Afterburner [network-independent card for protocols]

Many current implementations of protocols such as the Transmission Control Protocol/Internet Protocol (TCP/IP) are inefficient because data are often accessed more frequently than necessary. Three techniques that reduce the need for memory bandwidth are proposed. The techniques are copy-on-write, page remapping, and single-copy. Afterburner, a network-independent card that provides the services that are necessary for a single-copy protocol stack, is described. The card has 1 MByte of local buffers and provides a simple interface to a variety of network link adapters, including HIPPI and asynchronous transfer mode (ATM). Afterburner can support transfers to and from the link adapter card at rates up to 1 Gbit/s. An implementation of TCP/IP that uses the features provided by Afterburner to reduce the movement of data to a single copy is discussed. Measurements of the end-to-end performance of Afterburner and the single-copy implementation of TCP/IP are presented     

Hidden Markov modeling using a dominant state sequence with application to speech recognition

Approximate maximum likelihood (ML) hidden Markov modeling using the most likely state sequence (MLSS) is examined and compared with the exact ML approach that considers all possible state sequences. It is shown that for any hidden Markov model (HMM), the difference between the approximate and the exact normalized likelihood functions cannot exceed the logarithm of the number of states divided by the dimension of the output vectors (frame length), which is negligible for typically used values of vector dimension (128–256) and number of states (2–30). Furthermore, for Gaussian HMMs and a given observation sequence, the MLSS is typically the sequence of nearest neighbor states in the Itakura-Saito sense, and the posterior probability of any state sequence which departs from the MLSS in a single time instant decays exponentially with the frame length. Hence, for a sufficiently large frame length the exact and approximate ML approaches provide similar model estimates and likelihood values. The results and their implications on speech recognition are demonstrated in a set of experiments.     

The effects of treatment with PIXY321 (GM-CSF/IL-3 fusion protein) on human polymorphonuclear leukocyte function

During a phase I trial of the genetically engineered hematopoietic growth factor PIXY321 (granulocyte-macrophage colony-stimulating factor/interleukin-3 [IL-3] fusion protein), we examined the effects of PIXY321 treatment on human polymorphonuclear leukocyte (PMN) locomotive, respiratory burst, and phagocytic responses. PIXY321 treatment was associated with transient suppression of both unstimulated locomotion and chemotaxis responses to multiple stimuli, as well as significant transient enhancement of formyl peptide-stimulated H2O2 production. No effects on opsonic phagocytosis of Staphylococcus aureus were observed. In vitro exposure of control PMN to PIXY321 resulted in suppression of unstimulated locomotion/chemotaxis and enhancement of formyl peptide-stimulated H2O2 production but had no effects on phagocytosis. When patient cells were exposed in vitro to PIXY321 during treatment, suppression of chemotaxis and enhancement of H2O2 production were observed before PIXY321 treatment, but these effects diminished during treatment. The in vivo and in vitro exposure effects of PIXY321 treatment on PMN function are similar to those of the parent molecule, granulocyte-macrophage colony-stimulating factor (GM-CSF).     

TennisViewer: a browser for competition trees

A tennis novice watching a match for the first time might be surprised that the crowd erupts with cheers when a player wins one point, then barely applauds when he wins the next. The crowd is not necessarily fickle; some points are genuinely more important than others because a tennis match is hierarchically structured. One match consists of several sets. One set consists of several games. One game consists of several points. The match-winning point is the most important one. How can we make that importance visible? Our goal is to let a fan, a player, or a coach examine tennis data visually, extract the interesting parts, and jump from one item to another quickly and easily. The visualization tool should help parse the elements of a match. We developed an interactive system called TennisViewer to visualize the dynamic, tree-structured data representing a tennis match. It provides an interface for users to quickly explore tennis match information. The visualization tool reveals the overall structure of the match as well as the fine details in a single screen. It uses a 2D display of translucent layers, a design that contains elements of Tree-Maps and of the Visual Scheduler system, which was designed to help faculty and students identify mutually available (transparent) time slots when arranging group meetings. TennisViewer provides MagicLens filters to explore specialized views of the information and a time-varying display to animate all or part of a match     

USE OF DIELECTRIC SPECTROSCOPY TO ASSESS ADHESIVELY BONDED COMPOSITE STRUCTURES, PART I: WATER PERMEATION IN EPOXY ADHESIVE

This three-part series of papers describes studies on water penetration into epoxy resins and the use of low- and high-frequency dielectric spectroscopy as a method of monitoring the health of adhesive bond lines in bonded epoxy/graphite composite structures. The first paper is concerned with an investigation of the effect of water ingress into the epoxy adhesive resin used in the manufacturing of the adhesive bond. This study reports a dynamic mechanical thermal analysis (DMTA) and gravimetric analysis of the effects of water uptake and interprets the data in terms of various processes that can occur within the adhesive. Surprisingly high values of water absorption were observed in samples where the edges were unconstrained and presented direct access to the fibre matrix for the moisture. The study demonstrated the effects of postcure and leaching on the sorption and desorption processes and provided background data, which are necessary for the interpretation of the study described in Parts II and III of this series of articles.     

The Solar System Modeler

The authors undertook the Solar System Modeler project to improve comprehension and appreciation of the size, complexity, and splendor of the solar system. To do so, the Solar System Modeler must (1) accurately portray the orbital behavior of satellites, planets, comets, and other celestial bodies, and (2) function in a distributed virtual environment. Additionally, the system needs to: provide a flexible, 3D graphical user interface for immersive operation; assist the user in comprehending the state of the virtual environment; accurately portray the stars and their locations; graphically model all bodies throughout the solar system in 3D and to the same scale; and maintain an interactive frame rate. They describe how they met these requirements     

Explicit Ziv-Zakai lower bound for bearing estimation

The extended Ziv-Zakai bound for vector parameters is used to develop a lower bound on the mean square error in estimating the 2-D bearing of a narrowband planewave signal using planar arrays of arbitrary geometry. The bound has a simple closed-form expression that is a function of the signal wavelength, the signal-to-noise ratio (SNR), the number of data snapshots, the number of sensors in the array, and the array configuration. Analysis of the bound suggests that there are several regions of operation, and expressions for the thresholds separating the regions are provided. In the asymptotic region where the number of snapshots and/or SNR are large, estimation errors are small, and the bound approaches the inverse Fisher information. This is the same as the asymptotic performance predicted by the local Cramer-Rao bound for each value of bearing. In the a priori performance region where the number of snapshots or SNR is small, estimation errors are distributed throughout the a priori parameter space and the bound approaches the a priori covariance. In the transition region, both small and large errors occur, and the bound varies smoothly between the two extremes. Simulations of the maximum likelihood estimator (MLE) demonstrate that the bound closely predicts the performance of the MLE in all regions