Architectural Alternatives for Information Filtering in Structured Overlays

Content providers are naturally distributed and produce large amounts of new information every day. Peer-to-peer information filtering is a promising approach that offers scalability, adaptivity to high dynamics, and failure resilience. The authors developed two approaches that utilize the chord distributed hash table as the routing substrate, but one stresses retrieval effectiveness, whereas the other relaxes recall guarantees to achieve lower message traffic and thus better scalability. This article highlights the two approaches' main characteristics, presents the issues and trade-offs involved in their design, and compares them in terms of scalability, efficiency, and filtering effectiveness.     

Joint multi-cost routing and power control in wireless ad hoc networks

In this work we study the combination of multi-cost routing and adjustable transmission power in wireless ad hoc networks, so as to obtain dynamic energy- and interference-efficient routes to optimize network performance. In multi-cost routing, a vector of cost parameters is assigned to each network link, from which the cost vectors of candidate paths are calculated. Only at the end these parameters are combined in various optimization functions, corresponding to different routing algorithms, for selecting the optimal path. The multi-cost routing problem is a generalization of the multi-constrained problem, where no constraints exist, and is also significantly more powerful than single-cost routing. Since energy is an important limitation of wireless communications, the cost parameters considered are the number of hops, the interference caused, the residual energy and the transmission power of the nodes on the path; other parameters could also be included, as desired. We assume that nodes can use power control to adjust their transmission power to the desired level. The experiments conducted show that the combination of multi-cost routing and adjustable transmission power can lead to reduced interference and energy consumption, improving network performance and lifetime.     

Catalytic conversion of lignocellulosic biomass to fuels: Process development and technoeconomic evaluation

Levulinic acid (LA) has been identified as a platform chemical, which can be produced from lignocellulosic materials and transformed into liquid fuels, fuel additives and even other specialty chemicals. These conversions have been made possible through recent advances in heterogeneous catalysis. Taking advantage of novel chemistries and catalytic materials, we have developed a LA-based strategy to convert lignocellulosic biomass into liquid hydrocarbon fuels. To assess the economic potential of this approach, a process synthesis effort supported by detailed process simulation and capital/operational cost calculations has been undertaken. Furthermore, we study different feedstocks and perform sensitivity analysis studies for several process and economic parameters. Finally, we present the results of an energy efficiency analysis and discuss biomass transportation aspects.     

State-of-the-art graphene high-frequency electronics

High-performance graphene transistors for radio frequency applications have received much attention and significant progress has been achieved. However, devices based on large-area synthetic graphene, which have direct technological relevance, are still typically outperformed by those based on mechanically exfoliated graphene. Here, we report devices with intrinsic cutoff frequency above 300 GHz, based on both wafer-scale CVD grown graphene and epitaxial graphene on SiC, thus surpassing previous records on any graphene material. We also demonstrate devices with optimized architecture exhibiting voltage and power gains reaching 20 dB and a wafer-scale integrated graphene amplifier circuit with voltage amplification.     

JointMMCC: joint maximum-margin classification and clustering of imaging data

A number of conditions are characterized by pathologies that form continuous or nearly-continuous spectra spanning from the absence of pathology to very pronounced pathological changes (e.g., normal aging, mild cognitive impairment, Alzheimer's). Moreover, diseases are often highly heterogeneous with a number of diagnostic subcategories or subconditions lying within the spectra (e.g., autism spectrum disorder, schizophrenia). Discovering coherent subpopulations of subjects within the spectrum of pathological changes may further our understanding of diseases, and potentially identify subconditions that require alternative or modified treatment options. In this paper, we propose an approach that aims at identifying coherent subpopulations with respect to the underlying MRI in the scenario where the condition is heterogeneous and pathological changes form a continuous spectrum. We describe a joint maximum-margin classification and clustering (JointMMCC) approach that jointly detects the pathologic population via semi-supervised classification, as well as disentangles heterogeneity of the pathological cohort by solving a clustering subproblem. We propose an efficient solution to the nonconvex optimization problem associated with JointMMCC. We apply our proposed approach to an medical resonance imaging study of aging, and identify coherent subpopulations (i.e., clusters) of cognitively less stable adults.     

QoS experiences in native IPv6 networks

Deployment of IPv6 technology in research and commercial networks has accelerated in the last few years. Inevitably, as more advanced services take advantage of the new technology, IPv6 traffic gradually increases. Today, there is limited experience in the deployment of Quality of Service (QoS) for IPv6 traffic in backbone networks that support the Differentiated Services framework. As available software and hardware are designed to handle IPv4 packets, there is a need to accurately measure and validate performance of QoS mechanisms in an IPv6 environment. This paper discusses tests and technical challenges in the deployment of IPv6 QoS in core networks, namely the production dual stack gigabit‐speed Greek Research and Education Network (GRNET) and the IPv6‐only 6NET European test network, using both hardware and software platforms. In either case, we succeeded in delivering advanced transport services to IPv6 traffic and provided different performance guarantees to portions of traffic. The deployed QoS schema was common to IPv6 and IPv4; in most cases both v4 and v6 traffic exhibited comparable performance per class, while imposing no significantly different overhead on network elements. A major conclusion of our tests is that the IPv6 QoS mechanisms are efficiently supported with state‐of‐the‐art router cards at gigabit speeds. Copyright © 2008 John Wiley & Sons, Ltd.     

