Construction of sparse signal representations with adaptive multiscale orthogonal bases

We propose a new approach to construct adaptive multiscale orthonormal (AMO) bases of R^N that provide highly sparse signal representations. Our new multilayer AMO basis design produces a high proportion of small scale vectors. The basis vectors are built from small scale to large scales, layer by layer. For each layer, the basis vector maximizes a p-norm measure of sparsity. We compare the sparsity ratios SR (i.e. the percentage of negligibly small coefficients) obtained with AMO and Daubechies wavelet bases for seven families of piecewise smooth signals with randomly located discontinuities. The signals are composed of polynomial, sinusoidal and exponential pieces. In all cases, AMO bases produce a SR increase ranging from 6% to 37%. AMO bases have three main advantages over wavelets. First, they are found automatically by solving a sequence of optimization problems, which eliminates the problem of selecting a wavelet for a given signal. Second, they can provide a significantly sparser representation. Finally, they have the ability to produce zero coefficients for a larger family of piecewise smooth signals. The drawbacks of AMO bases are computational: the basis computation is more expensive, the basis vectors require storage space and no fast transform is known.     

H optimal approximation for causal spline interpolation

In this paper, we give a causal solution to the problem of spline interpolation using H^∞ optimal approximation. Generally speaking, spline interpolation requires filtering the whole sampled data, the past and the future, to reconstruct the inter-sample values. This leads to non-causality of the filter, and this becomes a critical issue for real-time applications.Our objective here is to derive a causal system which approximates spline interpolation by H^∞ optimization for the filter. The advantage of H^∞ optimization is that it can address uncertainty in the input signals to be interpolated in design, and hence the optimized system has robustness property against signal uncertainty.We give a closed-form solution to the H^∞ optimization in the case of the cubic splines. For higher-order splines, the optimal filter can be effectively solved by a numerical computation. We also show that the optimal FIR (finite impulse response) filter can be designed by an LMI (linear matrix inequality), which can also be effectively solved numerically. A design example is presented to illustrate the result.     

Threshold effect in the energy loss of hydrogen and helium ions transmitted in channeling conditions in gold single crystal

The energy loss of hydrogen and helium ions in the low energy range in single crystal gold targets is investigated experimentally. We found that the stopping power for helium ions shows a deviation from the proportionality with ion velocity predicted theoretically. This behavior has been also observed for protons in previous experiments. Additionally, we found that for a small range of velocity the energy loss for helium ions has the same magnitude as for protons for the same velocity. This last finding is in agreement with early theoretical prediction for the ion energy loss interacting with low electron densities, beyond of metallic densities, in the frame of the free electron gas model and the density functional theory. Both effects are explained due to two particular phenomena: particle channeling in monocrystalline targets, where the inhomogeneity of the spatial electron density distribution plays a fundamental role and the well-known threshold effect in the stopping power which is explained considering the electronic band structure properties of metallic targets.     

On finding minimum cost tree for multi-resource manycast in mesh networks

We investigate the problem of multi-resource manycast in mesh networks. The problem of multi-resource manycast extends the traditional manycast problem or k-Steiner tree problem, which finds a minimum cost tree spanning any k vertices. For the traditional manycast, all the vertices in the set of candidate destinations will be regarded as identical. However, the computing capability of the resource at each vertex may be not equivalent in the realistic networks. In this paper, we consider the problem of multi-resource manycast, in which the computing capability of the resource at a vertex is decomposed into discrete units. That is, each vertex may have multiple units of computing resources. The objective is to find a minimum cost tree spanning any k units of computing resources distributed in the networks. We show that multi-resource manycast is NP-Complete. The ILP formulation and approximation analysis are given for this problem. Simple polynomial-time heuristic algorithms are also proposed for the problem of multi-resource manycast. We investigate various approaches to implement multi-resources manycast in mesh networks, and verify the effectiveness of the approaches through simulation.     

