Multiple-band large-signal characterization of millimeter-wave double avalanche region transit time diode

A large-signal method based on non-sinusoidal voltage excitation model is used to study the DC and RF characteristics of Double Avalanche Region (DAR) Silicon Transit Time diode. A large-signal simulation program based on drift-diffusion model is developed for this study. The simulation results show the existence of several distinct negative conductance bands in the admittance characteristics separated by positive conductance. Thus the DAR device is capable of delivering RF power not only at the design frequency but also at several frequency bands higher than the design frequency band in the mm-wave regime. A comparative study with DDR Si device designed to deliver RF power at a particular mm-wave frequency band shows that DAR Si device is capable of delivering significantly higher RF power not only at the designed mm-wave frequency band, but also at higher frequency bands.     

Material characterization of SS 316 in low-cycle fatigue loading

This study deals with simulation of low-cycle fatigue (LCF), followed by evaluation of fatigue parameters, which would be suitable for estimating fatigue lives under uniaxial loading. The cyclic elastic–plastic stress–strain responses were analyzed using the incremental plasticity procedures. Finite-element (FE) simulation in elastic–plastic regime was carried out in FE package ABAQUS. Emphasis has been laid on calibration of SS 316 stainless steel for LCF behavior. For experimental verifications, a series of low-cycle fatigue tests were conducted using smooth, cylindrical specimens under strain-controlled, fully reversed condition in INSTRON UTM (Universal Testing Machine) with 8,800 controller at room temperature. The comparisons between numerical simulations and experimental observations reveal the matching to be satisfactory in engineering sense. Based on the cyclic elastic–plastic stress–strain response, both from experiments and simulation, loop areas, computed for various strain amplitude, have been identified as fatigue damage parameter. Fatigue strain life curves are generated for fatigue life prediction using Coffin–Manson relation, Smith–Watson–Topper model, and plastic energy dissipated per cycle (loop area). Life prediction for LCF has been found out to be almost identical for all these three criteria and correlations between predicted and experimental results are shown. It is concluded that the improvement of fatigue life prediction depends not only on the fatigue damage models, but also on the accurate evaluations of the cyclic elastic–plastic stress/strain responses.     

A journey towards reliability improvement of TiO2 based Resistive Random Access Memory: A review

A Resistive Random Access Memory (RRAM), where the memory performance principally originated from ‘resistive’ change rather than ‘capacitive’ one (the case with conventional CMOS memory devices), has attracted researchers across the globe, owing to its unique features and advantages meeting the demands of future generation high-speed, ultra low power, nano dimensional memory devices. A large family of semiconducting oxides have been investigated as insulator for Resistive Random Access Memory (RRAM), amongst which TiO₂ is one of the potential candidate, principally owing to some of its remarkable advantages e.g. wide band gap, high temperature stability and high dielectric constant with flexibility to offer both unipolar and bipolar switching, which are essential for RRAM device applications. In this review article, we tried to represent the long voyage of TiO₂ based RRAM, towards the improvement of the reliability aspects of the device performance in a comprehensive manner. Starting with the key factors like oxygen vacancies, Ti interstitials and electroforming, which are responsible for resistive switching phenomenon, various material preparation techniques for RRAM development have been discussed with emphasis on relative merits and bottlenecks of the process. The factors like electrode material and geometry, device structuring, doping, compliance current, annealing effect etc., which play the pivotal role in determining the switching performance of the device, have been reviewed critically. Finally, the article concludes with the comparison of different TiO₂ based RRAM devices followed by the prediction of possible future research trends.     

Optical control of large-signal properties of millimeter-wave and sub-millimeter-wave DDR Si IMPATTs

The authors have proposed a complete large-signal (L-S) model to investigate the optical modulation of high frequency properties of double-drift region (DDR) impact avalanche transit time (IMPATT) devices operating at different millimeter-wave and sub-millimeter-wave frequencies. Simulation is carried out based on the proposed model to study the effect of photo-irradiation of different values of incident optical power of different wavelengths from 600–1000 nm on the DC and L-S characteristics of DDR IMPATTs based on Si designed to operate at 94, 140, 220, 300 and 500 GHz. Two different optical illumination configurations such as top mount and flip chip are taken into account for the present study. Results show that the maximum optical tuning of L-S parameters of the device can be achieved for optical illumination of wavelength 700 nm, i.e. near the wavelength corresponding to the responsivity peak of Si in both top mount and flip chip configurations. Further, better photo-sensitivity of top mount structure is observed as compared to its flip chip counterpart. The simulation results are found to be in good agreement with the experimental results reported earlier. Suitable tunable light source information is also suggested to carry out the optical control experiment within the wavelength range under consideration.     

