This paper covers new architectures, technologies, and performance benchmarking together with prospects for high productivity and high performance computing enabled by photonics. The exponential and sustained increases in computing and data center needs are driving the demands for exascale computing in the future. Power-efficient and parallel computing with balanced system design is essential for reaching that goal as should support ∼billion total concurrencies and ∼billion core interconnections with ∼exabyte/second bisection bandwidth. Photonic interconnects offer a disruptive technology solution that fundamentally changes the computing architectural design considerations. Optics provide ultra-high throughput, massive parallelism, minimal access latencies, and low power dissipation that remains independent of capacity and distance. In addition to the energy efficiency and many of the fundamental physical problems, optics will bring high productivity computing where programmers can ignore locality between billions of processors and memory where data resides. Repeaterless interconnection links across the entire computing system and all-to-all massively parallel interconnection switch will significantly transform not only the hardware aspects of computing but the way people program and harness the computing capability. This impacts programmability and productivity of computing. Benchmarking and optimization of the configuration of the computing system is very important. Practical and scalable deployment of photonic interconnected computing systems are likely to be aided by emergence of athermal silicon photonics and hybrid integration technologies.

Triggered by the Great East Japan Earthquake in March 2011, the authors have been studying a resilient network whose key element is a movable and deployable ICT resource unit. The resilient network needs a function of robust and immediate connection to a wide area network active outside the damaged area. This paper proposes an application of digital coherent technology for establishing optical interconnection between the movable ICT resource unit and existing network nodes through a photonic network, rapidly, easily and with the minimum in manual work. We develop a prototype of a 100Gbit/s digital coherent transponder which is installable to our movable and deployable ICT resource unit and experimentally confirm the robust and immediate connection by virtue of the plug and play function.

An optoelectronic 32-bit serial-to-parallel converter with a novel conversion scheme and shared-trigger configuration has been developed for the label processing of 100-Gbps (25-Gbps $ imes 4 lambda)$ optical packets. No external optical trigger source is required to operate the converter, as the optical packet itself is used to perform self-triggering. Compared to prior optoelectronic label converters, the new device has a much higher gain even while converting labels at higher data rates, and exhibits tolerance to the voltage swing of received packets. The device response is presented together with the experimental demonstration of serial-to-parallel conversion for 4 different labels at 25 Gbps.

We describe wavelength-routed switching technology for 25-Gbit/s optical packets using a tunable transmitter that monolithically integrates a parallel-ring-resonator tunable laser and an InGaAlAs electro-absorption modulator (EAM). The transmitter provided accurate wavelength tunability with 100-GHz spacing and small output power variation. A 25-Gbit/s burst-mode optical-packet data was encoded onto the laser output by modulating the integrated EAM with a constant voltage swing of 2 V at 45$^{circ}$C. Clear eye openings were observed at the output of the 100 GHz-spaced arrayed-waveguide grating with error-free operation being achieved for all packets. The tunable transmitter is very promising for realizing a high-speed, large-port-count and energy-efficient wavelength-routing switch that enables the forwarding of 100-Gbit/s optical packets.

Graphene is attracting attention in electrical and optical research fields recently. We measured the optical absorption characteristics and polarization dependence of single-layer graphene (SLG) on sub-micrometer Si waveguide. The results for graphene lengths ranging from 2.5 to 200 $mu$ m reveal that the optical absorption by graphene is 0.09 dB/$mu$ m with the TE mode and 0.05 dB/$mu$ m with the TM mode. The absorption in the TE mode is 1.8 times higher than that in the TM mode. An optical spectrum, theoretical analysis and Raman spectrum indicate that surface-plasmon polaritons in graphene support TM mode light propagation.

