Switch Fabric Research Articles

Block-free optical quantum Banyan network based on quantum state fusion and fission

Optical switch fabric plays an important role in building multiple-user optical quantum communication networks. Owing to its self-routing property and low complexity, a banyan network is widely used for building switch fabric. While, there is no efficient way to remove internal blocking in a banyan network in a classical way, quantum state fusion, by which the two-dimensional internal quantum states of two photons could be combined into a four-dimensional internal state of a single photon, makes it possible to solve this problem. In this paper, we convert the output mode of quantum state fusion from spatial-polarization mode into time-polarization mode. By combining modified quantum state fusion and quantum state fission with quantum Fredkin gate, we propose a practical scheme to build an optical quantum switch unit which is block free. The scheme can be extended to building more complex units, four of which are shown in this paper.

Towards zero latency photonic switching in shared memory networks

SUMMARYOptical networks on chip based on silicon photonics have been proposed to reduce latency and power consumption in future chip multiprocessors. However, high performance chip multiprocessors use a shared memory model, which generates large numbers of short messages, creating high arbitration latency overhead for photonic switching networks. In this paper, we explore techniques that intelligently use information from the memory hierarchy to predict communication in order to setup photonic circuits with reduced or eliminated arbitration latency. Firstly, we present a switch scheduling algorithm, which arbitrates on a per memory transaction basis and holds open photonic circuits to exploit temporal locality. We show that this can reduce the average arbitration latency overhead by 60% and eliminate arbitration latency altogether for up to 70% of memory transactions. We then demonstrate that this switch scheduling algorithm operating with a central photonic crossbar or Clos switch has significant energy efficiency benefits over arbitration‐free photonic networks such as single writer multiple reader networks. Finally, we demonstrate that cache miss prediction can be used to predict 86% of more complex memory transactions involving multiple nodes or main memory. Copyright © 2014 John Wiley & Sons, Ltd.

Cost evolution model to design optical switching fabrics with wavelength converters

Optical switching is expected to provide high degree of flexibility in future Internet networking. Exploitation of sub-wavelength domain based on emerging optical technology is attractive from a functional and performance perspective but its feasibility must be proven also in relation to the costs of components. In this paper a new methodology to compare optical switching matrices based on different wavelength converters sharing strategies is presented, which introduces component cost evolution in time. The methodology is applied to compare different switch configurations by taking into account possible evolution in component costs relationship in relation to technology availability. Design guidelines are drawn based on extensive simulations.

A COMBINED LOW LATENCY AND WEIGHTED FAIR QUEUING BASED SCHEDULING OF AN INPUT-QUEUED SWITCH

Input queuing has become dominant and popular building blocks for high speed crossbar switches with many ports and fast line rates because they require minimum speed-up of memory bandwidth. Input Queued switches with finite Virtual Output Queues guarantees QoS performance in terms of throughput and average delay. A switch performs two functions Queuing and Scheduling. Queue Management algorithm manages the size of the queues and drops packets when necessary or appropriate. Scheduling algorithms determine next packet to transfer and solves conflicts with the switching fabric. Fairness and Starvation are another two properties of IQ switches and it is analyzed in finite VOQ in this works. Fairness performs fair allocation of bandwidth among flows and prevents flows from misbehaving flows. Starvation of VOQ prevents serving High priority queue. The motivation behind this study is to schedule the HoL packets queued in finite VOQs by Framing with Low Latency Queuing (LLQ) and Weighted Fair Queueing (WFQ). This queueing technique of VOQ is measured in terms of throughput and average delay by fair allocation of bandwidth with WFQ and Starvation-free queue with LLQ.

Scalable Open-Source Architecture for Real-Time Monitoring of Adaptive Wiring Panels

The first prototype of an adaptive wiring panel was recently introduced that implemented a reconfigurable switch fabric that allows dynamic routing of analog, digital, and power signals for space system applications. In this paper, a complete redesign and reimplementation of the adaptive wiring panel system is considered to address issues associated with scalability, reliability, and real-time monitoring of the switching fabric. The new system is demonstrated using 48 cells as opposed to the six cells of the first adaptive wiring panel prototype. The hardware and software systems are open source, and recommendations are provided to support further extensions to the system.

