Search

Abstract As a crucial nonlinear phenomenon, stimulated Raman scattering (SRS) plays multifaceted roles involved in forward and inverse problems. In fibre-optic systems, these roles range from detrimental interference that impairs optical performance to beneficial effects that enables various devices such as Raman amplifier. To obtain solutions of SRS, various numerical methods customized for different scenarios have been proposed. However, these methods are time-consuming, low-efficiency, and experience-orientated, particularly in combined scenarios consisting of both forward and inverse problems. Inspired by physics-informed neural networks, we propose SRS-Net, which combines the efficient automatic differentiation and powerful representation ability of neural networks with the regularization of SRS physical laws, to obtain universal solutions for SRS of forward, inverse, and combined problems. We showcase the intuitive solving procedure and high-speed performance of SRS-Net through extensive simulations covering different scenarios. Additionally, we validate its capabilities in experiments involving the high-fidelity modelling of a wavelength division multiplexing system spanning the C + L-band with approximately 10 THz. The versatility of the SRS-Net framework extends beyond SRS, indicating its potential as a promising universal solution in other engineering problems with nonlinear dynamics governed by partial differential equations.

To achieve the all-optical pattern matching for optoelectronic firewall which could process high-speed optical signals efficiently, a BPSK all-optical pattern recognition system with the data rate up to 42 Gbit/s was proposed and demonstrated by employing the commercial optical components.First, the input BPSK sequence was converted to a positive and a negative IM signals by adding in-phase and inverse-phase coherent carriers, respectively.Then the converted IM signals were coupled into time delay lines and optical AND logic gates implemented by HNLF to achieve the optical correlation operation.In the end, a high-level optical pulse indicating the pattern matched result was output.The numerical simulation and the experiment results demonstrate that the proposed system can recognize a 4-bit target BPSK sequence from an 8-bit 42 Gbit/s BPSK input sequence, which proves the feasibility of the all-optical pattern matching of BPSK signals and can be applied to high-speed optoelectronic firewall.

In recent years, Kramers-Kronig (KK) receiver has been widely studied because of its simple structure and the ability of digital signal processing (DSP) to recover complex value signals. One practical structure is detecting the optical signal with an AC-coupled photodetector. However, the direct current (DC) component of the photocurrent will be lost, in which case the KK algorithm ceases to be effective. To restore the lost DC component, we propose a simple data-aided method to estimate the DC component by the information of signal mean power stored in the designed data-aided sequences. We introduce a correction coefficient to evaluate the accuracy of our DC estimation by comparing it to the conventional DC sweeping method. Furthermore, the coefficient can be used to combat the DC estimation deviation. We carry out both simulations and experiments to study the performance of the proposed method. The results show that the DC component can be effectively recovered and the system performance is close to the results of the DC sweeping method. At the same time, in the case of low optical signal-to-noise ratio, optimization of correction coefficient can effectively improve the system performance.

The circular phased array is one of the main methods to generate radio frequency orbital angular momentum (RF-OAM) beams with high purity and capability of OAM-beams steering, but OAM beams generation and steering are only related to angle, not range dimensions. Therefore, the mode of OAM beams can be easily received in a wide range and intercepted by unintended eavesdroppers in the beam axis. As a promising solution, a range-angle dependent OAM beams transmission scheme is highlighted by the frequency diverse array (FDA) for the sake of the OAM secure transmission in this paper. In order to realize the range-angle dependent OAM beams, a special frequency interval between adjacent antenna elements in the FDA is designed and illustrated. The modes of OAM beams generated by the FDA can only be observed in the legal target location. Numerical results show that the scheme can effectively realize the safe transmission of OAM beams, i.e., reducing the possibility of the OAM beams being tapped, improving the OAM-beams transmission performance, and enhancing the energy efficiency with low complexity.

