how to use the huge bandwidth flexibly and efficiently is one of the hot topics of optical networks. Optical burst switching (OBS) is the best choice to statistically use the optical resources, OBS based transparent optical network attracting attentions. We proposed an integrated policy by using dual fiber links server model along with traffic spacing mechanism (TSM) by address differentiating and transmitting those traffics between adjacent nodes by IP directly, the others by OBS mechanism. The theoretical results show that this method can improve network performance and be practically used in fields.

Many recent studies have convincingly demonstrated that network traffic exhibits a noticeable self-similar nature which has a considerable impact on network performance. Therefore, we present a novel scheme for optical burst switching (OBS) networks, called logical cascaded private subnet. We define the concept of 'canoe' relative to 'cluster' in self-similar traffic, and introduce a new device named payload segregator at the edge node as a gateway to the core node in OBS Networks. According to the changes in the edge node framework, we put forward the concept of cluster private subnet and canoe private subnet in the core node correspondingly. Performance study indicates that, the new scheme is robust under self-similar traffic.

Most studies of optical burst switching (OBS) networks are under a fundamental assumption that full wavelength conversion is available throughout the network. In practice, however, economic and technical considerations are likely to dictate a more limited and sparse deployment of wavelength converters in the optical network. Therefore, we present a new start wavelength assignment policy for the absence of (full) wavelength conversion capabilities in the OBS Networks. The performance study indicates that, our new scheme is robust under wavelength continuity constraint as well as Poisson and Self-Similar Traffic.

To control optical phased array by steering the wavelength of the light source, this paper presents the method using 2-D photonic crystal waveguides, which replace the conventional waveguides to guide and form the beam in the array. As a result, we present a 1-D optical phased array, whose optical beam can scan an angle from -40 degree to +40 degree with a SNR about 15 dB by a relative simple device in ±3 nm tuning at 1550 nm.

Multi-granularity switching in optical layer has gain significant importance due to the reduction of size and cost of the optical cross-connect. In this paper, we propose a novel structure of MG-OXC, named two-layer MG-OXC, and also give a new graph model for describing the function of the all kinds of MG-OXC. The network states can be represented completely through our auxiliary graph. In such an auxiliary graph, the optical tunnel allocation in multi-granularity switching network is reduced to the min-cost path find problem, and different traffic grooming policies can be archived simultaneously. We compare the performance of the two-layer MG-OXC to the normal three-layer MG-OXC and single-layer MG-OXC by the algorithm we proposed through simulation, and conclude that the novel MG-OXC can get very close performance to threelayer MG-OXC but smaller size of switching matrix. The simulation also proved the algorithms we proposed performs better than traditional RWA algorithms dealing the routing and wavelength sub problems separately.

We demonstrate transmission of 36 10-Gbit/s WDM signals over 4134 km of standard single-mode fiber using all-Raman amplification and NRZ format. Wide-band fiber delay loop depolarizers are used in the Raman pump unit to reduce the depolarization dependent Raman gain. Keywords: Optical fiber transmission, Raman amplifier, non-return-to-zero, wavelength-division multiplexing, dispersion compensation, forward error correction INTRODUCTION In recent years ultra-long haul large capacity dense wavelength-division multiplexing (DWDM) optical transmission systems have attracted considerable attention due to their potential to greatly reduce the cost per bit × distance. One of the main limitations for such ultra-long haul transmission is the degradation of the link optical signal-to-noise ratio (OSNR) because of the amplified spontaneous emission (ASE) noise accumulation. The channel power cannot be increased unlimitedly due to the fiber nonlinearity impairment. All distributed Raman amplification (DRA) has recently demonstrated its potential in DWDM transmission [1-3]. In addition to its low noise, another attractive characteristics of Raman amplifier is the freedom in designing the amplification band through properly choosing the wavelength of the pump. Several record experiments have been demonstrated based on all-Raman amplification [1-4]. In this paper, we report a transmission experiment of 10Gb/s DWDM signals over 4000 km of standard single-mode fiber (SSMF) with 106 km spans using all-Raman amplification and non-return-to-zero (NRZ) format. In our Raman pump units, rather than polarization multiplexing, the WDM pumps are depolarized by fiber delay loops depolarizer. This Raman pumping scheme eliminates the polarization multiplexing and reduces the number of pump diodes by a half.