In this invited paper, software defined network (SDN)-based approaches for future cost-effective optical mobile backhaul (MBH) networks are discussed, focusing on key principles, throughput optimization and dynamic service provisioning as its use cases. We propose a novel physical-layer aware throughput optimization algorithm that confirms > 100Mb/s end-to-end per-cell throughputs with ≥2.5Gb/s optical links deployed at legacy cell sites. We also demonstrate the first optical line terminal (OLT)-side optical Nyquist filtering of legacy 10G on-off-keying (OOK) signals, enabling dynamic >10Gb/s Orthogonal Frequency Domain Multiple Access (OFDMA) λ-overlays for MBH over passive optical network (PON) with 40-km transmission distances and 1:128 splitting ratios, without any ONU-side equipment upgrades. The software defined flexible optical access network architecture described in this paper is thus highly promising for future MBH networks.

An optical flip-flop circuit with a single semiconductor optical amplifier (SOA) using two orthogonal polarization states is proposed. The optical set / reset input and output signals are at a single wavelength. The flip-flop circuit consists of an SOA, a polarization combiner, a polarization splitter, two directional couplers, and two phase shifters. No continuous light source is required to operate the circuit. In this paper, we theoretically analyze the operation performance. Polarization dependence in SOA is considered in the analysis at a single wavelength operation, and numerically simulated results are presented. We confirm that the flip-flop circuit with a feedback-loop length of 15~mm can be operated at switching time of around 3~ns by 1~ns set / reset pulses. The flip-flop performance is discussed from viewpoints of transient overshoot and contrast at the steady on-off states.

In conventional networks, service control function and network control function work independently. Therefore, stereotypical services are provided via fixed routes or selected routes in advance. Recently, advanced network services have been provided by assortment of distributed components at low cost. Furthermore, service platform, which unifies componentized network control and service control in order to provide advanced services with flexibility and stability, has attracted attention. In near future, network management system (NMS) is promising, which replies an answer quickly for such advanced service platforms when route setting is requested with some parameters: quality of service (QoS), source and destination addresses, cost (money) and so on. In addition, the NMS is required to provide routes exploiting functions such as path computation element (PCE) actually. This paper proposes scalable network architecture that can quickly reply an answer by pre-computing candidate routes when route setting is requested to a control unit like an Autonomous System (AS). Proposed architecture can manage network resources scalably, and answer the availability of the requested QoS-aware path settings instantaneously for the forthcoming service platform that finds an adequate combination of a server and a route. In the proposed method, hierarchical databases are established to manage the information related to optical network solution and their data are updated at fewer times by discretized states and their boundaries with some margin. Moreover, with multiple and overlapped overlay, it pre-computes multiple candidate routes with different characteristics like available bandwidth and the number of hops, latency, BER (bit error rate), before route set-up request comes. We present simulation results to verify the benefits of our proposed system. Then, we implement its prototype using OpenFlow, and evaluate its effectiveness in the experimental environment.

Many detailed studies ranging from networking to hardware as well as standardization activities over the last few years have advanced the performance of the elastic optical network. Thanks to these intensive works, the elastic optical network has been becoming feasible. This paper reviews the recent advances in the elastic optical network from the aspects of networking technology and hardware design. For the former, we focus on the efficient elastic network design technology related to routing and spectrum assignment (RSA) of elastic optical paths including network optimization or standardization activities, and for the latter, two key enabling technologies are discussed: elastic transponders/regenerators and gridless optical switches. Making closely-dependent networking and hardware technologies work synergistically is the key factor in implementing truly effective elastic optical networks.

For next generation planar lightwave circuit (PLC) devices, high function and high-density integration are required as well as downsizing and cost reduction. To realize these needs, high refractive index difference between a core and a clad $(Delta)$ is required. To use PLC for practical applications, silica-based PLC is one of the most attractive candidate. However, degradation of the optical properties and productivity occur when $Delta$ of the core becomes high. Thus, $Delta$ of most of the conventional PLC with GeO$_2$-SiO$_2$ core is designed less than 2.5%. In this paper, we report a silica-based ultra-high $Delta $ PLC with ZrO$_2$-SiO$_2$ core. 5.5%-$Delta$ ZrO$_2$-SiO$_2$ PLC has been realized with low propagation loss and basic characteristics has been confirmed. Potential of chip size reduction of the ZrO$_2$-SiO$_2$ PLC is shown.

