This paper presents a sensitivity analysis of a health index of power transformers tool, in order to identify the importance of each input to determine most important and those that can be dismissed to simplify the calculation process without significantly affecting the effectiveness of answers. To achieve objective, a methodology based on a study case is proposed which, starting from a set of real scenarios, analyzes the variation of the output (health index) when each input (tests performed in oil) is varied.
In the diagnosis and evaluation of Parkinson's disease (PD) the most frequent motor symptoms are: resting tremors, bradykinesia, gait disturbances, and freezing. In recent years, various types of sensors have been used to objectively quantify these symptoms in PD patients. This work presents a comparative analysis of the main characteristics of sensors developed for the data acquisition and measurement of different types of movement on an ongoing basis in the patient's daily activities. This work also shows the application of sensors in an evaluation of postural tremor in hands on a group of young adults. The analysis developed indicates that the measurement of motor disorders, using some kind of body sensor and portable technologies, set the trend in monitoring of the symptoms PD.
IoT (Internet of Things) devices tend to work intermittently, and the leakage power reduction in the sleeping state is essential. Although giving the reverse body bias to the back gate of transistors is an efficient technique to reduce the leakage power, it has not been commonly used dynamically because of the large timing overhead when a common CMOS process is used. However, recent SOI technologies enabled to use dynamic body bias control with an acceptable overhead. Here, we focus on Silicon On Thin Box (SOTB), a type of SOI technology which can control body biasing widely with a small overhead. For the dynamic body bias control, we analyzed timing overhead of a simple microcontroller and dynamically reconfigurable processor with SOTB technology. Evaluation results revealed that the sleep down time and the wake-up time are about hundreds microseconds both for the processor core, memory module, and smaller elements. We could not observe a certain relationship between the area of target modules and the timing overhead.
In many countries the health system consists of many organizations, both public and private, that must cooperate to achieve the common goal of population health. Even though in countries where public or private sector control most of the health centers, frequently each organization has a particular medical record format. This happens in Venezuela and in some South American countries. This medical record format diversity, makes it difficult to properly attend patients when they move from one health center to another. It also complicates decision-making processes related to public health policies and research, because the medical records can't be shared between organizations. Relevant medical record information is also difficult to combine. To address this problem, we propose a EMR's Ecosystem in which electronic medical records are centralized at the national level and local replicas are kept in every organization. This article is focused on a prototype of synchronization with priorities and on data sharing mechanisms proposed to implement this ecosystem.
This document presents the design and implementation of a wireless sensor network, using the zigbee communication standard, for monitoring the climatic variables in an orchid greenhouse. The objective is to develop a system of low-cost and low-energy of supervision, so that floriculturists have the possibility to monitor the climatological conditions inside the greenhouse and take necessary preventive measures on orchids, optimizing growing time and productivity.
A qualitative assessment of the accuracy for different de-embedding methods is presented in this paper. For the realization of this work, four procedures are analyzed. The open, short, open-short and multiline methods are applied to S-parameters measurements of four microstrip lines and dummy structures. The samples are fabricated on silicon substrate and have the same cross section, but differ in length. The characteristic impedance of the lines is extracted up to 65 GHz by using the ABCD-parameters, then, it is analyzed to determine the most reliably de-embedding method for each transmission line (TL). Also, the dependent frequency S-parameters of the de-embedded lines are obtained and it is shown that below 10 GHz, the extracted data agree well for the different methods and the accuracy is high.
This paper presents a systematic methodology to characterize and model on-chip inductors, built over a ground shield, directly from S-parameter measurements. Excellent model-experiment correlation is achieved up to 50 GHz for the circuit network parameters as well as for the Q-factor.
Traditional methods to identify critical interconnects regarding electromigration are based on current density criteria. As these can be too conservative and lead to a design with oversized interconnects, it can be a problem in designs with area restriction. To guarantee a more accurate design, the chip designer needs to consider besides the critical current density the performance variation. In this work, a method to investigate and identify oversized lines is proposed. The method is applied to evaluate a bandgap circuit. Although the designed interconnects obey the foundries' limits, the results indicate that they can be reduced without affecting neither the circuit performance nor reliability.
In this work we present the design of a digitally controlled ring type oscillator in 0.5 μm CMOS technology for a low-cost and portable radio-frequency diathermy (RFD) device. The oscillator circuit is composed by a low frequency ring oscillator (LFRO), a voltage controlled ring oscillator (VCRO), and a logic control. The digital circuit generates an input signal for the LFO, which generates a voltage ramp that controls the oscillating output signal of the VCRO in the range of 500 KHz to 1 MHz. Simulation results show that the proposed circuit exhibits controllable output characteristics in the range of 500 KHz-1 MHz, with low power consumption and low phase noise, making it suitable for a portable RFD device.
This work presents the simulation of different heavy ion impacts on the control transistors of two different types of magnetic random access memory cells (MRAM): a 32nm Bulk MRAM and 28nm FDSOI MRAM with buried oxide (BOX) thickness of 20nm. The Magnetic Tunnel Junction (MTJ) element of both MRAMs is simulated using a Compact Model described in C++. The control transistors are described using TCAD tools. The Linear Energy Transfer (LET) characteristic of simulated heavy-ions ranges from 10 to 80MeV-cm2/mg. The results show the robustness of these MRAM cells against these type of heavy ions, which is good for applications in harsh conditions.