In-Band Label Extractor Based on Cascaded Si Ring Resonators Enabling 160 Gb/s Optical Packet Switching Modules

Photonic integration of optical packet switching modules is crucial to compete with existing electronic switching fabrics in large data center networks. The approach of coding the forwarding packet information in an in-band label enables a spectral-efficient and scalable way of building low-latency large port count modular optical packet switching architecture. We demonstrate the error-free operation of the four in-band label extraction from 160 Gb/s optical data packets based on photonic integrated silicon-on-insulator ring resonators. Four low-loss cascaded ring resonators using the quasi-TM mode are used as narrowband filters to ensure the detection of four optical labels as well as the error-free forwarding of the payload at limited power penalty. Due to the low-loss and less-confined optical quasi-TM mode the resonators can be very narrowband and have low insertion loss. The effect of the bandwidth of the four ring resonators on the quality of the payload is investigated. We show that using four rings with 3dB bandwidth of 21 pm and only an insertion loss of 3 dB, the distortion on the payload is limited (<;1.5 dB power penalty), even when the resonances are placed very close to the packet's central wavelength. We also investigate the optical power requirements for error-free detection of the label as function of their spectral position relative to the center of the payload. The successful in-band positioning of the labels makes this component very scalable in amount of labels.

Aqueous Processing of LiNi0.5Mn0.3Co0.2O2 Composite Cathodes for Lithium-Ion Batteries

IntroductionRecently, there has been growing interest in switching fabrication of composite electrodes for lithium-ion batteries (LIBs) to aqueous systems due to advantages in low cost and environmental effect 1. Aqueous processing for conventional graphite anode is relatively mature, but remains challenging for the diverse array of LIB cathodes. Some success has been achieved on LiFePO4 2-4, LiCoO2 5 , LiNi1/3Mn1/3Co1/3O2 (NMC333) 6, lithium-and manganese-rich NMC 7. In this work, NMC532 cathodes were fabricated through aqueous processing with different water soluble binder and dried at various conditions. The effect of binders and residual moisture on the electrode performance was evaluated. Experimental NMC532 aqueous suspension was mixed with a planetary mixer and coated by a slot-die coater. The cathode consiss of 90wt% NMC532, 5 wt% Denka carbon black, 4 wt% water soluble binder, and 1 wt% carboxymethyl cellulose (CMC) as dispersant/binder. Four binders were investigated. Al current collected was treated by corona plasma for improved wettability of the aqueous suspension 8.The electrodes were also dried at various temperatures resulting in different residual moisture in the electrodes when being assembled into coin cells. Electrode performance was evaluated in both half and full coin cells. Half cells were assembled with NMC532 and Li metal as the cathodes and anodes, respectively. A12 graphite anode was selected as the anode in full cells. A Celgard 2325 separator was placed between the cathode and anode. The electrolyte was 1.2 M LiPF6in ethylene carbonate: diethyl carbonate (3/7 wt ratio). The cells were cycled between 2.5 and 4.2 V (VSP, BioLogic) for rate performance and cyclic performance. Results Figure 1 shows the NMC532 performance with various binders from half coin cells. They were similar at low C-rates and were better than that with PVDF from NMP-based processing at high C-rates. The one with latex also demonstrated identical cyclic performance to that with PVDF. This indicates comparable performance could be obtained on NMC532 from aqueous processing.Figure 2 shows the effect of drying temperature on discharge capacity of NMC. The various drying temperatures lead to different residual moisture. According to Figure 2, NMC532 dried at 120oC for 2 h showed the best performance. The residual moisture was 45 ppm at that condition. The drying condition is similar to that with conventional BMP-based processing. This indicates that it doesn’t require extra effort to remove residual moisture from electrode fabricated via aqueous processing. Aqueous processing of composite cathodes could provide comparable performance to those via conventional NMP-based processing. Acknowledgment This research at Oak Ridge National Laboratory (ORNL), managed by UT Battelle, LLC, for the U.S. Department of Energy under contract DE-AC05-00OR22725, was sponsored by the Office of Energy Efficiency and Renewable Energy Vehicle Technologies Office (VTO) Applied Battery Research 9ABR) subprogram (Program Managers: Peter Faguy and David Howell). References (1) Li, J.; Daniel, C.; Wood, D. Journal of Power Sources 2011, 196, 2452.(2) Li, J.; Armstrong, B. L.; Daniel, C.; Kiggans, J.; Wood Iii, D. L. Journal of Colloid and Interface Science 2013, 405, 118.(3) Lee, J.-H.; Kim, J.-S.; Kim, C.; Zang, D. S.; Choi, Y.-M.; Park, W. I.; Paik, U. Electrochemical and Solid-State Letters 2008, 11, A175.(4) Li, C. C.; Peng, X. W.; Lee, J. T.; Wang, F. M. Journal of the Electrochemical Society 2010, 157, A517.(5) Li, C.-C.; Lee, J.-T.; Tung, Y.-L.; Yang, C.-R. Journal of Materials Science 2007, 42, 5773.(6) Loeffler, N.; von Zamory, J.; Laszczynski, N.; Doberdo, I.; Kim, G.-T.; Passerini, S. Journal of Power Sources 2014, 248, 915.(7) Wu, Q.; Ha, S.; Prakash, J.; Dees, D. W.; Lu, W. Electrochimica Acta 2013, 114, 1.(8) Li, J.; Rulison, C.; Kiggans, J.; Daniel, C.; Wood III, D. L. J. Electrochem. Soc. 2012, 159, A1152.

