This paper presents a very compact printed antenna with simple and novel radiator for Ultra-Wideband applications. The proposed antenna is fed by a coplanar waveguide feed line to achieve wide bandwidth performance. The impedance bandwidth of the antenna is from 2.88 to 12.82GHz for which S11 < -10dB, which constitutes about 127%. The total size of the fabricated antenna is 18 × 27 × 0.787mm3. The radiation patterns of the antenna are measured and presented. Also the effects of dimensional parameters on the performance of the antenna are investigated through a parametric study. The agreement obtained between the measured and simulated results is excellent.
Cued Speech is a visual mode of communication that uses handshapes and placements in combination with the mouth movements of speech to make the phonemes of a spoken language look different from each other and clearly understandable to deaf individuals. The aim of Cued Speech is to overcome the problems of lip reading and thus enable deaf persons to wholly understand spoken language. In this study, automatic phoneme recognition in Cued Speech for French based on hidden Markov model (HMMs) is introduced. The phoneme correct for a normal-hearing cuer was 82.9%, and for a deaf 81.5%. The results also showed, that creating cuer-independent HMMs should not face any specific difficulties, other than those occured in audio speech recognition.
The reduction of Specific Absorption Rate (SAR) with metamaterial is performed by the finite-difference time-domain method (FDTD) with Lossy-Drude model by CST Microwave Studio in this paper. The metamaterials can be achieved by arranging split ring resonators (SRRS) periodically. The SAR value has been observed by varying the distances between head model to phone model, different width, different thickness, and different height of metamaterial design. Metamaterial has achieved 42.12% reduction of the initial SAR value. These results can endow with supportive information in designing the wireless communications equipments for safety compliance.
Hidden Markov Model (HMM) is a well-known classification approach which its parameters are conventionally learned using maximum likelihood (ML) criterion based on expectation maximization algorithm. Improving of parameter learning beyond ML has been performed based on the concept of discrimination among classes in contrast to maximizing likelihood of each individual class. In this paper, we propose a discriminative feature transformation method based on genetic algorithm, to increase Hidden Markov Model likelihoods in its training phase for a better class discrimination. The method is evaluated for phoneme recognition on clean and noisy TIMIT. Experimental results demonstrate that the proposed transformation method results in higher phone recognition rate than well-known feature transformations methods and conventional HMM learning algorithm based on ML criterion.
Cell assignment is an important issue in the area of resource management in cellular networks. The problem is an NP-hard one and requires efficient search techniques for its solution in real-time. In this letter, in order to minimize the cabling and handoff costs, we use two novel discrete particle swarm optimization (PSO) algorithms, the barebones (BB) and the exploiting barebones (BBExp) PSO variants. The impact of network and algorithm parameters on the solution accuracy and computational cost of the methods is investigated. Comparisons with optimization methods in the literature demonstrate the benefits of our proposal.
A low voltage low power successive approximation register (SAR) analog-to-digital converter (ADC) based on a novel rail-to-rail comparator is proposed in this paper. The power consumption of the comparator is significantly reduced through dynamic operation while the speed is augmented by using an efficient regenerative latch. No biasing circuits are needed and there are no floating nodes in the comparator throughout the conversion process. The digital-to-analog converter (DAC) is formed from a binary array of MIM capacitors. The 250KS/s ADC implemented in a 0.18µm process consumes only 1.35µW of power at a supply voltage of 0.8V.
Two novel architectures for pipelined floating point fast Fourier transform on FPGA are presented. The new radix-22 two-path delay feedback (R22TDF) architecture leads to 50% area saving for floating point complex adders compared with the radix-22 single-path delay feedback (R22SDF) architecture. Besides a new hybrid architecture is presented which mixes the R22TDF and R22TDF butterfly structures and is flexible and efficient for FPGA implementation.
A new low distortion class-D switching power amplifier (PA) is presented using a double-band hysteresis control (DBHC) technique. This PA utilizes two Schmitt trigger comparators to realize 4-level comparison. The THD is lowered due to reduction of overshoot and Hysteresis-band. Simulation results, using a 0.13µm CMOS technology, show that in this PA the THD is improved, while it supports relatively high output power. The performance of DBHC-PA is also experimentally tested and its operation is verified by a PCB prototype amplifier. Measurements illustrate 58% THD improvement compared to conventional class-D PA.