In order to maintain the security of optical network,it is urgent to research and deploy photonic firewall which can effectively process high-speed optical signals to replace existing electronic firewall which consumes high energy.After a brief review of the principle of photonic firewall,the pattern matching technology based on correlation operation and all-optical logic gates were summarized,and the matching systems was compared according to their advantages and disadvantages.Besides,the realization methods of all-optical logic gates were introduced.Finally,the next research direction and development trend of all-optical pattern matching were prospected:more efficient system structure or non-linear devices with faster response speed.The survey shows that photonic firewall based on all-optical pattern matching has a good application prospect in network security.

A chirp-rate-tunable microwave photonic pulse compression system for multi-octave linearly chirped microwave waveform (LCMW) is proposed in this paper. The key components are a dual-parallel Mach-Zehnder modulator (DPMZM) and a dispersive loop (DL) structure based on a linear chirped fiber Bragg grating. By controlling the output of the DL to change the accumulated group velocity dispersion, the chirp rate of the system can be tuned. By adjusting the wavelength of the optical carrier to select the single sideband modulated optical signal (positive or negative sideband), the chirp direction (down or up chirped) can also be tuned. The performance of the multi-octave LCMW pulse compression can be improved by eliminating the second-order harmonic distortion through configuring the bias voltages of the DPMZM. Detailed theoretical analysis is conducted. A verification simulation demonstrates the pulse compression of the multi-octave LCMWs with the chirp rates of 23.877, 11.939, 3.979, and -3.979 GHz/ns. In the proof-of-concept experiment, the matched chirp rates 23.229, 12.051, -22.568, and -11.355 GHz/ns are obtained. The multi-octave LCMW pulse compressions with the compression ratios of 14.3, 29.3, 15.0, and 28.6 are verified. The simulation and experiment results are in good agreement with the theoretical analysis.

Orbital angular momentum (OAM) beams can be generated at many microwave frequencies by using antenna array. However, the complexity of the system and inability to transmit over long distances limit the applications of OAM beams in the radio domain. Here, a significantly simplified global positioning system (GPS) ceramic antenna array is employed to generate OAM beams without the assistance of phase shifting devices. Simulation of the GPS ceramic antenna array verifies that the OAM beams with the mode of ±1 show small divergence angle and very standard spiral phase distribution. To test the transmission capability of the proposed antenna array, a measurement system that can detect the one-dimension spatial distribution of electromagnetic characteristic for the OAM beams is set up. Even considering the enormous loss in cables and power divider, the receiving antenna still can receive the electromagnetic characteristic of the OAM beams at a place more than 2 m away from the antenna array. Experimental results presented in this paper demonstrate the excellent performance of the OAM beams when generated with the proposed ceramic antenna array configuration.

A photonics-assisted super-octave microwave phase shifter is proposed. The super-octave frequency range is achieved by eliminating the second-order distortions (SDs) with the techniques of polarization interleaving and second-order optical sidebands (2-OSBs) phase processing. Polarization interleaving is based on a polarization-multiplexing dual-drive Mach-Zehnder modulator (PM-DMZM) and single sideband with carrier (SSB+C) modulation. 2-OSBs phase processing is implemented by introducing opposite phase shifts to the ±2-OSBs in a Fourier-domain optical processor (FD-OP). The proposed phase shifter is experimentally verified by proof-of-concept experiments. Experimental results show that flat phase shifts cover a full 360° range with less than 3-dB power variation over 11.5-25 GHz. Besides, excellent long-term stability is presented with a maximum phase/power fluctuation of only 1.49°/0.57 dB. Compared with the conventional FD-OP-based photonic microwave phase shifter, the proposed one improves the second-order spurious free dynamic range (SFDR2) by 13.57 dB.

