Yolanda Campos-Roca

Addressing voice recording replications for Parkinson's disease detection

A general subject-based Bayesian approach has been proposed.Special treatment is provided for the probit model.Latent variables are used to provide a predictive model that can handle replications.A Gibbs sampling-based method is derived to compute the model parameters.The approach is used to discriminate healthy people from people suffering PD. A clinical expert system has been developed for detection of Parkinsons Disease (PD). The system extracts features from voice recordings and considers an advanced statistical approach for pattern recognition. The significance of the work lies on the development and use of a novel subject-based Bayesian approach to account for the dependent nature of the data in a replicated measure-based design. The ideas under this approach are conceptually simple and easy-to-implement by using Gibbs sampling. Available information could be included in the model through the prior distribution. In order to assess the performance of the proposed system, a voice recording replication-based experiment has been specifically conducted to discriminate healthy people from people suffering PD. The experiment involved 80 subjects, half of them affected by PD. The proposed system is able to discriminate acceptably well healthy people from people with PD in spite that the experiment has a reduced number of subjects.

A D-band frequency doubler MMIC based on a 100-nm metamorphic HEMT technology

A coplanar single-ended frequency doubler based on a 100 nm metamorphic HEMT technology is presented. For an input power of 4.8 dBm, this doubler demonstrates an output power between 2.6 and -0.3 dBm over the bandwidth from 105 to 145 GHz, that is, a 3-dB bandwidth of 32% has been achieved. To the knowledge of the authors, this is the first reported multiplier based on MHEMT technology at D-band or higher frequencies.

A 200 GHz Medium Power Amplifier MMIC in Cascode Metamorphic HEMT Technology

A 200 GHz power amplifier is presented. The millimeter-wave monolithic integrated circuit (MMIC) has been realized in a 35 nm InAlAs/InGaAs cascode metamorphic high electron mobility transistor (MHEMT) process in grounded coplanar waveguide technology (GCPW). The amplifier demonstrates an output power of 14 mW with 11.4 dB compressed power gain at 200 GHz. This represents an increase in output power in comparison to previous reported MHEMT-based MMIC amplifiers. The small-signal gain demonstrates a peak value of 20 dB and is above 15.9 dB from 185 to 215 GHz.

A 200 GHz driver amplifier in metamorphic HEMT technology

A driver amplifier operating at 200 GHz has been design and manufactured with compact size, high gain, broadband performance and high power added efficiency (PAE). This monolithic microwave integrated circuit (MMIC) is realized in grounded coplanar waveguides (GCPW) technology in conjunction with a 35 nm gate length metamorphic high electron mobility transistor technology (mHEMT). The four-stage driver amplifier provides more than 20 dB linear gain between 180 GHz and 270 GHz (40 % bandwidth) and PAE higher than 3.3% with more than 7.4 dBm saturated output power at 200 GHz.

Multibias scalable HEMT small-signal modeling based on a hybrid direct extraction/particle swarm optimization approach

A new multibias HEMT small-signal model extraction method is proposed. The approach, based on scaling rules, combines direct extraction techniques and a particle swarm optimization algorithm. This method has been successfully tested with PHEMTs and MHEMTs, leading to accurate and scalable models up to 70 and 120GHz, respectively.

A broadband 175–245 GHz balanced medium power amplifier using 50-nm mHEMT technology

In this paper, a balanced millimeter-wave monolithic integrated circuit (MMIC) medium power amplifier (MPA) is presented. This amplifier was developed by using an advanced 50 nm metamorphic high electron mobility transistor (mHEMT) process, where a grounded coplanar waveguide (GCPW) technology and a compact cascode configuration were adopted to achieve high gain and high output power at the WR-4 waveguide band (170–260 GHz). This two-stage MPA exhibits a 13.4 dB peak measured small-signal gain at 228 GHz and an excellent 3-dB bandwidth of 33.8% from 175 to 245 GHz. Good input and output return losses with higher values than 9.8 dB and 9.3 dB were measured. At 198 GHz a maximum output power of 9.5 dBm with 7.5 dB associated large signal gain is obtained, while at 235 GHz an output power-referred 2.3-dB compression point of 6.6 dBm is exhibited. A linear output power higher than 4 dBm (at 1-dB compression point) is realized from 198 GHz to 240 GHz (19% relative bandwidth, RBW).

Impact of Metallization Layer Structure on the Performance of G-Band Branch-Line Couplers

Five different versions of grounded coplanar waveguide (GCPW) branch-line couplers on GaAs were EM simulated, processed and investigated for operation in G-band (140-220 GHz), based on different layer structures (including 2 or 3 metallization layers) and layouts. The best results have been obtained with the structures based on a continuous galvanic metal in the central conductor line and ohmic connections between top ground planes, reducing the insertion losses by 0.8 dB. A measured amplitude imbalance less than 0.5 dB from 160 to 200 GHz (22%), with a coupler insertion loss lower than 1.3 dB was achieved for a 3 metallization layer branch-line coupler.

Scalable HEMT small-signal model extraction based on a hybrid multibias approach

A multibias approach for HEMT small-signal model extraction is proposed. The method is based on scaling rules and uses a hybrid direct extraction/particle swarm optimization approach. By the use of a reduced number of S-parameter measurements in the optimization procedure, the computational effort is kept low. The extraction procedure is verified with measurements on metamorphic HEMTs, leading to accurate and scalable models in the millimeter-wave range. Furthermore, the robustness of the proposed method against measurement uncertainties is verified by using PHEMT synthetic data, intentionally affected by errors.