Volume 18, Number 4, 2015



On Additional Constrains in Lossless Compression of Text Files
pp. 299–311


Abstract. Coding algorithms are generally aimed at minimizing the output code length encoding speed once the code has been designed. Moreover, most codes use a binary alphabet. This paper examines other issues related to coding, such as additional constraints imposed on the channel. Code generation will be considered where there is a limit on code words. Limits the application of such systems is a practical example of data compression where fast decoding is essential. When all code words correspond to a single word in memory (usually 32 bits, but there are situations that take 64-bit) can be used canonical decoding. If the deadline cannot be guaranteed, however, required the use of slower decoding methods. This paper also deals with the alphabetic code generation, where lexicographic arrangement of words by their code symbols must correspond to the original order in which the symbols were taken coding system. When an alphabetic code is used to compress a database that can be sorted in the same, order they would have had if the database records were first decompressed and then sorted. It also corresponds to alphabetic code trees binary search trees, which have applications in a wide variety of search problems. Assumption that the symbols are sorted by probability is not suitable for this scenario. The problem of finding codes for non-binary alphabets channel will be examined in detail. The subsequent experimental results cover the problem of alphabetic coding and of limited length coding. Read the pdf







Ligia FRUNZA, Stefan FRUNZA, Constantin Paul GANEA, Irina ZGURA, Florentina NEATU, Vasile I. PARVULESCU
Dielectric Properties of LDH-Type Layered Materials Containing Zn or
Mg Ions: On the Non Monotonous Temperature Dependence of Relaxation Times

pp. 312–328


Abstract. The electrical behavior of layered double hydroxides containing Al or Ga as the trivalent ions and Mg or Zn as bivalent ions was evaluated using broadband dielectric spectroscopy in a wide temperature range. Besides conduction effects a relaxation peak is observable at high frequencies which is assigned to the reorientational fluctuations of water molecules adsorbed on the oxide surface or in the interlayer voids. The Maxwell-Wagner-Sillars peak, superimposed to the conductivity phenomenon is observable at low frequencies. A non-monotonous temperature dependence of the relaxation rates of the relaxation process has been found. A quantitative description of this dependence was possible based on a model assuming two competing processes: rotational fluctuation of water molecules and formation of additional defects. Reasonable values for the characteristic parameters were obtained. Thus the water behavior in the studied layered oxides is similar to that observed for water in other porous materials. However, the activation energy of the rotational fluctuation, the pre exponential factor and the number of defects are higher whereas the value of the energy of defect formation is lower in the layered oxide materials than for water confined to nanoporous molecular sieves, porous glasses or in bulk ice. Read the pdf







P-Type Doping of 4H-SiC for Integrated Bipolar and Unipolar Devices
pp. 329–342

Abstract. P-type 4H-SiC layers formed by ion implantation need high temperature processes, which generate surface roughness, losing and incomplete activation of dopants. Due to dopant redistribution and channeling effect, it is difficult to predict the depth of the formed junctions. Vapor-Liquid-Solid (VLS) selective epitaxy is an alternative method to obtain locally highly doped p-type layers in the 1020 cm-3 range or more. The depth of this p-type layers or regions is accurately controlled by the initial Reactive-Ion-Etching (RIE) of the SiC. Lateral Junction Field Effect Transistor (JFET) devices are fabricated by integrating p-type layers created by Al ion implantation or VLS growth. The p-type VLS layers improve the access resistances on the electrodes of the fabricated devices. Read the pdf








Marius VOLMER, Marioara AVRAM, Andrei Marius AVRAM
Detection of Magnetic Nanoparticles for Lab-on-a Chip Applications
pp. 343–355

Abstract. Detection of magnetic carriers used as labels for biosensing applications represent key features in lab-on-a-chip (LOC) devices. This paper presents a review of some experimental techniques used by us to characterize magnetic nanoparticles and to detect them using Permalloy based planar Hall effect sensors. We found that sensitivities better than 0.2x10-3 emu/mV can be obtained for specific detection schemes. A freeware micromagnetic simulator, SimulMag, was used to complete the view on these experiments. Read the pdf








Roxana-Daniela AMARIUTEI, Liviu GORAS, Monica RAFAILA, Andi BUZO, Georg PELZ
On the Suppression of Output Oscillations in a Software Controlled DC-DC Buck Converter
pp. 356–375


Abstract. In this paper unwanted output voltage oscillations due to the nonlinearities in a software controlled DC-DC Buck converter are analyzed and conditions for their suppression are given These conditions are correlated with the PID parameters designed using an algorithm that ensures imposed transient behavior for input steps and load variations. Read the pdf






Maria-Alexandra PAUN, Claudio FALCO, Florin UDREA
On the Necessary Verification of Fabrication Stages for SOI CMOS
Humidity Sensors

pp. 376–387


Abstract. A humidity sensor, which is part of an air quality sensor system, based on MWCNTs/MMA composite doped with KOH has been studied. A standard SOI CMOS process, with only one post-processing step for the membrane etching, was used to fabricate the sensor presented. In order to have a homogeneous distribution of the CNTs in the solution, the tip sonication method was employed. The current voltage and resistance voltage characteristics have been investigated, for different
temperatures. Within the sensing layer of the proposed humidity sensor, the parabolic dependence of the resistance with the voltage applied has been proven for positive voltage values. The corresponding first and second order coefficients have been extracted. Read the pdf






Anca Gabriela PURDILA, Gheorghe PRISTAVU, Vlad ANGHEL, Gheorghe BREZEANU, Emilian CEUCA
Low-Voltage Controlled Floating Buck Converter with Imbalanced Stacked MOSFETs Switch (ISMS) for High LED Driving Capability
pp. 388–407

Abstract. This paper presents a cost effective technique for extending LED driving capability of floating buck converters, using an imbalanced stacked MOSFETs switch (ISMS). The ISMS is comprised of a low-voltage transistor and a power MOSFET, biased at a constant voltage, and can be driven by low output capability control blocks. Unlike prior-art, the proposed ISMS does not rely on synchronous switching and identical devices. The overall turn ON/OFF times impact on the converter's switching frequency is investigated. The proposed technique is validated by implementation in floating buck converters with two high performing control circuits. Hence, a critical conduction mode control block and a variable OFF time controller, fabricated in 40 V and 5 V CMOS technologies, respectively, are used for driving the converter. Measurements are performed for a wide range of LED strings (6-48) and up to 250 V input voltage. Efficiency results reveal a peak value of 97% and switching frequencies between 25 KHz and 170 KHz. Converters with ISMS are experimentally proven to drive up to eight times more LEDs than their regular versions, with higher efficiencies. Read the pdf.