A generalized analytical model based on multistage scattering phenomena for estimating the impact ionization rate of charge carriers in semiconductors

A generalized analytical model based on multistage scattering phenomena has been developed in this paper for estimating the impact ionization rate of charge carriers in semiconductors. The probabilities of impact ionization initiated by electrons and holes have been calculated separately by taking into account all possible combinations of optical phonon scattering and carrier-carrier collisions prior to the impact ionization. Finally the analytical expressions of impact ionization rate of electrons and holes have been developed by using the aforementioned impact ionization probabilities. The impact ionization rates of electrons and holes in 4H-SiC have been calculated within the field range of \(2.5\times 10^{8}\) – \(6.5\times 10^{8}\hbox { V m}^{-1}\) by using the analytical expressions of those developed in the present paper. Those are also calculated by using the analytical expressions developed by some other researchers earlier without considering the multistage scattering phenomena. Finally the theoretical results obtained from the analytical model proposed in this paper and the analytical model developed by earlier researchers within the field range under consideration have been compared with the ionization rate values calculated by using the empirical relations fitted from the experimentally measured data. Closer agreement with the experimental data has been achieved when the impact ionization rate of charge carriers in 4H-SiC are calculated from the proposed model as compared to the earlier one.     

Language independent unsupervised learning of short message service dialect

Noise in textual data such as those introduced by multilinguality, misspellings, abbreviations, deletions, phonetic spellings, non-standard transliteration, etc. pose considerable problems for text-mining. Such corruptions are very common in instant messenger and short message service data and they adversely affect off-the-shelf text mining methods. Most techniques address this problem by supervised methods by making use of hand labeled corrections. But they require human generated labels and corrections that are very expensive and time consuming to obtain because of multilinguality and complexity of the corruptions. While we do not champion unsupervised methods over supervised when quality of results is the singular concern, we demonstrate that unsupervised methods can provide cost effective results without the need for expensive human intervention that is necessary to generate a parallel labeled corpora. A generative model based unsupervised technique is presented that maps non-standard words to their corresponding conventional frequent form. A hidden Markov model (HMM) over a “subsequencized” representation of words is used, where a word is represented as a bag of weighted subsequences. The approximate maximum likelihood inference algorithm used is such that the training phase involves clustering over vectors and not the customary and expensive dynamic programming (Baum–Welch algorithm) over sequences that is necessary for HMMs. A principled transformation of maximum likelihood based “central clustering” cost function of Baum–Welch into a “pairwise similarity” based clustering is proposed. This transformation makes it possible to apply “subsequence kernel” based methods that model delete and insert corruptions well. The novelty of this approach lies in that the expensive (Baum–Welch) iterations required for HMM, can be avoided through an approximation of the loglikelihood function and by establishing a connection between the loglikelihood and a pairwise distance. Anecdotal evidence of efficacy is provided on public and proprietary data.     

BIST/Digital-Compatible Testing of RF Devices Using Distortion Model Fitting

Testing of Radio Frequency (RF) circuits for nonlinearity specifications generally requires the use of multiple test measurements thereby contributing to increased test cost. Prior RF test methods have suffered from significant test calibration effort (training for supervised learners) when using compact tests or from increased test time due to direct specification measurement. On the other hand, due to aggressive technology scaling, there are plenty of digital transistors available that can be used to simplify testing of Analog/Mixed-Signal (AMS) and RF devices. In this paper, an RF test methodology is developed that: (a) allows RF devices to be tested for several distortion specifications using distortion model fitting algorithms in test time comparable to what can be achieved using supervised learning techniques while retaining the accuracy of direct specification measurement, (b) allows multiple RF specifications to be determined concurrently from a single data acquisition and (c) allows digital-compatible testing/BIST to be performed using digital testers or on-chip built in self-test (BIST) circuitry. With regard to (a), a key benefit is that no training of supervised learning algorithm is necessary. The proposed method based on distortion model fitting is shown to give excellent results across common RF performance metrics while providing ~10× improvements in test time compared to previous methods.