An alternative Space Vector Modulation (SVM) strategy which reduces the Common Mode (CM) perturbations when using Matrix Converters (MC) is presented in this work. The proposed method maintains the conventional SVM pattern but the zero vectors, which have been shown to introduce not only high CM voltage values but also high CM voltage time derivative. Hence the three zero vectors are replaced by three rotating ones, in such a way that the fundamental output voltage vector and the input current direction remains both unchanged. Results of the proposed SVM strategy driving a Permanent Magnet Synchronous Machine (PMSM) are provided to confirm the effectiveness of the method.
Our challenge is to fabricate practical all-inorganic electroluminescent devices using colloidal quantum dots with presenting new aspects of electroluminescence mechanisms. The electric field effect on the excitation of a quantum dot with a hot electron that results in saturated chromatic electroluminescence was studied by time-resolved spectral transient properties at room temperature. Fast and very slow radiative recombination transition pathways are newly distinguished in the electroluminescence related to carrier confinement in the quantum dot under an electric field.
This paper introduces a novel compact wideband CPW-fed slot antenna comprising of an E-shaped patch that is excited with a trident shaped feed-line structure. The antenna design exhibits a very wide operating bandwidth of over 168% with S11 < -10dB in the frequency range 1.6 to 18.9GHz. The antenna's radiation characteristics are stable across this wide bandwidth. The proposed antenna exhibits high polarization purity with cross-polarization level that is well below -40dB. A comprehensive parametric study was carried out to get insight on the effects of various antenna parameters that enable optimization of the antenna's performance. This antenna easily fulfills the requirements for ultra-wideband (UWB) wireless communication systems as specified by FCC over the designated band 3.1GHz-10.6GHz.
This paper reports a general on-wafer noise figure (NF) de-embedding technique with the analysis of two gain definitions. As implemented in this work, all elements involved in NF measurement were determined and classified as a multi-stage network, and the well-known Friis law is applied to correct the noise contributions coming from other stages. With the two gain definitions, the effects of impedance match on NF are investigated. The result shows an NF of 3.80dB obtained with the de-embedding method and 6.06dB without the de-embedding method. This result is for vector measurement using the available gain, which gives 0.18dB improvement in NF as opposed to scalar measurement utilizing the insertion gain. Furthermore, NF was also measured at 5.63dB, 5.76dB, and 4.75dB under three different source impedances, namely, short, open, and load, respectively.
In this paper, a low power 24mW 5Gb/s differential output transimpedance amplifier (TIA) is realized in 0.18µm CMOS technology for optical interconnect application. The TIA is a fully balanced and differential architecture design that can improve immunity with the common mode noise attributed to the power supply. The differential gain achieved is 66dBΩ with a -3dB bandwidth of 4.0GHz for a 0.5pF photodiode capacitance (Cpd) by implementing both the active inductor peaking and capacitive degeneration techniques. The TIA core consumes only 24mW power from a single 1.8V power supply while achieving the sensitivity of -19.0dBm for a bit error rate (BER) of 10-12.
The oscillating motions of a wavefront on a two- dimensional (2D) switch line, i.e., a transmission line periodically loaded with electronic switches (the switch is open for voltages greater than a fixed threshold and closed otherwise), are discussed. In a two-dimensional switch line, the amplitude of the wavefront decays more rapidly than in a one-dimensional switch line, and the oscillation frequency depends on the propagation orientation. The rapid decay of the wavefront, supported by inherent frequency entrainment, is useful for generating high-frequency voltage waves.
We have been developing a Josephson/CMOS hybrid memory, which enables sub-nanosecond access time, in order to overcome a memory bottleneck in single-flux-quantum (SFQ) digital systems. In our previous study, we obtained the access time of about 1.2ns in a 16-kb hybrid memory system using a 0.35µm CMOS process, but observed unexpected double peaks in histograms of access-time measurements. In this study, we designed a 64-kb hybrid memory system using a 0.18µm CMOS process. We considered that the double peak effect is due to parasitic capacitances at the bonding pad of Josephson and CMOS chips and reduced them in the new design. Measured access time is about 1.4ns, which agrees well with simulation results. The double peaks in the histogram were completely removed in the new results.