Study on the Delay Guaranteed Scheduling Mechanism of the Central Stage Buffered Multiple-Plane and Multiple-Stage Packet Switching Fabric Based on Service Identification

Study on the Delay Guaranteed Scheduling Mechanism of the Central Stage Buffered Multiple-Plane and Multiple-Stage Packet Switching Fabric Based on Service Identification

Monolithic Silicon Integration of Scaled Photonic Switch Fabrics, CMOS Logic, and Device Driver Circuits

We demonstrate 4 × 4 and 8 × 8 switch fabrics in multistage topologies based on 2 × 2 Mach–Zehnder interferometer switching elements. These fabrics are integrated onto a single chip with digital CMOS logic, device drivers, thermo-optic phase tuners, and electro-optic phase modulators using IBM's 90 nm silicon integrated nanophotonics technology. We show that the various switch-and-driver systems are capable of delivering nanosecond-scale reconfiguration times, low crosstalk, compact footprints, low power dissipations, and broad spectral bandwidths. Moreover, we validate the dynamic reconfigurability of the switch fabric changing the state of the fabric using time slots with sub-100-ns durations. We further verify the integrity of high-speed data transfers under such dynamic operation. This chip-scale switching system technology may provide a compelling solution to replace some routing functionality currently implemented as bandwidth- and power-limited electronic switch chips in high-performance computing systems.

광 스위치 구조 분석 평가와 파장 변환기를 이용한 회선 경합 회피 실험

본 논문에서는 여러 가지 광 스위치들의 기능, 구조 및 장단점들을 비교 분석하여 이를 토대로 광 스위치를 선택하고, 이 스위치 모듈에 있는 파장변환기를 이용하여 블록킹을 피하기 위한 회선 경합 회피 스위칭 실험을 하였다. 스위치 패브릭의 성능을 평가하는 기본적인 기준은 넌블록킹(nonblocking), 모듈성(modularity), 업그레이드 능력(upgrade ability) 그리고 광 손실(power loss)인데 이들을 종합적으로 분석 평가하여 스위치를 선택하였다. 이렇게 선택된 스위치 패브릭은 블록킹 발생을 피하기 위한 3가지의 경우에 대하여 파장 변환을 통한 스위칭을 함으로써 회선경합 회피를 할 수 있음을 보여주었다. 이 실험의 결과는 파장 변환기의 제어를 통하여 경로를 바꾼 광신호가 출력단에서 바뀐 광파워(power)의 피크임을 보여줌으로써 라인 경합을 피하여 스위칭 되었음을 확인하였다. 또한 스위치보드에서 시간에 따른 광 파워의 변화를 분석하여 스위칭에 필요한 스위치의 채널 설정 시간과 해제 시간을 알아내었다. 이에 대한 분석 결과를 망설계시 반영하면 경제적이고 효율적인 네트워크 구조로 디자인할 수 있다. This paper presents the line contention avoidance experiments with an optical switch, which was selected based on the comparison analysis and evaluation of the various characteristics. For example, the function, structure, strengths and weaknesses of the optical switches. After considering the nonblocking, modularity, upgrade ability and optical power loss of the several kind of the switch fabrics, a switch was selected. The selected switch fabric by using wavelength converters was controlled to avoid contention of the optical lines. In this experiment shows an example of three cases. As a result of this experiment, optical signal shows a changed peek of optical power in output. By showing a peak it confirms that the contention was avoided. By analyzing of changed optical power according to the channel setting time and release time to control of the switch could be determined. If this analysis applied to the network design, economical and efficient structures can be formed.