With the explosion of traffic demand in recent years, the integration of satellite optical networks and terrestrial networks (ISTN) creates a promising networking solution for future low-latency, high-rate, and high-capacity communications. Owing to the high cost of deploying and maintaining a satellite optical network, it is critical to carefully design and plan the network to ensure the performance of the network. Thus, a semi-physical simulation platform based on software-defined networks (SDNs) is developed to simulate a satellite optical network and evaluate the performance of the proposed label-based advanced orbiting system (AOS) frame switching method that adheres to the Consultative Committee for Space Data Systems’ recommended standard (CCSDS). The semi-physical simulation platform has two major innovations: (1) adapting and integrating network protocols between the CCSDS and open system interconnect (OSI) reference models, particularly at the data link layer, and (2) the foundation for an SDN-based satellite optical network. In the control plane, real-time VxWorks Simulators serve as controllers to establish and manage various network protocols and the link manager protocol (LMP). Here, network protocols include open shortest path first (OSPF) for routing managing and controlling messages, constraint shortest path first–traffic engineering (CSPF-TE), and constraint-label distribution protocol (CR-LDP) for routing data services. LMP is used to assign and reserve satellite optical link resources. The performance of the architecture and protocols is evaluated via a semi-physical simulation platform.

In order to solve the data center link congestion problem,based on the characteristics of the flow distribution and flow types,a flow identification and scheduling scheme based on optical interconnect structure,named HCFD (host-controller flow detection),was proposed to identify the elephant flow which has a large impact on the network performance,and use the SDN controller to make forward strategy,and schedule the network traffic reasonably.The implementation of the scheme was to use the Netfilter framework in Linux kernel protocol on the host side to mark the flow that exceeds the threshold amount.Then,the classification model was used in the controller side to classify the marked flow.Finally,the appropriate forwarding strategy was developed based on the above results.With the advantage of the photoelectric network,mechanisms of flow depth fusion and switching could be realized.The scheme which integrates the advantage of the existing research results,was expected to identify elephant flow more accurately and comprehensively.It can effectively alleviate the network congestion,make full use of network bandwidth,reduce end-to-end delay and packet loss rate.

The sub-tree-based optical multicasting scheme provides a spectrum-efficient approach to providing emerging multicast services in optical data center networks. Moreover, multicast services are usually replicated and maintained in multiple geographically distributed data centers to improve its access efficiency and reliability. Therefore, the source data centers of all constructed sub-trees for a multicast demand are not confined to a common data center and can be independently determined by the requested distribution. In this paper, we study the problem of multicast service provisions while leveraging multicast service backups among multiple geographically distributed data centers. A novel distributed sub-tree-based optical multicasting (DST-OM) scheme is proposed. An integer linear program model is developed for the DST-OM scheme with the aim of minimizing the total spectrum consumption of all multicast demands in elastic optical data center networks. We also define the minimum spectrum sub-tree (MSST) problem for the DST-OM scheme. Two modulation-level-aware heuristic algorithms are developed to address the MSST problem. Numerical results show that the DST-OM scheme achieves higher spectrum efficiency and lower blocking probability than the conventional common source sub-tree-based optical multicasting scheme and the single-tree-based optical multicasting scheme.

To design a disaster-resilient optical data center network, the concept of k-node (edge) content connectivity has been proposed and researched. Based on this concept, the k-node (edge) content connected elastic optical data center network (KC-EODCN) has been defined. To realize the KC-EODCN, individually establishing k independent end-to-content paths for each user is a direct method with lower complexity. However, this method ignores the relevance among the users who request the same level of k-node (edge) content connectivity and causes more spectrum consumption. In this paper, a survivable k-node (edge) content connected virtual optical network (KC-VON) is defined. The users who require the same level of k-node (edge) content connectivity are organized into one KC-VON. The KC-VON satisfies the k-node (edge) content connectivity requirement that each user has at least k virtual independent end-to-content paths. The KC-VON embedding scheme is proposed instead of individually establishing k independent end-to-content paths for each user. Moreover, the KC-VON construction algorithm is developed to realize the KC-VON's construction. An integer linear program model and heuristic KC-VON embedding algorithms are developed to realize the virtual link mapping. Numerical results show that the KC-VON embedding scheme can greatly reduce the spectrum consumption more than the method of individually establishing k independent end-to-content paths for each user while realizing the KC-EODCN.