Landé g-factor and cyclotron effective mass in a cylindrical GaAs-(Ga,Al)As quantum pillbox under the influence of an axis-parallel applied magnetic field

We have performed a theoretical study of electronic states and the parallel Landé g-factor in a quantum dot (QD), assumed to be in the form of a pillbox, in the presence of an uniform magnetic field applied parallel to the pillbox axis. The quantum pillbox is assumed to consist of a finite length cylinder of GaAs material surrounded by Ga_1-xAl_xAs which describe the realistic finite potential confinement. The calculations have been performed by using the Kummer confluent hypergeometric functions. This study is performed for different radii and lengths of the cylindrical GaAs pillbox and the limit cases have been studied to prove the validity of the model. Our results are in good agreement with the previous theoretical results [F.E. López, B.A. Rodríguez, E. Reyes-Gómez, L.E. Oliveira, in press] in the limit geometry of a quantum well wire (QWW).     

Quasi-analytical study of the energy levels in double quantum wells

In this work we develop an analytical expression to the eigenvalue equation for the double quantum well (DQW), to describe the energy gap between resonant levels. The calculation of the energy gap does not involve the evaluation of the wavefunctions for the DQW nor the transfer or tunneling integrals in contrast to how it is usually done.     

Crossed Andreev reflection in superconducting junctions

In this work crossed Andreev reflection (CAR) and elastic cotunneling (EC) are studied for junctions (N₁ISIN₂), where N₁ and N₂ are normal metals, S is a high T_c superconductor and I is an insulator. This study is carried out based on the analytical solutions of Bogoliubov de Gennes equations for anisotropic superconductors. The influence of different pair potential symmetries on the CAR and crossed conductance is analyzed. We show that CAR and EC are higher in d_x²-y² symmetry than in s symmetry. In the case of normal electrodes without magnetization, EC is the predominantly process for d_x²-y² symmetry, while in s symmetry, both processes decay with the same amplitude.     

Growth mechanism of iron nanoparticles on (0 0 0 1) sapphire wafers

Laser ablation of a high purity (99.7%) iron target was used to accomplish the depositions of iron nanoparticles on the (0 0 0 1) face of single crystal sapphire wafers. The nanoparticles were characterized in situ by means of X-ray photoelectron spectroscopy (XPS). The growth mechanism was determined by applying the QUASES-Tougaard methodology to the extended part of the background intensity of the Fe KMM peak in XPS spectra. The heights of nanoparticles obtained are between 3.5 and 6.5 nm. In the first 150 laser pulses, the height of the nanoparticles remained constant while the coverage was increased.     

Exponential H∞ filtering for time-varying delay systems: Markovian approach

This paper discusses the H_∞ filtering problem for a class of deterministic systems with time-varying delays, where the stochastic property of time-varying delays described by Markovian approach is taken into consideration in filter design. Firstly, the delay interval is separated into several subintervals, which can be described by Markov process. Then, a new H_∞ filtering method for deterministic system with time-varying delay is given, whose filter can switch with time delay in terms of Markov process. Sufficient conditions for the existence of H_∞ filter are obtained as linear matrix inequalities, where the mode transition rates are known exactly or inexactly. Finally, numerical examples are used to demonstrate the utility of the given methods.     

Milling time effects on the magnetic and structural properties of the Fe70Si30 system

The magnetic and structural properties of the Fe_0,70Si_0,30 alloy are analyzed by using X-ray diffraction and Mössbauer spectrometry. The samples were prepared by mechanical alloying in a planetary ball mill pulverisette 5 during 1, 5, 9, 15, 75 and 100 h milling, with a ball mass to powder mass relation of 10:1. The X-ray diffraction patterns showed a coexistence of BCC, DO₃ and FeSi structural phases; however, for 1-5 h process, we observed additional peaks which correspond to iron and silicon elements. These results indicate us that the alloy is still a non-fully consolidated sample. The grain size decreases drastically from 1 to 5 h, then slowly up to 15 h, and finally it remains nearly constant. The lattice parameter for each phase remains nearly constant (2838 A for the BCC structure). The shape of hyperfine field distributions used to describe the Mössbauer spectra is discussed.     

