Search

Nowadays, the digital chip design flow starts with formal specifications, which are mapped to Register Transfer Level (RTL) models using different underlying implementation variants and (micro-) architectures. By doing so, a hardware designer predicts and resolves time-critical parts to achieve an RTL-design that intentionally meets all constraints after synthesis. However, wrong predictions can be detected only later in the design flow, thus leading to long design iterations. Classical methods estimating delay in early design stages are constrained to the type of components or are computationally expensive for larger designs. In this paper, we propose a Machine Learning-based approach to estimate pin-to-pin delays for RTL combinational circuits. To gain accuracy, we combine slew and delay estimation. To that end, a training set is built using features of components generated by a model-driven hardware generator framework. Ground truth labels for delays, slews, and their interdependencies are extracted using open-source tools for logic synthesis and static timing analysis. Evaluations in unseen designs show that the delay estimation has on average an accuracy of 87% and it is 13x faster compared with results of synthesis and timing analysis tools. Based on the estimation, critical areas of the design can be detected and proper microarchitecture decisions can be taken earlier in the design flow.

Despite the high configurability of IPs and hardware generators, code modifications are still required to introduce aspect-oriented instrumentation to satisfy emerging design requirements such as on-chip debug and functional safety. These code modifications lead to escalated development, verification efforts and deteriorate the code reuse. This paper proposes a highly efficient aspect-oriented design automation approach that leverages graph-grammar-based model transformations. With the proposed approach, main design functionalities and aspect-oriented instrumentation are separately developed, automatically integrated and verified. To demonstrate the applicability, industrial SoCs were transformed to support on-chip debug. Experimental results confirm the efficiency of the approach. Further, reduced code is needed with the proposed automation approach, which also replaces the error-prone manual RTL coding. Finally, the transformation scripts are applicable to different SoCs, which promotes the overall code reuse.

Driven by Moore’s law, the chip design complexity is steadily increasing. Electronic Design Automation (EDA) has been able to cope with the challenging very large-scale integration process, assuring scalability, reliability, and proper time-to-market. However, EDA approaches are time and resource-demanding, and they often do not guarantee optimal solutions. To alleviate these, Machine Learning (ML) has been incorporated into many stages of the design flow, such as in placement and routing. Many solutions employ Euclidean data and ML techniques without considering that many EDA objects are represented naturally as graphs. The trending Graph Neural Networks are an opportunity to solve EDA problems directly using graph structures for circuits, intermediate RTLs, and netlists. In this paper, we present a comprehensive review of the existing works linking the EDA flow for chip design and Graph Neural Networks.

The Singular Spectrum Analysis technique has been applied to the amplitude of the first Schumann resonance measured at the Sierra Nevada station (Spain) in order to obtain the principal components of its time evolution. The results of this study confirm the appearance of the annual and semiannual components that have been pointed out for this resonance and also reveal other components corresponding to 45 to 120 days variations, which matches the variations of athmospheric waves like the Madden-Julian oscillation and the Kelvin waves. A preprocessing of the measurements is required since the station has experienced some technical failures and there are thus some gaps in the measured data. The application of the technique has been made taking one data per month and one data per day, and the results have been compared.

In this paper, a hybrid Doherty power amplifier (DPA) using dynamic input power splitting for 3.5 GHz is presented and compared to another Doherty with a conventional Wilkinson input divider sharing the same output structure. The amplifier is designed in a novel hybrid configuration, where a MMIC consisting of two parallel GaN transistors with input prematching are fully integrated on chip and used as the main and peak amplifiers. The remaining passive structures are implemented at PCB level on a RO4350B. The results at 3.5 GHz under continuous wave operation show 45 dBm of maximum output power with over 41% of PAE and Drain efficiency over 47% in a 6dB output power backoff (OBO) and 7 dB of gain for maximum output power.

This article presents the perception of the causes of motorcyclist accidents in the city of Medellin. Traffic accidents, for many researchers, are considered a public health problem, which is why the Ministry of Transportation has implemented strategies to mitigate this problem. this article presents the perception of the different drivers on the road about the cause traffic accidents. For this, surveys were carried out on the Likert scale and the responses were compared with the behaviors of the different users in one of the routes of maximum accident in the city. It was found that half of the motorcyclists surveyed did not take any driving course to obtain their license, therefore the lack of expertise in conjunction with the age of the motorcyclists become risk factors for both them and other drivers’ actors. Almost unanimously, drivers of private vehicles and drivers of public service vehicles like buses have the perception that motorcyclists are reckless and generate traffic accidents, which is correct with the data found and with the different studies carried out.