In this paper, the novel shortest energy consumption Routing Spectrum Assignment (SEC-RSA) algorithm based on Distance Adaptive Modulation (DAM) is introduced. According to the distance between source and destination for a determinate lightpath the modulation pattern is modified. The energy consumption is computed for the k-shortest path, then the one with lowest energy consumption is adopted. The algorithm can dynamically commute to another modulation as the connections requests or the numbers of lightpath between source and destination enhances. The results show that the energy consumption decreases when SEC-RSA algorithm is used in comparison with traditional ones. Moreover, energy consumption decreases as the order of modulation increases. The outcomes additionally figure out that energy consumption decrease when block connections are considered.
This paper proposes methods for improving of magnitude response in a two-stage corrector filter structure. The corrector filters improve the comb passband and stopband characteristic in all odd folding bands. We propose here two methods to improve the stopband characteristic in all folding bands, while keeping a low passband deviation. The first proposed method is based on introducing additional zeros in the center of all even folding bands. The second method separates double zeros within all even folding bands. In both cases the resulting structure is multiplierless. We contrast the proposed methods with some existing methods to demonstrate the benefits of the proposed design technique in aliasing rejection.
In this article proposing a low-cost system for measuring physiological variables in this case temperature, blood pressure and graphic heart rate, which will be a tool for doctors and technicians working in rural areas and presents isolated; the use of consumer electronics such as in tables for viewing, and systems based on FPGA acquirer of data, provide the right to provide services in these regions qualities. After the experiments conducted an error less than 1 % is obtained in the measurements of each of the variables analyzed.
The different benefits of WLANs have achieved their massive implementation in different sectors: domestic business and intelligent cities, among others. However, security in data transmission is one of their biggest problems, as they are more susceptible to suffer different types of wireless network attacks. This disadvantage motivates users to use Virtual Private Networks (VPN) in order to reinforce the security of the data in the network. VPN overload data traffic including another layer, in this article presents the evaluation of VPN QoS parameters over a Wireless Network (WLAN), using different Raspberry Pi models. Additionally, we analyzed the QoS parameters with different traffic conditions and the CPU consumption in the VPN. The contribution of the results and their analysis determine the correlation between the parameters and the Raspberry Pi models. Getting better performance on Raspberry Pi Zero and Raspberry Pi 2.
This article proposes a new hardware implementation for a Reconfigurable Neural Network for systems in which the topology needs flexibility. The used architecture is a MultiLayer Perceptron, where the entry of a layer depends on the output of the previous layer. The approach allows flexibility in the number of network inputs, neurons, layers and in the activation function executed by neurons. Despite having been developed for FPGAs, the implemented circuit allows its implementation in ASIC, since it does not use proprietary internal blocks of the FPGA. By submitting the network to approximation tests, its operation and flexibility has been checked and validated.
This paper presents logarithmic CMOS image sensor with wide output voltage swing range. Wide output voltage swing range is achieved by increasing the number of transistors connected as diodes of logarithmic CMOS image sensor. A prototype has been fabricated in 0.35μm AMS opto process. Four different pixels were fabricated using 20μm × 20μm Nwel/Psub photodiodes with one, two, three and four transistors in series. The sensitivity measured were 136mV/Dec, 191mV/Dec, 281mV/Dec and 425mV/Dec for the pixels using one, two, three and four transistors in series respectively. This results shows that the sensitivity is increased by N approximately, were N is the number of transistors in series at pixels. Considering that the supply voltage the four-transistor pixel presents total output voltage swing range of 2.65V over a range of 120dB being the supply voltage 3.3V.
Closed-form expressions for the capacitance and inductance related to vias (vertical transitions) in interconnects implemented in PCB technology are presented. The importance of this contribution relies on the fact that these parasitic effects critically impact the performance of data channels by introducing undesired signal reflections. The development of the proposal is based on the systematic analysis of measurements and 3D electromagnetic simulations of actual high-speed links. Hence, a set of physically-based equivalent circuit models valid for different via pad sizes (scalable models) is obtained. This allows for the performance assessment and optimization of electrical transitions in channels in a fast and accurate way.
Visual function testing using electroretinogram (ERG) signal is used to detect abnormalities of the retina. This is accomplished by measuring, characterizing and analyzing the bio potential responses from various retinal cells formed due to a visual stimulation. The objective of this study is to improve the existing ERG signal characterization models by identifying and including a key component called i-wave in the photopic response. To validate the proposed characterization model, a signal analysis and processing algorithm based on Multi Resolution Analysis has been developed to reliably separate various underlying components of this signal. Finally, the fidelity of this separation is quantitatively and qualitatively assessed by computing the Pearson's correlation coefficient and the corresponding scatter plots between the composite and the reconstructed ERG signal.
In this paper, the floating body effect (FBE) is experimentally investigated on triple gate n-channel Bulk FinFETs for 1T-FBRAM (1 transistor Floating Body Dynamic Random Access Memory) application. A difference between the Direct Current (DC) and the Alternating Current (AC) measurements corresponding with the real memory operation is shown. The large hysteresis under DC measurement related to floating body effects is observed using AC measurement only when the bulk contact is floating. Moreover, it is shown that the stored holes are not in the fins but probably in the ground plane below the junctions. Furthermore, the floating body effect in the Bulk FinFET is observed when the latter is operated like a biristor, confirming the charge storage in the junction capacitor.
This work addresses the effect of gate misalignment on the electrical characteristics of MOSFET structures with trapezoidal gate shapes. Differential Triangular test structure are used for the extraction of misalignments between gate and the other transistor structures, as a function of the differences in drain currents. Three-dimensional numerical simulation was used for comparison and verification of measurements. These simulations and measurements show how the analytical model developed fits properly to designed structures.