Generating realistic laminate fiber angle distributions for optimal variable stiffness laminates

The advent of advanced fiber placement technology has made it possible, through the use of fiber steering, to exploit the anisotropic properties of composite materials to a larger extent than was previously possible. Spatial variation of stiffness can be induced by steering composite fibers in curvilinear paths to give beneficial load and stiffness distribution patterns. Buckling of composite panels is one area where fiber steering has been proven to be very effective. Fiber angles and predefined fiber angle variations are used in most of the research on fiber steered composites reported in the literature, however, from an optimization point of view it is attractive to design such variable stiffness (VS) structures in terms of lamination parameters (LPs). This results in a two-step design approach. In the first step a VS composite is designed in terms of LPs, and in the second step the LPs are converted into fiber angle distributions for each layer in the laminate. A methodology is proposed to convert a known LP distribution for a VS composite laminate into a realistic design in terms of fiber angles, with minimum loss of structural performance, whilst satisfying a constraint on in-plane fiber angle curvature. The proposed conversion process is formulated as an optimization problem and can be used for any number of equi-thickness plies. The methodology was tested by converting a known optimal LP design for a sample structure, a square plate under bi-axial compression into a fiber angle design. The effect of the in-plane curvature constraint, the number of layers in the laminate, and the choice of objective function for the conversion process were studied for a balanced symmetric lay-up.     

Autoregressive energy-efficient context forwarding in wireless sensor networks for pervasive healthcare systems

Driven by technological advances in low-power network systems and medical sensors, we have witnessed during the recent years the adoption of wireless sensor networks (WSNs) in electronic healthcare. Improving the quality of electronic healthcare and the prospects of ‘ageing in place’ through WSNs requires solving difficult problems in scale, energy management, and data acquisition. Medical and pervasive healthcare application (or mobile healthcare application) based on WSNs is influenced by many factors such as transmission errors and power consumption. We propose a multivariate context forwarding model that achieves energy-efficient WSN operation. A node adopts multivariate autoregression for forecasting contextual information (bio-signals or vital parameters) and locally decides whether context retransmission is required or not. This scheme is applied in patient telemonitoring systems where accurate yet energy-aware transmission of bio-signals to a remote control unit is crucial. Simulation results are reported indicating the capability of the proposed model in minimizing energy consumption in WSNs having as application domain the electronic healthcare systems.     

Tight bounds for oblivious routing in the hypercube

We prove that in anyN-node communication network with maximum degreed, any deterministic oblivious algorithm for routing an arbitrary permutation requires (N/d) parallel communication steps in the worst case. This is an improvement upon the (N/d ^3/2) bound obtained by Borodin and Hopcroft. For theN-node hypercube, in particular, we show a matching upper bound by exhibiting a deterministic oblivious algorithm that routes any permutation in (N/logN) steps. The best previously known upper bound was (N). Our algorithm may be practical for smallN (up to about 2¹⁴ nodes).C. Kaklamanis was supported in part by NSF Grant NSF-CCR-87-04513. T. Tsantilas was supported in part by NSF Grants NSF-DCR-86-00379 and NSF-CCR-89-02500.     

Functional ceramic and nanocomposite fibers, cellular articles and microspheres via radiation curable colloidal dispersions

We present here new cellular ceramics, fibers and microspheres produced by novel and general microshaping techniques employing colloidal dispersions in solvent free radiation curable monomer mixtures. High loading, low viscosity dispersions of functional nanoparticles such as TiO₂ and α-Fe₂O₃ and colloidal Fe(C₂O₄)·2H₂O metal salt are achieved with comb-polyelectrolyte surfactants and/or mildly polar resins. TiO₂ dispersions are spun and solidified “on the fly” by UV radiation into continuous ceramic/polymer nanocomposite fibers with dimensions below 10 μm. Fe(C₂O₄)·2H₂O dispersions provide a UV curable alternative to α-Fe₂O₃ dispersions which can only be thermally cured. The Fe(C₂O₄)·2H₂O nanocomposites transform to α-Fe₂O₃ below 550 °C. Novel cellular Al₂O₃, TiO₂ and α-Fe₂O₃ articles with porosity >80% and precise replication of the pore-forming agents have also been produced from such dispersions. Al₂O₃ nanocomposite microspheres are produced by emulsification of the dispersions in an appropriate medium and UV curing.