Network virtualization holds a great promise for next-generation network, because it enables sharing of physical infrastructure among different users/services, in a logical isolated way, to achieve higher efficiency and greater flexibility of resources. By using optical network of extremely large capacity, more services can be carried out, but that leads to increase in substrate failures. This problem can be addressed by using survivable virtual network embedding (SVNE), more commonly with single-path provisioning (SPP) than multipath provisioning (MPP), although the former consumes more resources. However, only a few schemes consider using one-to-multi node mapping in SVNE. Now, one-to-multi node mapping and MPP scheme can be easily implemented in SVNE for software defined network (SDN) and network function virtualization (NFV) technology. So, this paper investigates the feasibility of implementing the SVNE scheme, with MPP and one-to-multi node mapping, to achieve higher reliability and better efficiency in elastic optical network. This scheme considers four key points: 1) the number and resource requirement of physical nodes that carry one virtual node, 2) the spectrum requirement of each path of one virtual link, 3) the number of paths considering shared risk group, and 4) graph theory, which are used in the scheme. Three contrasting schemes are formulated as integer linear programming models and their simulation are carried out. The results show that the proposed survivable multipath virtual network embedding scheme against multiple failures, with SDN/NFV, achieves significant gain in terms of spectrum efficiency and survivability.

Electromagnetic waves carrying an orbital angular momentum (OAM) are of great interest. However, most OAM antennas present disadvantages such as a complicated structure, low efficiency, and large divergence angle, which prevents their practical applications. So far, there are few papers and research focuses on the problem of the divergence angle. Herein, a metasurface antenna is proposed to obtain the OAM beams with a small divergence angle. The circular arrangement and phase gradient were used to simplify the structure of the metasurface and obtain the small divergence angle, respectively. The proposed metasurface antenna presents a high transmission coefficient and effectively decreases the divergence angle of the OAM beam. All the theoretical analyses and derivation calculations were validated by both simulations and experiments. This compact structure paves the way to generate OAM beams with a small divergence angle.

Facing the growing high data rate and large communication capacity demands, optical communications are widely recognized to be used to implement satellite communications. For a space-based optical backbone network, an appropriately designed protocol stack is important. In this paper, we proposed a protocol stack that is suitable for a space-based optical backbone network. Following this, we then used software to simulate this stack, built a hardware platform to test it, and finally, analyzed the results. The results showed that the proposed protocol stack was well designed to provide efficient control and management of the space-based optical backbone network. It could improve management efficiency by collecting resources and automatically calculating and building route paths. It could also facilitate data forwarding in intermediate satellite nodes with limited source and power, by using an advanced orbiting systems (AOS) frame switching scheme to avoid unnecessary processes, such as unpacking, upper-layer processing and repacking for passing services. The protocol stack could also support the use of unidirectional links to improve the link resource utilization. Finally, it could also provide transparent transmission for different kinds of services.

This paper investigates the detection and tracing problems in a multi-relay network, where a source wants to exchange the information with a destination through multiple potential malicious relay nodes. We find that the destination could probabilistically detect the maliciousness in a multi-relay network if and only if the network satisfies a non-manipulable condition. Notice that the non-manipulable condition of the whole network is complicated to be checked. We divide the whole network into several sub-networks, and check the non-manipulability of the small-scale sub-networks in turn, which finally constitutes to a simplified checking method of the non-manipulable condition for the whole relay network. Furthermore, we propose a tracing algorithm to pinpoint the malicious relays individually in a multi-relay network if a tracing condition is satisfied. No pre-shared secret is needed in our proposed detecting and tracing schemes. Numerical examples are presented to validate the effectiveness of the proposed schemes.

A thermally tunable slot-coupled dielectric resonator antenna (DRA) has been designed and prepared by placing a thermosensitive ceramic resonator onto the slot. Typical magnetic resonance occurs in the resonator, which is closely related to its dielectric constant. Because the dielectric constant of the ceramic resonator decreases as the temperature increases, the resonance frequency of the proposed DRA increases as the temperature increases. The simulated results are in good agreement with the measured ones, which confirms the thermally tunable behavior. This approach provides a way for designing the frequency tunable antennas.