A novel driving scheme utilizing the resonance effect of the micro-mirror is presented to reduce the drive voltage of the silicon-based MEMS (microelectromechanical systems) torsion-mirror optical switch. A mixed bias of DC (direct current) and AC (alternating current) is adopted to drive the torsion mirror. Both the numerical simulation and experiment results show that the driving voltage of the optical switch is reduced by nearly 50%, from more than 270V to 130V.

Optical generation of 50 GHz mm-wave WDM sources with low phase noise, low fiber dispersion, and wide tunable range is investigated for radio-on-fiber system. Uplink signal transmission with phase noise of -95 dBc/Hz at 100 kHz is tested.

We introduce an optical WDM communication system that uses 3 erbium doped fiber amplifiers (EDFAs) with 12 channels and a total transmission length of 250 km. Eight of these 12 channels have a bit rate of 140 Mb/s each with a CMI code pattern, and the other 4 channels have a bit rate of 622 Mb/s each with an NRZ code pattern. This system works well with a low bit error rate. In China, this WDM system has the largest number of channels. The system setup and measured results are also shown.

The network architecture where each node integrates an IP router and an OXC (optical cross-connect) is an attractiveparadigm for the next generation optical Internet. LP wavelength routing or POW (packet over wavelengths) is based onthis kind of architecture. The goal of this architecture is to switch as much traffic as possible directly by means of OXCs,because IP forwarding is relatively expensive and causes large end-to-end delay by comparison. Based on analyzing theproblems and shortcomings of the previous node structure, a novel node structure is proposed to overcome the problemsencountered by the previous scheme. Then two IP switching protocols, IFMP (Ipsilon Flow Management Protocol) andGSMP (General Switch Management Protocol), are extended to support [P wavelength routing networks. Finally, theperformance of an IF wavelength routing network is analyzed with self-similar network traffic model. The simulatedresults show that the higher the network traffic self-similar degree, the higher switching gain the network gets.Keywords: IP over WDM, IP wavelength routing, Self-similar traffic, Switching gain

In this paper, we discuss three GVD compensation schemes, including the post-compensation, pre-compensation and hybrid-compensation schemes with considering combined impacts of GVD, Kerr effect and PMD. By means of the split-step Fourier method, we numerically solve the nonlinear coupling equations of two orthogonal polarization modes. With respect to the statistical character of PMD in the optical fiber, we employ the parameters of average eye-opening penalty to evaluate the combined impacts induced by GVD, Kerr effect and PMID. The numerical results are obtained on the base of 1000 times computation with average DGD less than 35% bit duration. Simulations show that, for the 10Gbit/s NRZ system, the GVD pre-compensation scheme performs best with more than 1dB of average eye-opening penalty better than the other two schemes when input power is greater than lOdBm. However, for the 4OGbitJs NRZ system, when input power P0 > 5.0dBm , the GVD post-compensation scheme performs best with more than 1dB average eye-opening penalty better than the other two schemes. This is different than the previous conclusion that GVD hybrid-compensation scheme has best performance without considering the impact of PIVID. We also define the parameter of eye-opening penalty at outage probability 10-5 to reflect the impact of DGD at the trailing of the Maxwellian distribution. The results also verify that GVD pre-compensation performs best in the lOGbitJs system and post-compensation performs best in the 40Gbit/s system. In both lOGbitJs and 40Gbit/s systems, either average eyeopening penalty or eye-opening penalty at outage probability iO indicates that the system performance with the best GVD compensation scheme can be improved with the increased input power up to 10dBm, and the performance will be worse if input power is greater than 13dBm. We analyze this behavior and conclude that PMD can be suppressed by the increased Kerr effect within a certain power range.

We experimentally demonstrate a mechanism of broad-angle and efficient unidirectional excitations of surface plasmon polaritons (SPPs). By combining the plasmonic critical angle phenomenon and grating effects, unidirectional excitation of SPPs with broad incident angular width up to 25° and high extinction ratio better than 10 dB was experimentally demonstrated. A peak extinction ratio of 17 dB was achieved. The proposed structure was also numerically simulated showing good agreement with experimental measurement. The operation principle was numerically and experimentally clarified through compare and analysis.

A unidirectional plasmonic coupler with ultra-broadband and wide-angle coupling efficiency is proposed and demonstrated. Employing the plasmonic critical angle (PCA) phenomenon in conjunction with the effect of subwavelength metallic gratings, a wide incident angular full-width-half-maximum (AFWHM >28°) coupling efficiency larger than 50% and extinction ratio better than 15 dB over an ultrabroad wavelength range 725–985 nm (bandwidth >260 nm) are achieved. The operation principle and performance of the structure are clarified and theoretically illustrated.

