Large-scale simulation studies in image pattern recognition

Many obstacles to progress in image pattern recognition result from the fact that per-class distributions are often too irregular to be well-approximated by simple analytical functions. Simulation studies offer one way to circumvent these obstacles. We present three closely related studies of machine-printed character recognition that rely on synthetic data generated pseudo-randomly in accordance with an explicit stochastic model of document image degradations. The unusually large scale of experiments - involving several million samples that makes this methodology possible have allowed us to compute sharp estimates of the intrinsic difficulty (Bayes risk) of concrete image recognition problems, as well as the asymptotic accuracy and domain of competency of classifiers     

Magnet Safety Assessment for ITER

The ITER magnet system consists of structurally linked sets of toroidal (TF) and poloidal (PF) field coils, central solenoid (CS), and various support structures. The coils are superconducting, force flow Helium cooled with a Kapton-Glass-Epoxy multilayer insulation system. The stored magnetic energy is about 100GJ in the TF system and 20GJ in the PF-CS. Coils and structure are maintained at 4 K by enclosing them in a vacuum cryostat. The cryostat, comprising an outer envelope to the magnets, forms most of the second radioactivity confinement barrier. The inner primary barrier is formed by the vacuum vessel, its ports and their extensions. To keep the machine size within acceptable bounds, it is essential that the magnets are in close proximity to both of the nuclear confinement barriers. The objective of the magnet design is that, although local damage to one of the barriers may occur in very exceptional circumstances, large scale magnet structural or thermal failure leading to simultaneous breaching of both barriers is not credible. Magnet accidents fall into three categories: thermal (which includes arcing arising from insulation failure and local overheating due to discharge failure in the event of a superconductor quench), structural (which includes component mechanical failure arising from material inadequacies, design errors and exceptional force patterns arising from coil shorts or control failures), and fluid (Helium release due to cooling line failure). After a preliminary survey to select initial faults conceivable within the present design, these faults are systematically analyzed to provide an assessment of the damage potential. The results of this damage assessment together with an assessment of the reliability of the monitoring and protective systems, shows that the magnets can operate with the required safety condition.     

Design and performance evaluation of a banyan network basedinterconnection structure for ATM switches

This paper presents a new self-routing packet network called the plane interconnected parallel network (PIPN). In the proposed design, the traffic arriving at the network is shaped and routed through two banyan network based interconnected planes. The interconnections between the planes distribute the incoming load more homogeneously over the network. The throughput of the network under uniform and heterogeneous traffic requirements is studied analytically and by simulation. The results are compared with the results of the baseline network and another banyan network based parallel interconnection network. It is shown that, for the proposed design, a higher degree of heterogeneity results in better performance     

Asymmetric dilation of multiwavelength cross-connect switches forlow-crosstalk WDM optical networks

We predict that, for wavelength division multiplexing optical-network applications, an asymmetrically dilated configuration of a 2×2 cross-connect is significantly better in terms of overall crosstalk when the levels of the bar-port crosstalk and the cross-port crosstalk are significantly different from each other, as is the case with optical-frequency filters which utilize grating-assisted coupling. As a verification, we present a simulation study with 2×2 polarization-diversified acousto-optic tunable filters. We present a recursive method to extend the principle of asymmetric dilation to larger-size cross-connect switches, and make a recommendation for an asymmetrically dilated 4×4 cross-connect configuration     

Optimal deterministic timeouts for reliable scalable multicast

Reliable multicast protocols suffer from the problem of feedback implosion. To avoid this problem, the number of receivers sending feedback in case of loss must be small. However, losses experienced by different receivers are strongly correlated, since receivers share common resources in the multicast tree. One approach to feedback implosion avoidance relies on delaying feedback at the receivers. We present deterministic timeouts for reliable multicast (DTRM), a distributed algorithm to compute optimal deterministic timeouts for each receiver in a multicast tree as a function of the tree topology and the sender-to-receiver round-trip delays. DTRM has several desirable properties. First, feedback implosion is provably avoided for a single loss anywhere in the tree, provided delay jitter is bounded. Second, the computation of the timeouts can be entirely distributed; receivers and intermediate nodes only rely on local topology information. Third, the timeouts computed by DTRM are optimal with respect to the maximum response time     

Finite element beam propagation method for three-dimensionaloptical waveguide structures

A beam propagation method (BPM) based on the finite element method (FEM) is described for longitudinally varying three-dimensional (3-D) optical waveguides. In order to avoid nonphysical reflections from the computational window edges, the transparent boundary condition is introduced. The present algorithm using the Pade approximation is, to our knowledge, the first wide-angle finite element beam propagation method for 3-D waveguide structures. To show the validity and usefulness of this approach, numerical results are shown for Gaussian-beam excitation of a straight rib waveguide and guided-mode propagation in a Y-branching rib waveguide     

Positive temperature dependence of the electron impact ionizationcoefficient in In0.53Ga0.47As/InP HBTs

Impact ionization is a major limiting factor to the maximum operating voltage of InGaAs-based, high-speed transistors. In this work, data on the positive temperature dependence of the electron impact ionization coefficient α_n in In_0.53Ga_0.47As at medium-low electric fields are reported for the first time. The increase of α_n with temperature is opposite to the behavior normally observed in most semiconductors. This anomalous behavior implies the onset of a positive feedback between power dissipation and avalanche generation which may adversely affect the power handling capability of In_0.53Ga _0.47As-based devices, and which should be taken into account in device thermal modeling. In the experimental procedure, based on the measurement of the multiplication factor M-1 in npn In_0.53Ga _0.47As/InP Heterojunction Bipolar Transistors (HBT), particular care has been taken in order to rule out possible spurious, temperature-dependent contributions to the measured multiplication current     

A closed form theory for calibrating a dual six-port networkanalyzer

The application of a general self-calibration procedure, the so-called TNA (through, network, attenuator) method, to a dual six-port network analyzer is presented for the first time. This innovation results from strictly separating the establishment of the complex measurement ability and the determination of the analyzers system error parameters. As will be seen, the well-known TRL (through, reflect, line) method is a special case of the general TNA technique     

Application of neural networks and machine learning in networkdesign

Communication network design is becoming increasingly complex, involving making networks more usable, affordable, and reliable. To help with this, we have proposed an expert network designer (END) for configuring, modeling, simulating, and evaluating large structured computer networks, employing artificial intelligence, knowledge representation, and network simulation tools. We present a neural network/knowledge acquisition machine-learning approach to improve the END's efficiency in solving the network design problem and to extend its scope to acquire new networking technologies, learn new network design techniques, and update the specifications of existing technologies     

Densification involved in the UV-based photosensitivity of silicaglasses and optical fibers

A comprehensive survey of photosensitivity in silica glasses and optical fiber is reviewed. Recent work on understanding the mechanisms contributing to germanium or aluminum doped fiber photosensitivity is discussed within the framework of photoelastic densification models