Microwave bandpass filter structure has been designed and fabricated by filling the periodically metallic apertures with dielectric particles. The microwave cannot transmit through the metallic subwavelength apertures. By filling the metallic apertures with dielectric particles, a transmission passband with insertion loss 2 dB appears at the frequency of 10-12 GHz. Both simulated and experimental results show that the passband is induced by the Mie resonance of the dielectric particles. In addition, the passband frequency can be tuned by the size and the permittivity of the dielectric particles. This approach is suitable to fabricate the microwave bandpass filters.

A magnetically tunable metamaterial perfect absorber (MPA) based on ferromagnetic resonance is experimentally and numerically demonstrated. The ferrite-based MPA is composed of an array of ferrite rods and a metallic ground plane. Frequency dependent absorption of the ferrite-based MPA under a series of applied magnetic fields is discussed. An absorption peak induced by ferromagnetic resonance appears in the range of 8–12 GHz under a certain magnetic field. Both the simulated and experimental results demonstrate that the absorption frequency of the ferrite-based MPA can be tuned by the applied magnetic field. This work provides an effective way to fabricate the magnetically tunable metamaterial perfect absorber. [ABSTRACT FROM AUTHOR]

Space-division multiplexing (SDM) has been proposed to increase fiber transmission capacity to cope with the increasing bandwidth requirements in the network. The use of multicore fiber (MCF) is considered to be a simple and feasible way to implement SDM technology. The intercore crosstalk (XT) in the MCF affects the performance of network transmission. Currently, bidirectional transmission of MCF has been proven to have lower XT than traditional unidirectional transmission of MCF. But there are few research studies on routing, core, and spectrum allocation strategies with bidirectional transmission. Thus, to get more XT reduction, we propose two different routing, core, and spectrum allocation strategies according to the traffic loads in the network based on the bidirectional transmission of MCF. The simulation results show that the proposed algorithm has lower blocking rate and higher resource utilization than the benchmark algorithm in the two stages of different traffic loads in the network. [ABSTRACT FROM AUTHOR]

We have proposed an 8 phase-shift-keying (PSK) optical millimeter-wave (MMW) signal generation scheme with frequency quintupling enabled by an electro-optical phase modulator (PM) and a wavelength selective switch. In this scheme, the radio frequency (RF) carrier with 8PSK signal is modulated on the lightwave via the PM with simple precoding but constant Euclidean distance, and is recovered to the original 8PSK signal with the frequency-quintupling RF carrier based on the phase periodicity of an 8PSK signal. By optimizing the voltage amplitude of the 8PSK RF driving signal, the generated frequency-quintupled 8PSKMMWsignal is maximized at a given laser launch power. A radio over fiber simulation link is built to demonstrate the theoretical analysis. The simulated constellations and bit-error-rate curves show that the frequency-quintupling 8PSK MMW signal has good transmission performance. The simulation results agree well with our theoretical prediction. [ABSTRACT FROM AUTHOR]

The light-tree-based optical multicasting (LT-OM) scheme provides a spectrum- and energy-efficient method to accommodate emerging multicast services. Some studies focus on the survivability technologies for LTs against a fixed number of link failures, such as single-link failure. However, a few studies involve failure probability constraints when building LTs. It is worth noting that each link of an LT plays different important roles under failure scenarios. When calculating the failure probability of an LT, the importance of its every link should be considered. We design a link importance incorporated failure probability measuring solution (LIFPMS) for multicast LTs under independent failure model and shared risk link group failure model. Based on the LIFPMS, we put forward the minimum failure probability (MFP) problem for the LT-OM scheme. Heuristic approaches are developed to address the MFP problem in elastic optical networks. Numerical results show that the LIFPMS provides an accurate metric for calculating the failure probability of multicast LTs and enhances the reliability of the LT-OM scheme while accommodating multicast services. [ABSTRACT FROM AUTHOR]