This paper introduces a new structure of WDM local area network. Each node uses WDMA (Wavelength Division Multiplexing Access) to access the network. The switching router of WDM LANs proposed before often use PSC (Passive Star Coupler), but now has been replaced by AWG (Arrayed Waveguide Grating), while each user equipped one tunable transmitter and one tunable receiver is reserved. As we all know, in WDM LAN, destination collision is inevitable when two or more then two nodes want to communicate with the same destination node and this degrades the network performance. So, in all related papers, transmission and receiving procedure must be pre- scheduled in order to preventing from destination collision. However, if there is a configuration, which has the possibility of permitting more than one to access the same destination simultaneously, access scheduling can be more efficient and system performance can be improved. Obviously, adding receiver number or transmitter number is impractical, thus, the only solution is to introduce buffer at the router output side. This is a distinctive difference of the network router we proposed from others. We believe this design can improve the network performance and simulation and numerical results have confirmed that.

Integrated-Optic Arrayed Waveguide Grating (AWG) N X N multiplexers are a new focus in All Optical Networks (AON) for their unique N input and N output geometry and have been used in a variety of applications. All AWGs fabricated so far have shown loss-imbalance among FDM channels due to non-uniform diffraction efficiency. It is fatal in AON because it will accumulate after each AWG. Uniform diffraction efficiency obtained in ideal design limit is one ideal solution, but it is nearly impossible to achieve the ideal limit with present fabrication technology. This paper applies a new method -- shifted connection -- on equalizing loss-imbalance of AWG- based OADM. The basic idea is simple and the result is of great significance. We theoretically analyze the effect of 8 various shifted configurations and predict each effectiveness. We find the relationship between the loss-imbalance and shift (delta) is a quadratic equation. We extend the method to cascaded AWGs. Also we deduce the limit of the equalization, which is independent of m and AWG's loss-imbalance parameter (alpha) , (beta) . Experiment results accord with theory very well and show that we successfully reduce a worst case loss- imbalance (9.5 dB) to a best case loss-imbalance (3 dB) in a single AWG-based OADM. We find that equalization configuration with foldback path is superior to loopback path configurations, when no more than N/2 wavelengths are used. The method is very simple and effective in equalizing AWG's loss-imbalance and the idea can be applied to other equalizations in AON.

We propose a novel passive optical network which can dynamic control of transmission sequence at high speed and high crosstalk rejection. The experimental system using two wavelength dilated acousto-optic tunable filters has been demonstrated. The wavelength dilation, doubling the distance between channels, provides substantial improvement in both interchannel and intrachannel crosstalk at the condition of fast interchannel switching speed.© (2000) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Acousto-optic tunable filter (AOTF), acousto-optic switch (AOS) and acousto-optic equalizer with high acousto-optic interaction efficiency are important in WDM networks. The technologies of unidirectional and focused transducer were adopted for the design of integrated acousto-optic device. Extensive usage for the acousto-optic device such as optical switches, gain equalizer and optical add-drop multiplexer have been studied. The concept that changing the radius of focus transducer can adjust the switching speed and FWHM of the filter was proposed for the first time.

In this article, we introduce a three-input three-output OXC which uses minimum number of small 2/spl times/2 optical switches. The blocking, loss, loss uniformity and cross-talk characteristics are also demonstrated.

Power management is discussed based on a WDM optical test network, in which ring laser clamped EDFAs are used to compensate loss and only the optical attenuators is needed for power regulation. All channels at the output of each node are adjusted to a preset power level, so signal power fluctuations and differences between channels are not passed to the later nodes. The transmission performance is also studied and a simple expression is derived for evaluating the possible input power to the fiber and the possible transmission distance in optical networks. The accumulation of optical amplifier noise and the maximum input power for the stable gain of the lasing-clamped EDFA are the dominant factors limiting the network scale, and restrict the preset input power to the fiber.

Two types of four-channel tunable optical add-drop multiplexers(OADM) are demonstrated which are based on fiber Bragg gratings(FBGs),optical circulators and switches. Both of them can dynamically select the drop channel or just bypass all WDM channels. The crosstalk from the adjacent channel is lower than -23dB and the isolation between the drop and add port is higher than 40dB.

