MapReduce programming model has simplified the implementation of many data parallel applications. The simplicity and quality services provided by implementations attract a lot enthusiasm among distributed computing communities. From years experience in applying to various scientific applications we identified set extensions improvements its architecture that will expand applicability more classes In this paper, present Twister an enhanced runtime supports iterative computations efficiently....

Abstract To solve the problem of huge channel state information (CSI)data caused by a large number antennas in large-scale MIMO systems,a CSI compressed feedback network model (CFNet)with encoder-decoder structure is proposed this paper. Specifically, encoder efficiently compresses data to reduce burden transmission and storage. A key component, Asymmetric Expansion Convolution (AEC) module, combines asymmetric convolution dilated extract feature from different directions, thereby increasing...

Artificial intelligence (AI) for fluid mechanics has become attractive topic. High-fidelity data is one of most critical issues the successful applications AI in mechanics, however, it expensively obtained or even inaccessible. This study proposes a high-efficient forward generation method from partial differential equations (PDEs). Specifically, solutions PDEs are first generated either following random field (e.g. Gaussian field, GRF, computational complexity O(NlogN), N number spatial...

In order to provide fault-tolerance and guarantee reliability, data redundancy should be introduced in distributed storage systems. The emerging coding techniques for such as fractional repetition codes, the required more efficiently than conventional replication scheme. this letter, we extend construction of codes present a new scheme, termed general codes. proposed can applied systems which capacities nodes may different. Based on combinatorial structure known group divisible design, code...

We report here a polarization-independent drop filter (PIDF) based on photonic crystal self-collimation ring resonator (SCRR). Despite of the large birefringence associated with polarization-dependent dispersion properties, we demonstrate PIDF multiple-beam interference theory and polarization peak matching (PPM) technique. The performance was also investigated finite-difference time-domain (FDTD) technique, excellent agreement between simulation. For designed wavelength 1550 nm, free...

Although catalytic dehydrogenative aromatization from cyclohexanones and NH3 is an attractive synthetic method for primary anilines, using a hydrogen acceptor was indispensable to achieve satisfactory levels of selectivity in liquid-phase organic systems without photoirradiation. In this study, we developed highly selective synthesis anilines via efficient acceptorless heterogeneously catalyzed by Mg(OH)2-supported Pd nanoparticle catalyst which Mg(OH)2 species are also deposited on the...

Novel opal hydrogels with water-tunable photonic bandgap (PBG) exhibiting responses to external stimuli were self-assembled from polystyrene-co-poly(N,N-dimethylacrylamide) (PS-co-PDMAA) microspheres. The polymeric microspheres narrow size distribution successfully prepared in water, consisting of two regions. inner region is rich PS which hard and hydrophobic; the outer PDMAA soft hydrophilic. self-assembly PS-co-PDMAA hydrogel ready induced by centrifugation resulted highly ordered...

Regenerating code is a class of storage codes that achieve the optimal trade-off between capacity and repair bandwidth, which are two important performance metrics in data systems. However, existing constructions regenerating rely on expensive computational operations such as finite field multiplication. The high coding complexity limit their applications large-scale practical In this paper, we show it possible to full potential with low complexity. particular, propose new codes, called...

The key current challenges for the industrial application of all optical switching networks are energy consumption, transmission rate, spectrum efficiency, and throughput. consumption problem is mainly researched in this paper. From perspective components modules, node equipment, network levels, different enabling technologies proposed to overcome problem, which also evaluated through experimental demonstrations. First, high-sampling-rate digital-to-analog converters (DACs) WSS-based ROADM...

A metasurface composed of regularly arranged silicon (Si) nanospheres (NSs) with coupling was investigated both theoretically and numerically based on the Mie theory, simple Lorentz line shape model finite-difference time-domain technique. By deliberately controlling strength between Si NSs through design lattice constants a rectangular lattice, polarization beam splitters, converters analyzers good performance can be successfully constructed. square as well large incidence angle employed to...

Mobile ad hoc networks are self configuring wireless without any fixed infrastructure. The members within such a network move independently from each other and their movements not restricted in case (in car to ad-hoc nodes partly drive the same direction). Hence, it is impossible predict routes, link expiration or failure times. We describe simplified scenario with two stationary terminals one terminal motion. This simple approach allows statistical derivation forecast average distance...

The next generation of MPSoC points to the integration thousands IP cores, requiring high performance interconnect for throughput communications. Optical on-chip enables significantly increased bandwidth and decreased latency in MPSoC. However, interface between electrical photonic devices implies strong layout constraints that may impact system scalability. In this paper, we propose a novel optical named Chameleon. simplifies allows cores be adapted according communication requirements....

SUMMARY The many‐core design research community has shown high interest in optical crossbar on chip for more than a decade. Key properties of crossbars, namely (1) contention‐free data routing, (2) low latency communication, and (3) potential bandwidth through the use wavelength division multiplexing, motivate several implementations this type interconnect. These demonstrate very different scalability power efficiency abilities depending three key factors: network topology, considered...

Optical Network-on-Chip (ONoC) is an emerging technology considered as one of the key solutions for future generation on-chip interconnects. However, silicon photonic devices in ONoC are highly sensitive to temperature variation, which leads a lower efficiency Vertical-Cavity Surface-Emitting Lasers (VCSELs), resonant wavelength shift Microring Resonators (MR), and results Signal Noise Ratio (SNR). In this paper, we propose methodology enabling thermal-aware design optical interconnects...

Optical Network-on-Chip (ONoC) is an emerging technology considered as one of the key solutions for future generation on-chip interconnects. However, silicon photonic devices in ONoC are highly sensitive to temperature variation, which leads a lower efficiency Vertical-Cavity Surface-Emitting Lasers (VCSELs), resonant wavelength shift Microring Resonators (MR), and results Signal Noise Ratio (SNR). In this paper, we propose methodology enabling thermal-aware design optical interconnects...

Mobile radio systems beyond the third generation will evolve into all-IP systems, integrating Internet and mobile system advantages. The BRAIN project is developing a architecture which combines local coverage broadband access based on HIPERLAN/2 with several wider-coverage enabling full of seamless IP-based services for users in hot spot areas move. End-to-end QoS provision one major challenges design such must be supported by application, network, wireless layers. This article proposes...

Simulation studies of grid scheduling strategies require representative workloads to produce dependable results. Real production have shown diverse correlation structures and scaling behavior, which are different than the characteristics available supercomputer cannot be captured by Poisson or simple distribution-based models. We present models that able reproduce various structures, including pseudo-periodicity long range dependence. By conducting model-driven simulation, we quantitatively...