H∞ filter design for nonlinear networked control systems with uncertain packet-loss probability

This paper is concerned with the problem of designing H_∞ filters for a class of nonlinear networked control systems (NCSs) with transmission delays and packet losses. This problem is investigated under the following assumptions: (i) the delay is constant and known, (ii) the packet-loss process is described by a Bernoulli distributed white sequence, and (iii) the probability of the packet-loss process is unknown but bounded. The first two assumptions are well known in the literature of NCSs, while the last one may often arise in practical networks due to modeling or measurement errors. Under the aforementioned assumptions, delay-dependent conditions for the solvability of the addressed problem are presented in terms of linear matrix inequalities. Numerical examples are also included to illustrate the effectiveness of the proposed method.     

A hierarchical Path Computation Element (PCE)-based k-random-paths routing algorithm in multi-domain WDM networks

We propose an inter-domain routing algorithm for multi-domain wavelength-division multiplexing (WDM) mesh networks. On the basis of the hierarchical Path Computation Element (PCE) architecture, the proposed algorithm introduces a k random path strategy in the parent PCE to adjust the algorithm optimization level. Simulation results indicate that the proposed algorithm outperforms previous schemes in terms of blocking probability and resource utilization, while preventing the increase in PCEP signaling overhead.     

Delay-dependent H∞ filtering for discrete-time singular Markovian jump systems with time-varying delay and partially unknown transition probabilities

This paper is concerned with the H_∞ filtering design for a class of discrete-time singular Markovian jump systems with time-varying delay and partially unknown transition probabilities. The class of systems under consideration is more general and covers the singular Markovian delay systems with completely known and completely unknown transition probabilities as two special cases. A mode-dependent filter is constructed and by defining an appropriate stochastic Lyapunov functional combined with using the discrete Jensen inequality, a delay-dependent bounded real lemma (BRL) for the considered systems is established in terms of linear matrix inequalities (LMIs). Based on this, a sufficient condition on the existence of the desired filter which guarantees the admissibility and the H_∞ performance of the corresponding filtering error system is presented by employing the LMIs technique. Some numerical examples are provided to illustrate the effectiveness of the developed theoretical results.     

Coffee certification criterion using the photoacoustic technique

The photoacoustic (PA) technique allows a spectroscopic analysis of organic opaque samples and its thermal characterization; so this technique gives information about the pigmentation and the roasting modelling parameters, important for industry and commercialization process, when it is applied to green and powdered coffee samples, respectively. A spectrum is particular of each sample; so is possible to identify different characteristics in coffee powder, for example: purity, caffeine content, roasting level, origin conditions, etc. Powdered coffee PA spectroscopic measurements and correlation analysis that allow proposing a discrimination criterion are reported in this work. Also, the thermal diffusivity of green coffee samples was measured by open cell PA technique, in order to study the relation of this parameter with the bean cellular structure, which was looked through an optical microscopy. The results presented will allow a possible method for certifying organic coffee, which is cultivated without chemicals and has greater price in the international market and positive impact on the environment.     

Piezoelectric models for semiconductor quantum dots

The importance of fully coupled and semi-coupled piezoelectric models for quantum dots are compared. Differences in the strain of around 30% and in the electron energies of up to 30 meV were found possible for GaN/AlN dots.     

Input match and load tank digital calibration of an inductively degenerated CMOS LNA

This paper presents a technique for independently tuning the center frequency and quality of the input match for a CMOS low noise amplifier implemented using the inductive source degenerating topology. This technique allows for the input match to be centered to the desired frequency in the presence of process shifts and parasitic elements. The complete calibration loop is presented. The implemented calibration circuits allow for fast calibration, low additional power consumption during calibration, and negligible additional power consumption during operation. In addition, the center frequency of the load tank is tuned to match the desired center frequency of operation using the same calibration loop.     

Comparison of ZnO thin films deposited by three different SILAR processes

ZnO thin films were deposited on amorphous glass substrates using successive ionic layer adsorption and reaction (SILAR). The employed baths were ammonium zincate, with NH₃ and NH₄OH as complex agents and ZnSO₄ and Zn(NO₃)₂ as precursor sources. The comparison between films deposited by three processes was studied by means of X-ray diffraction (XRD), optical absorption, and micro Raman. The band gap is between 3.14 and 3.30 eV. By μ-Raman spectroscopy, the principal vibration modes around 435 and were determined in all the films.