A novel scheme of OADM(Optical Add/Drop Multiplexer) is proposed. It has the property of arbitrary add/drop channel compositions ( up to 4 wavelengths). In order to supress more of the homodyne crosstalk, we proposed a novel scheme of utilizing OLM(Optical Loop Mirror) so that we can utilizes the demultiplexer back-and-forth, thus utilizes it both as a demultiplexer and a multiplexer.The results show that we get a 5dB improve on the crosstalk depression, that means a 3 times larger network scale than normal scheme of demultiplexer+switch array+coupler. We found that the reflection on the input port of demultiplexer is the main cause of homodyne crosstalk.

The cascade EDFAs have to be used if the distance between nodes is extremely long in WDM network. However, the gain spectrum of normal EDFA is not flat. This will become much serious as the number of cascade EDFAs is huge. The gain spectrum is studied both theoretically and experimentally in this paper. A special designed FP filter is used to compensate the gain difference after cascade EDFAs used.

The integrated acoustically-tunable optical filter (IAOTF) is rapidly gaining importance in the field of WDM networks because it has the combined features of narrow bandpass , rapid switching time and broad tuning range . The low loss surface acoustic wave (SAW) transducer and the weigthed acoustic waveguide is studied using coupled-mode theory . It is proved that the theory is effective in the analysis and design of acousto-optic devices .

In a wide area WDM network with a huge capacity, dispersion limits the scope of the network. Among new ideas of dispersion compensation, dispersion compensation fiber (DCF) is the best. However, DCF can not balance the dispersion of all channels because of dispersion slope. There is surplus dispersion in some channels, so that further compensation of the surplus dispersion is needed after the using of DCF. The FP resonator is a dispersion-tunable device. The surplus dispersions of all the channels can be compensated after the suing of FP resonator as a tunable receiver. The numerical results show that the radius of WDm network can be made three times as large as it would have been without this device.

FP resonator has a very nice character which can be easily used as an OADM device. It can drop the selected channels and reflect the other channels which don't satisfy the resonance requirements. It can be sued as both channels drop and bypass device. In this paper, the design of good reflection spectrum is considered. The factors which affect the reflection spectrum of FP resonator are carefully studied. When using two FP resonators which have both nice reflection and transmission spectrum, it can be realized an OADM node without a circulator.

One parameter of the optical fiber is the crosstalk caused by its non-zero resistance band. The tunable flat passband filter is needed to improve the two characters, as the channel spacing becomes narrower, and the protocol of WDM network becomes faster optical packet switch one. This is because the flat passband can tolerate shorter channel spacing so that a large capacity and less tuning time are received. The structure of flat passband combines the multimode and basic mode to form the flat passband filter. This device can be designed with large tunable area and a very narrow passband.

A compact dichroic surface plasmon polariton (SPP) splitting scheme composed of two cascaded subgratings of different dimensions is proposed and investigated. The normal incident photons illuminated on the dichroic splitting structure are converted to surface plasmons traveling to the left or right depending on the wavelength. The operation principle is clarified and a coupled-mode model is developed to facilitate structure design. The generated SPPs propagating to the left and right directions on the metal surface can be manipulated nearly independently by altering the left and right subgrating, respectively. Proof-of-principle demonstrations show that a remarkable property of high splitting ratios is achieved with 43.0 dB at wavelength 1310 nm (left:right power contrast) and 35.7 dB at wavelength 1550 nm (right:left power contrast). The proposed splitting concept has general applicability across other operating wavelengths, such as the visible and near-infrared range.

Optical phase modulation based on directly modulated reflection-mode optically injection-locked VCSEL is investigated based on standard OIL rate equations and reflection-mode OIL model. The phase information of both static and dynamic state is simulated. The difference of static state phase information between transmission- and reflection-mode OIL is numerically analyzed. With specific OIL parameters, the output power of directly modulated OIL-VCSEL remains constant and phase deviation of 0.934π rad is obtained. Results show that a directly modulated OIL-VCSEL can function as a key component in QPSK or 8PSK transmitters. Preliminary 2.5 Gb/s PSK modulation characteristic is demonstrated experimentally.