COMPARISON OF PACKAGING TECHNOLOGIES FOR RF MEMS SWITCH

The present paper describes an integrated approach for design, fabrication and encapsulation of RF MEMS switches in view of the optimal performance subsequent to packaging. 'Top and bottom contact' fabrication approaches are explored using different RF MEMS switch topologies. In the 'bottom contact package (BCP)' the packaging cap alignment is less critical as compared to the top contact packaging (TCP) approach where contact via is an integral part of the cap. In this case, the connection layout through silicon via holes is independent of the cavity geometry. For the devices under consideration, bulk etched silicon cavity height has been optimized to 50µm for optimal RF performance, e.g., isolation and insertion loss. Parasitic effects of top silicon cap are reduced by altering CPW impedance. Mechanical parameter damping is simulated for different cavity heights and found to be independent from cavity height after 20µ mo nwards.

The Wire-Speed Multicast Switch Fabric Based on Distributive Lattice

The Wire-Speed Multicast Switch Fabric Based on Distributive Lattice

Analysis and Design of Microring-Based Switching Elements in a Silicon Photonic Integrated Transponder Aggregator

In this paper, we present and investigate a new architecture of a silicon photonic transponder aggregator as a new interconnect subsystem enabling the implementation of colorless, directionless, and contentionless ROADMs. Such subsystem is based on a microring resonator switching fabric integrated in a silicon photonics platform to achieve high functional integration together with reduction of cost, footprint, and power consumption. In the proposed device, microring resonators perform simultaneous add and drop of wavelength channels which suffer from two detrimental effects: residual dropped signal crosstalk and residual added signal crosstalk, respectively. Considering three microring-based switching elements, the transfer matrix method has been used to compute the add/drop transfer functions of the switches as a function of their geometrical parameters. The two crosstalk effects have been evaluated jointly with other important transmission parameters, such as bandwidth, insertion losses, side lobe suppression, adjacent channel rejection, extinction ratio, and group dispersion. In addition, device sensitivity with respect to the ring-waveguide coupling coefficients has been calculated. Finally, the performance of the different switches has been assessed to demonstrate that, by a proper design, the proposed transponder aggregator can support 100 Gb/s DP-QPSK modulated signal transmission.

Asymmetric Evolution of Interacting Solitons in Parity Time Symmetric Cells

We show the evolution of two fundamental bright solitons after being simultaneously launched into nonlinear triangular parity time (PT) symmetric waveguides, at mirrored transverse input positions. Numerical simulations show, whether below or above the phase transition, in general the retarded soliton propagating along the gain side and the advanced soliton evolving along the loss side of either waveguide interfere incoherently, within the interacting regime. In non-interacting regime, however, each soliton evolves as it has been launched individually. When beyond the phase transition, the unstable solitons become self-trapped after traversing relatively short distances. The incoherent interference alters the projectile of each soliton from that of the one launched individually at the same initial input position. The temporal dynamics of interacting solitons along PT symmetric waveguide, due to the interplay of nonlinearity and the active elements, are not periodic. This is in contrast to the interaction dynamics along the purely index guided cell. The results suggest that appropriately designed nonlinear PT symmetric waveguides, below or above the phase transition, can play the role of key elements in design and fabrication of optical gates, isolators, and switches.