Optical network on chip (ONoC) is an attractive solution for multicore/many-core processor systems due to its high power efficiency and enormous bandwidth. However, as increasing numbers of cores need to be interconnected, the scalability of a many-core processor on a single chip is limited by its process yield and power density. A multichip architecture is, therefore, proposed to improve the scalability of ONoC. In multichip architectures, the throughput and traffic delay rely on both the intrachip and interchip networks. To exploit the advantages of multichip systems, first we propose a multichip ONoC architecture for a many-core processor system that employs a nesting ring topology. The design principles of multichip systems of different sizes are then investigated to achieve higher throughput and lower delay. These principles include the number of chips and the number of cores per chip, which are considered jointly for the first time. Finally, we evaluate the performance of the proposed architecture, implemented in 240-core and 400-core systems, respectively, and compare it to two other traditional ONoC architectures with respect to throughput and end-to-end (ETE) delay. The results show that the proposed multichip system exhibits good scalability, achieves high throughput, and provides low ETE delay. [ABSTRACT FROM AUTHOR]

A scheme for optical parallel encryption/decryption of quadrature phase shift keying (QPSK) signals is proposed, in which three QPSK signals at 10 Gb/s are encrypted and decrypted simultaneously in the optical domain through nondegenerate four-wave mixing in a highly nonlinear fiber. The results of theoretical analysis and simulations show that the scheme can perform high-speed wiretapping against the encryption of parallel signals and receiver sensitivities of encrypted signal and the decrypted signal are -25.9 and -23.8 dBm, respectively, at the forward error correction threshold. The results are useful for designing high-speed encryption/ decryption of advanced modulated signals and thus enhancing the physical layer security of optical networks. [ABSTRACT FROM AUTHOR]

As the scale of the intra-data center network (DCN) grows even larger, the traditional electrical switching has reached a bottle neck in terms of energy consumption, bandwidth provision, and end-to-end latency. Different approaches have been made by employing the optical switch instead of the electrical ones to solve the bandwidth as well as the energy efficiency and the latency problem. We propose a DCN architecture based on cascaded microelectromechanical systems switches for dynamic DCN connectivity provisioning. This architecture provides high port count, which attributes to the demands of the intradata center traffic. Multiple points to one point switching scenario is experimentally demonstrated through this data center interconnect. Numerical simulation is employed to investigate the performance of the proposed architecture. The results show that the blocking probability and latency decrease as the scale of the architecture is upgraded. [ABSTRACT FROM AUTHOR]

A system of generating and receiving orbital angular momentum (OAM) radio beams, which are collectively formed by two circular array antennas (CAAs) and effectively optimized by two intensity controlled masks, is proposed and experimentally investigated. The scheme is effective in blocking of the unwanted OAM modes and enhancing the power of received radio signals, which results in the capacity gain of system and extended transmission distance of the OAM radio beams. The operation principle of the intensity controlled masks, which can be regarded as both collimator and filter, is feasible and simple to realize. Numerical simulations of intensity and phase distributions at each key cross-sectional plane of the radio beams demonstrate the collimated results. The experimental results match well with the theoretical analysis and the receive distance of the OAM radio beam at radio frequency (RF) 20GHz is extended up to 200 times of the wavelength of the RF signals, the measured distance is 5 times of the original measured distance. C The presented proof-of-concept experiment demonstrates the feasibility of the system. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of spectrum-aware survivable strategies with failure probability constraints under static traffic in flexible bandwidth optical networks. The joint failure probability between primary and backup paths must be below the maximum tolerable joint failure probability for each traffic demand. We develop integer linear program (ILP) models for dedicated-path protection and shared-path protection in order to minimize the total number of frequency slots consumed, and we also propose a spectrum-aware dedicated protection (SADP) algorithm and a spectrum-aware shared protection (SASP) algorithm. Simulation results show that the ILP model solutions consume minimum number of frequency slots, but lead to higher average joint failure probability compared to the SADP and SASP algorithms. Moreover, both the SADP and SASP algorithms achieve a better performance in terms of total number of frequency slots consumed as compared to a conventional dedicated-path protection algorithm and a conventional shared-path protection algorithm, respectively, but lead to higher average joint failure probability. [ABSTRACT FROM PUBLISHER]