In this paper we summarize the work conducted in our group in the area of E- and W-band optical high-capacity fiber-wireless links. We present performance evaluations of E- and W-band mm-wave signal generation using photonic frequency upconversion employing both VCSELs and ECLs, along with transmission over different type of optical fibers and for a number of values for the wireless link distance. Hybrid wireless-optical links can be composed of mature and resilient technology available off-the-shelf, and provide functionalities that can add value to optical access networks, specifically in mobile backhaul/fronthaul applications, dense distributed antenna systems and fiber-over-radio scenarios.

In this paper, we investigate a combination scheme of subcarrier intensity-modulation (SIM) with spatial modulation (SM) for optical wireless communication. Using computer simulation, the performances of the proposed SIM/SM scheme are investigated and compared with those of the conventional SIM scheme in the additive white gaussian noise (AWGN) as well as in outdoor environment with turbulence induced fading characteristics. Numerical results show that the proposed SIM/SM scheme can outperform the conventional SIM in an environment with different spectral efficiencies. When the spectral efficiency is varied from 2bits/s/Hz to 4bits/s/Hz, an Eb/N0 gain of 2dB to 5dB is achieved, when the bit error rate of 10-5 is maintained. It shows that the employment of SM may further improve the power efficiency of SIM, when the number of subcarriers increases according to the spectral efficiency. When the spectral efficiency is 4bits/s/Hz, the SIM/SM scheme for 0.5 of log-irradiance variance in the log-normal turbulence channel shows the same performance as SIM with variance of 0.3. This means that the SIM/SM can be an alternative choice in even worse environments.

The stochastic behavior of inter-core crosstalk in multi-core fiber is discussed based on a theoretical model validated by measurements, and the effect of the crosstalk on the Q-factor in transmission systems, using multi-core fiber is investigated theoretically. The measurements show that the crosstalk rapidly changes with wavelength, and gradually changes with time, in obedience to the Gaussian distribution in I-Q planes. Therefore, the behavior of the crosstalk as a noise may depend on the bandwidth of the signal light. If the bandwidth is adequately broad, the crosstalk may behave as a virtual additive white Gaussian noise on I-Q planes, and the Q-penalty at the Q-factor of 9.8dB is less than 1dB when the statistical mean of the crosstalk from other cores is less than -16.7dB for PDM-QPSK, -23.7dB for PDM-16QAM, and -29.9dB for PDM-64QAM. If the bandwidth is adequately narrow, the crosstalk may behave as virtually static coupling that changes very gradually with time and heavily depends on the wavelength. To cope with a static crosstalk much higher than its statistical mean, a margin of several decibels from the mean crosstalk may be necessary for suppressing Q-penalty in the case of adequately narrow bandwidth.

Optical processing of optical labels is expected for increasing processing speed in network routers. We previously proposed optical waveguide circuits for recognition of optical QAM codes by detecting a null output port. The circuits are based on a recognition circuit for QPSK codes. In the device, however, optical or electrical inverters with large dynamic range are required. In this paper, we propose optical circuits to recognize optical QAM codes by maximum output with a post-processor consisting of thresholders and logical circuits. The recognition function of the waveguide circuit is numerically proved by FD BPM.

A reduction in the intensity deviation of a nine-channel optical frequency comb block (OFCB) is demonstrated, by adopting an asymmetric differential drive method for an InP-based dual drive Mach-Zehnder modulator. The generation of a tailored OFCB with an intensity deviation of less than 0.8dB is confirmed by using the modulator.