This paper presents a 4-channel densely spaced optical frequency division multiplexing transmission/distribution system. Good bit error rate (BER) and eye diagrams are obtained. In more than 24 hours of continuous operation, no error bit is detected (BER < 1 X 10-14). The minimum received optical power measured before the demultiplexing Fabry-Perot filter is -30 dBm (BER equals 1 X 10-9).© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In this paper, wavelength control precision will be discussed when crosstalk is considered in a WDM or FDM optical fiber communication system. The control precision of an example, 16 * 2.488 Gb/s system with channel spacing (200 GHz) and the utilized F-P filter with FSR (6000 GHz) and F (300), will be calculated.

We find and verify that unidirectional surface plasmon polaritons (SPPs) are highly excited by subwavelength metallic gratings with back-side illumination around the plasmonic critical angle (PCA). The conditions and physical origins of the SPP excitation around the PCA are clarified. Moreover, the experimental results support our numerical calculations very well. Potential applications are also suggested.

A grating coupling scheme from free-space light into silicon waveguide with a remarkable property of wide flat-top angular efficiency is proposed and theoretically investigated. The coupling structure is composed of two cascaded gratings with a proper distance between their peak angular efficiencies. A quantitative semi-analytical theory based on coupled-mode models is developed for performance prediction and validated with the fully vectorial aperiodic Fourier modal method (a-FMM). With the theory, wide flat-top angular response is achieved and the conditions are pointed out. Proof-of-principle demonstrations show that the -1 dB angular width, a figure of merit to evaluate the flat-top performance, is broadened to almost 3 to 4 times, and meanwhile the -3 dB angular width, i.e., angular-full-width-half-maximum (AFWHM), is widened to nearly more than twice, compared with the reference gratings composed of the same number of periodic defects. We believe this work will find applications in biological or chemical sensing and novel optical devices.

An optically amplified WDM transmission system with 8 channels, 151 km nonregenerated transmission fiber and 2 EDFAs is reported. Good BER performance and eye diagrams are obtained with the code patterns being 140 Mb/s CMI PRBS.

The cross conversion between surface plasmon polaritons (SPPs) and quasi-cylindrical waves (CWs) is theoretically reexamined. Except for the CW-to-SPP conversion, we find the SPP-to-CW conversion, as well as the reflection and transmission of the CW, plays an indispensable role in the interaction between SPPs and light via periodic grooves. The completeness of the whole scattering coefficients is emphasized by an SPP-CW model proposed to quantitatively predict the SPP excitation efficiency for any number of periodic grooves.

A metallic nano-slit surrounded with asymmetric grooves is proposed as the plasmonic concentrator for oblique incident light. A theoretical model based on the surface plasmon polariton (SPP) coupled-mode method is derived for the extraordinary optical transmission (EOT) through such a structure under oblique incidence. The model is quantitatively validated with the finite element method. With the model, the physical insight of the EOT is then interpreted, i.e., the major contributions to the transmission include the vertical Fabry-Perot resonance of the slit, and the interference among slit modes excited by the incident light, by SPPs generated from groove arrays and their first-order reflections. This is quite different from the EOT through a nano-slit surrounded with symmetric grooves under normal incidence.

A 60-GHz bidirectional radio-over-fiber (RoF) system using two-carrier-injected distributed feedback (DFB) laser is proposed and demonstrated to realize optical single sideband (SSB) modulation for downlink. An injection-locked Fabry-P'erot laser is also carried out to realize wavelength reuse in uplink. Transmission of 2.5 Gb/s on a 60-GHz carrier for downlink and 622-Mb/s baseband signal for uplink are both successfully demonstrated over 50-km single mode fiber without chromatic dispersion compensation.

Fibre nonlinearity mitigation of non-return-to-zero on-off keying (NRZ-OOK) signals with diverse-vestigial-sideband (DVSB) filtering maximum likelihood sequence estimation (MLSE) is demonstrated by experiments and simulations, in which DVSB filtering is realised with a Mach-Zehnder (MZ) filter. In a 10.709-Gbit/s NRZ-OOK system, experimental results show that, at a fixed optical signal-to-noise ratio, the Q -factor is improved significantly with DVSB-MLSE compared to that with conventional MLSE. Simulations reveal that the performance of DVSB-MLSE can be further improved by optimising the parameters of DVSB filters for different launch powers.

The mitigation of intra-channel nonlinearities in 10 Gbaud coherent NRZ-QPSK systems via a nonlinear electrical equaliser based on Volterra theory is experimentally demonstrated. Without the requirement of prior knowledge of the fibre link, the equaliser significantly improves the system performance compared to digital dispersion compensation only.