As is well known, elastic optical network is quite attractive due to its great spectral efficiency and flexibility, which can allocate appropriate size spectrum based on efficient modulation modes such as Orthogonal Frequency Division Multiplexing (OFDM) modulation, etc. These characters are able to better meet the needs of the intra-data-center optical interconnection. To properly analyze and operate elastic optical networks with highly dynamic traffic of the intra-data-center, considering the continuity constraint and the contiguity constraint, efficient methods are required for the on-line routing and spectrum allocation (RSA) issue. In this paper, we propose two methods (AFD and CSL) for assessing the status (spectrum fragmentation) of the network, and present two algorithms (BSPR and HSR) which use the proposed assessment value to perform on-line RSA applied to intra-data-center networks which adopt the Software-Defined Network (SDN) control mode. Simulation results reveal that the proposed approaches allow solving the on-line RSA problem more efficiently with lower blocking probability of the networks. [ABSTRACT FROM AUTHOR]

The capability of multicast directly in the optical domain at the routing nodes is a target for energy saving, fast service provisioning, and effective network resource utilization. An optical network node architecture supporting wavelength conversion, unicast, and multicast is presented and a shared multicast/wavelength conversion module is designed in the intermediate node. Wavelength conversion and one-to-six WDM multicast of 25-Gb/s QPSK signals based on four-wave mixing are experimentally demonstrated using a semiconductor optical amplifier with only two pumps. All six multicast signals achieve error-free performance with a power penalty <;4.3 dB. [ABSTRACT FROM PUBLISHER]

Data centers have to sustain the rapid growth of data traffic due to the increasing demand of bandwidth- hungry Internet services. The current fat tree topology causes communication bottlenecks in the server interaction process, resulting in power-hungry O-E-O conversions that limit the minimum latency and the power efficiency of these systems. As a result, recent efforts have advocated that all optical data center networks (DCNs) have the capability to adapt to traffic requirements on demand. We present the design, implementation, and evaluation of a cascaded microelectromechanical systems switches-based DCN architecture which dynamically changes its topology and link capacities, thereby achieving unprecedented flexibility to adapt to dynamic traffic patterns. We analyze it under a data center traffic model to determine its suitability for this type of environment. The proposed architecture can be scaled to 3300 input/output ports by available experimental components with low blocking probability and latency. The blocking probability and latency are about 0.03 and 72 ms at a moderate traffic load for 32 input/output ports based on our numerical results, which are much smaller than the results for 4 input/output ports which are 0.13 and 235 ms, respectively. [ABSTRACT FROM AUTHOR]

Abstract  The emergence of new services demands for a multicast function in optical networks. At the same time, wavelength converters are introduced to increase the efficiency of wavelength usage. It is because of the high cost and complex architecture of optical multicast and wavelength conversion technology, that a new switch structure is introduced, in which optical splitters and wavelength converters are shared per-node. In order to accommodate this architecture, a multicast routing and wavelength assignment algorithm in a splitter–converter-sharing optical network and a changing link weight policy to balance network traffic are proposed. By extending RSVP-TE (Resource ReSerVation Protocol-Traffic Engineering) and OSPF-TE (Open Shortest Path First-Traffic Engineering), an optical multicast mechanism is provided, and message type, signaling flow, and finite state machine model are given. Simulations of NSFNET show that, when the number of splitters and converters are 50% and 12.5% of the full equipment respectively, the performance is close to the ideal case. Using a changing link weight policy can improve performance greatly, when there are enough splitters and converters. [ABSTRACT FROM AUTHOR]