Scalable, Low-Energy Hybrid Photonic Space Switch

A scalable monolithically integrated photonic space switch is proposed which uses a combination of Mach-Zehnder modulators and semiconductor optical amplifiers (SOAs) for improved crosstalk performance and reduced switch loss. This architecture enables the design of high-capacity, high-speed, large-port count, low-energy switches. Extremely low crosstalk of better than -50 dB can be achieved using a 2 × 2 dilated hybrid switch module. A “building block” approach is applied to make large port count optical switches possible. Detailed physical layer multiwavelength simulations are used to investigate the viability of a 64 × 64 port switch. Optical signal degradation is estimated as a function of switch size and waveguide induced crosstalk. A comparison between hybrid and SOA switching fabrics highlights the power-efficient, high-performance nature of the hybrid switch design, which consumes less than one-third of the energy of an equivalent SOA-based switch. The significantly reduced impairments resulting from this switch design enable scaling of the port count, compared to conventional SOA-based switches.

Effective Stress Modeling of Membranes Made of Gold and Aluminum Materials Used in Radio-Frequency Microelectromechanical System Switches

Microelectromechanical system switches are becoming more and more popular in the electronics industry; there is a need for careful selection of the materials in the design and fabrication of switches for reliability and performance issues. The membrane used for actuation to change the state of an RF switch is made mostly using gold or aluminum. Various designs of membranes have been proposed. Due to the flexure-type structures, the design complexity increases, which makes stress analysis mandatory to validate the reliability and performance of a switch. In this paper, the effective stress and actuation voltage required for different types of fixed-fixed membranes is analyzed using finite element modeling. Effective measures are presented to reduce the stress and voltage.

Scalable optical switch with internal fiber network

We present architecture of a large dynamic optical packet/burst switch comprising a number of smaller switching fabrics that are interconnected by internal fiber links. The switching fabrics are based on the broadcast-and-select architecture, while the internal interconnection network is a full mesh. An extensive performance study has been performed and results regarding scalability, packet/burst loss rate and achievable throughput for four scheduling algorithms are shown.

A Buffer-less Energy Consumption Optimal Model of Self-routing Grouping Switching Fabric

考虑到缓存是路由交换通信设备能耗的主要来源，该文从交换网络无缓存节能的构想出发，提出了一种自路由群组交换结构的无缓存能耗优化模型。该模型采用G倍统计线复用路由单元和Banyan类交换互连建立群组交换结构，在满足业务对阻塞率和时延性能要求的基础上，采用无内部缓存、路由单元输入输出端口数扩展和级间互连线优化设计来达到节能的目的。相同网络性能下，缓存能耗和结构能耗的量化比较表明该无缓存结构能耗小。仿真实验结果证明：结构参数G越大，网络级数k越小，路由单元能耗越小，在Banyan类交换互连结构中，分治(Divide-and-Conquer, DC)互连方式的级连线能耗最小。

GreenRouter: Reducing Power by Innovating Router's Architecture

High speed routers in Internet are becoming more powerful, as well as more energy hungry. In this paper, we present a new architecture of router named GreenRouter which separates a line-card into two parts: network interface card (DB) and packet processing card (MB), connected by a two-stage switch fabric in traffic flows' ingress and egress direction respectively. Traffic from all DBs shares all the MBs in GreenRouter, thus can be aggregated to a few active MBs on demand and other MBs can be shut down to save power. Several key issues to this new architecture are addressed. We evaluate the power saving efficiency and give preliminary simulation results. GreenRouter can well adapt the traffic fluctuation and real trace evaluations over one week shows that up to 63.7% power saving can be achieved while QoS constraints are guaranteed.

RF MEMS Switches With $\hbox{RuO}_{2}$ –$\hbox{Au}$ Contacts Cycled to 10 Billion Cycles

We present improvements in RF microelectromechanical switch design and fabrication that demonstrated improved lifetimes in cycled switches. First, implementation of RuO2-Au contact metallurgy into an existing design showed improved switch lifetime over switches with Pt-Au, Ir-Au, and Au-Au contacts. Second, the switch design was changed to reduce impact upon switch closure, and the fabrication process was changed to avoid the use of polymer sacrificial materials while including the RuO2-Au contact metallurgy. Switches with the new design were cycled to 10 billion cycles with a resistance less than 4 Ω, an insertion loss of 0.4 dB, and an isolation of 28.0 dB at 10 GHz. We propose that the catalytic behavior of the RuO2 film prevents or delays the failure of the switches due to accumulation of carbon at the contacts. Additionally, the reduced impact upon closure prevented significant contact evolution during cycling.