We have designed a novel fiber-optic light concentrator with scattering layers and evaluated the light concentration characteristics by ray-trace simulations as functions of the parameters of the incident light angle and wavelength, as well as the waveguide structure. Unlike well-known luminescent solar concentrators, in our models, illuminating light is directly captured through the proposed waveguide structure. The optical efficiency in our fiber-optic models is remarkably improved in long-length regions compared with that in simple slab waveguides. In addition, the waveguide length required to effectively collect light is extended to 300mm and 1.5m for optical fibers with 1- and 10-mm core diameters, respectively, which are ten times longer than those in slab waveguides with an equivalent scale. Because of the cylindrical structure of optical fibers, we have also evaluated the sensitivity of our models to surrounding light. Consequently, an obvious directional property containing single or three peaks of the sensitivity is clarified, and their widths can be tuned by changing the width of the scattering parts. These results suggest that our models are suited for sensor devices such as optical receiving antennas, rather than simple light concentrators. Finally, we model a fiber-optic probe as an application and evaluate the light concentration characteristics when the concentrator is serially concatenated with a normal optical fiber.

We analyzed polarization characteristics of gammadion-shaped planar chiral nano-gratings (PCNGs), using Jones matrix and FDTD simulation. Optical activity (OA) was found to take place at wavelengths where long-lifetime modes appeared in the chiral layer. Among two kinds of resonance phenomena that concern the extension of the lifetime, guided-mode resonance and Fabry-Perot resonance, the latter was found to be a key to generate practically-important, broad peaks in the OA spectrum. Through the calculation of dispersion relations of Bloch modes in the chiral layer, we showed that the interference of multiple modes with group velocity dispersion played a critical role in the generation of such long-lifetime modes.

We theoretically study the performance of multiple-input multiple-output (MIMO) free-space optical (FSO) systems using subcarrier quadrature modulation (SC-QAM) signaling. The system average symbol-error rate (ASER) is derived taking into account the atmospheric turbulence effects on the MIMO/FSO channel, which is modeled by log-normal and the gamma-gamma distributions for weak and moderate-to-strong turbulence conditions. We quantitatively discuss the influence of index of refraction structure parameter, link distance, and different MIMO configurations on the system ASER. We also analytically derive and discuss the MIMO/FSO average (ergodic) channel capacity (ACC), which is expressed in terms of average spectral efficiency (ASE), under the impact of various channel conditions. Monte Carlo simulations are also performed to validate the mathematical analysis, and a good agreement between numerical and simulation results is confirmed.

Optical Network-on-Chip (ONoC) is a promising emerging technology, which can solve the bottlenecks faced by electrical on-chip interconnection. However, the existing proposals of ONoC are mostly built on fixed topologies, which are not flexible enough to support various applications. To make full use of the limited resource and provide a more efficient approach for resource allocation, RONoC (Reconfigurable Optical Network-on-Chip) is proposed in this letter. The topology can be reconfigured to meet the requirement of different applications. An 8×8 nonblocking router is also designed, together with the communication mechanism. The simulation results show that the saturation load of RONoC is 2 times better than mesh, and the energy consumption is 25% lower than mesh.

Silicon photonics Network-on-Chips (NoCs) have emerged as an attractive solution to alleviate the high power consumption of traditional electronic interconnects. In this paper, we propose a fully optical ring NoC that combines static and dynamic wavelength allocation communication mechanisms. A different wavelength-channel is statically allocated to each destination node for light weight communication. Contention of simultaneous communication requests from multiple source nodes to the destination is solved by a token based arbitration for the particular wavelength-channel. For heavy load communication, a multiwavelength-channel is available by requesting it in execution time from source node to a special node that manages dynamic allocation of the shared multiwavelength-channel among all nodes. We combine these static and dynamic communication mechanisms in a same network that introduces selection techniques based on message size and congestion information. Using a photonic NoC simulator based on Phoenixsim, we evaluate our architecture under uniform random, neighbor, and hotspot traffic patterns. Simulation results show that our proposed fully optical ring NoC presents a good performance by utilizing adequate static and dynamic channels based on the selection techniques. We also show that our architecture can reduce by more than half, the energy consumption necessary for arbitration compared to hybrid photonic ring and mesh NoCs. A comparison with several previous works in term of architecture hardware cost shows that our architecture can be an attractive cost-performance efficient interconnection infrastructure for future SoCs and CMPs.