In this paper we are presenting a power amplifier based on an RF power LDMOS transistor at 2.1 GHz with a bandwidth of 700 MHz implementing broadband impedance matching networks (BIMN) in band-pass form. The device parasitics modelled as a series RLC structure are absorbed in the input and output matching networks by proper design of matching networks in band-pass form using two different synthesis methodologies. The S 21 of the amplifier shows 11.2 dB of maximum small signal gain with 2.4 dB ripple along a 700 MHz bandwidth at 2.1 GHz and S 11 better than −7dB for 1. 64-2.34GHz. The band-pass matching networks are entirely designed by transmission lines modeling the behavior of the resonant parallel shunt and series LC structures that characterize a band-pass circuit. The amplifier should be able to deliver 1W of maximum output power at saturation.

Face-to-face learning activities involve multiple acoustic interactions between students and teachers. These interactions provide information about their level of engagement with the educational activity. Hence, the acoustic interaction between the different actors in a classroom produces relevant information to evaluate their performance in a quantitative and qualitative manner. Traditionally, this information is collected through surveys and interviews. Technology is rarely used to record learning sessions due to the cost related to the installation of the equipment and the complexity of the analysis of the recordings. Here, we developed a web application that guided different types of student-teacher interaction during a class. The application recorded all the acoustic interactions and labeled the information so that subsequent analysis of the information can be efficiently processed.

User engagement and information understanding was assessed for a basic electronics tutorial designed using Augmented Reality (AR). The main idea was to compare user performance in collaborative environments between two different versions of the AR application: a desktop version and a mobile version. Preliminary testing of both versions revealed that users were interested in AR technology but did not intuitively understand how to better utilize the system in comparison with a static video. Moreover, in spite there was not a statistically significant difference between information understanding between both conditions, the mobile application was rated significantly lower than the desktop application in user engagement for collaborative environments.

Assistive technology in learning environments needs to foster information recall strategies as a primary advantage over traditional learning systems. Moreover, multi-modal priming can be used as a well-known strategy to reinforce integration and association of memory for different stimulus in different modalities. A game-based memory training was designed to aid the visual vocabulary acquisition of Japanese Kanji using audio and tactile stimuli priming. Preliminary results revealed that the memory training showed a statistically significant increase in information recall. However, further research is needed to validate if this type of priming strategy holds in long-term information recall.

This work proposes and illustrates a graphical approach aimed at studying a wide range of features of the ELF horizontal magnetic field signal recorded at the Sierra Nevada station (Spain). In addition to the traditional long-term variations in the parameters of the first three Schumann resonances (their amplitudes, central frequencies and widths), many other properties such as the saturations of the magnetometers, anomalous values for the parameters or spectra with any kind of particularities are taken into consideration in this work. These features can provide us with complementary information about the long-term variation of Schumann resonances, give an estimation of the extent up to which the results obtained are reliable and be correlated with the occurrence of lightning events or with changes in the electrical properties of the ionosphere. The scheme proposed in this work allows to instantaneously display the variations of all these features within a desired period of time.

La denominada mision HIPAE (Helicopter-borne In-situ Pollution Assessment Experiment) desarrollo una prueba de concepto dentro de una aeronave de la Fuerza Aerea Colombiana, sobrevolando el Valle de Aburra para transportar dos tipos de contadores de particulas PM2.5 y PM10, asi como dos versiones de las plataformas en desarrollo llamadas Simple para medir variables meteorologicas (humedad relativa, presion barometrica, temperatura), altitud, geo-posicion y ocho tipos de gases CO 2 , H 2 , NO 2 , NH 3 , C 2 H 6 OH, CH 4 , C 4 H 10 , C 3 H 8 . Adicionalmente, un experimento con nano filtros demostro su capacidad para capturar material particulado, el cual fue analizado mediante microscopia electronica de barrido combinada con espectroscopia de rayos-X (EDX). Los resultados de EDX arrojaron informacion valiosa sobre la morfologia y quimica a nivel de particula en la atmosfera urbana por encima de la altura de las estaciones de medicion tradicionales. Fue posible visualizar en los datos altas concentraciones de compuestos de aerosol y gases como CO, NO 2 y CH 4 , cuyos valores fueron menores en areas rurales y forestales en comparacion con areas urbanas segun lo esperado. La plataforma Simple mostro un comportamiento adecuado manteniendose dentro de sus niveles de incertidumbre, indicando la utilidad de los datos adquiridos como primer paso a siguiente ejercicio para ser utilizadas en aeronaves comerciales o militares con el objetivo de suministrar constantemente, a los modelos meteorologicos y quimicos de transporte, informacion in-situ para actividades de asimilacion de datos basadas en ensamble, tanto secuencial (EnKF) como variacionalmente (4DenVar), como en actividades de fusion de datos para la toma de decisiones.

Graphene and 2D materials are being used in the design of light sources due to their wide absorption spectrum, ultrafast recovery time and high modulation depth. In this work, we present a experimental design of an modelocking Erbium-Doped-Fiber-Laser (EDFL) that uses graphene deposited onto D-shaped optical fiber, acting as a saturable absorber, in order to generate ultra-short pulses in an important region for optical communications: 1558.5 nm. To best of our knowledge, this is the first ultrashort pulses laser designed an experimentally validated in Colombia. The results shows a pulsed laser with a repetition rate of 33.3 MHz and a spectral width of 10 nm.

In this paper we present the performance characterization of Erbium-doped Fiber Pulsed (EDFL) laser with monolayer graphene as satuble absorber, when the Pulse Repetition Rate (PRR) and Spectral Width (SPW) are changed by introducing cavity dispersion variations by cavity length increases. For cavity lengths between 4.86 m and 24.86 m, the obtained PRR was between 8 MHz and 41 MHz, and the SPW changing between 18.40 nm and 2.08 nm. The SPW and the PRR performance showed a inversely proportional behaviour respect to cavity length L. For the case of the PRR, the fitting curve obtained was 200 L−1 with a Pearson correlation coefficient r=1.

In this paper, different human/machine strategies are tested in order to evaluate their performance in underwater threat recognition. Sonar images collected using synthetic aperture sonar (SAS) and side scan sonar (SSS) during real mine countermeasures exercises are used. Data are collected over a test area on the Belgian Continental Shelf, where several targets were deployed. Image resolution is divided in three categories: (1) up to 5cm pixel size, (2) between 5cm and 10cm pixel size, (3) larger than 10cm pixel size. Soil complexity is also evaluated and used to build up different strategies. Results demonstrate the utility of considering the human operator as an integral part of the automatic underwater object recognition process, as well as how automated algorithms can extend and complement human performances.

Transformers are essential devices in power and distribution systems, since they reduce energy losses produced in the transmission stage. The magnetic system is the most-important part to its operation because energy conversion process will always be defined and bounded by the electromagnetic characteristics of its ferromagnetic core. This paper proposes an experimental method based on electromagnetic principles to obtain an approximated magnetization curve of an electric transformer, which can be used as a reference to analyze magnetic-system-behavior evolution through time. Moreover, the obtained magnetization curve considers voltage and current harmonic components produced during the test. Therefore, a detailed division of input transformer power is done, assuming the total active power are magnetic losses and the total non-active power corresponds to magnetization power. Obtained results showed the proposed method relevance, after comparing it with a magnetization curve which rejects harmonic content of signals and a magnetization curve released by a ferromagnetic material manufacturer.

One of the tasks of the Music Information Retrieval area is the characterization of a musical genre. For this purpose, it is necessary to have a dataset of songs with enough information to perform the task. In order to know if there is enough information on the dataset, an analysis of the dataset is required. In this paper we perform an analysis of the Salsa dataset using unsupervised machine learning techniques and then extend it with annotations made by an expert.

Several organizations interested in auditing public decisions and the use of resources have increased the pressure on governments to publish the data produced by its operations to improve transparency, governance and the interaction of citizens with institutions. The answer to these demands has been an increasing number of governments and institutions planning, designing and implementing data management and publication projects without enough knowledge of methodologies and experience to carry out the planning of this type of project. Therefore, many methodologies and architectures for the implementation of open government platforms have emerged, but there is still a need to have guidelines that combines management, political and technical aspects. Consequently, this paper proposes a project planning guide to carry out the preparation and publication of government data considering the main technical elements. The preparation of the guide was carried out using as base the evaluation of methodologies, standards and academic references on open government data management. The main contribution of the work is the guide and the description of the process that can be followed for publishing open data, taking into account both the technical and management aspects involved.