Reearch

Digital Systems Design Area Research

Lines of Research

Nuclear Medicine Electronics Systems

Coordinator: Vicente Herrero Bosch

Nuclear medicine equipment uses very specific radiation sensors that require very high performance electronic conditioning and data acquisition systems. The high output density of these sensors, as well as the processing speed required to capture the signals they provide, make it necessary to adopt high-tech solutions that are not commercially available:

The analog electronics required for the low-level connection of the sensor as well as for initial processing to reduce the complexity of the system requires integration in an ASIC (Application Specific Integrated Circuit).
The high rate of information provided by the sensors requires a high-speed acquisition system where the signals are converted to the digital domain and specific processing is performed to obtain the necessary characteristics from them. The use of FPGAs (programmable logic devices) allows the modification of algorithms through firmware updates.
The methods of reconstructing three-dimensional images from the information provided by the acquisition system have a high computational cost, so GPU (Graphics Processing Unit) based systems are used to accelerate these tasks.

High-Speed Data Acquisition and Processing Systems

Coordinator: Raúl Esteve Bosch

The electronics needed for some applications require the design of ad-hoc digital systems that could not be implemented with existing commercial systems, due to their characteristics, both in terms of processing speed, power consumption, and number of inputs and outputs.

Within the Electronic Systems Design Area, an important capability is the design and implementation of reconfigurable systems and architectures for data acquisition and processing in high-speed applications, using high-speed buses, such as PCI Express, and advanced programmable logic devices, such as FPGAs, through the use of high-level languages.

Modelado y Simulación Paralela de Sistemas Bioinspirados

Coordinator: Joaquín Cerdá Boluda

This is an eminently multidisciplinary line in which there is room for new and revolutionary computational models, most of which are inspired by biological systems, such as Genetic Algorithms, Neural Networks or Cellular Automata. Emerging and promising computational alternatives such as Quantum Computing also fall into this line. Of special importance is the implementation and simulation of these alternative computing models on parallel architectures, especially FPGAs and graphics processors (GPUs), as well as the development of compilers and development environments to facilitate work with the new paradigms.

RF MEMS Device Design and Simulation of Microfabrication Processes

Coordinator: Jorge Daniel Martínez Pérez

This line is focused on two fundamental areas:

The development of new microfabrication techniques and the emergence of RF MEMS in the late 1990s have generated new expectations and promising applications for wireless communications devices with very low insertion loss, negligible power consumption, low cost, micrometer dimensions and virtually weightless. RF MEMS are devices with lateral sizes between 1-1000 microns, combining electronic and mechanical elements, usually moving parts that are driven by electrostatic, thermal, piezoelectric or magnetic actuation mechanisms). These components can be used as passive elements (i.e. switches and varactors) with performances far superior to those offered by semiconductor devices or electromechanical relays. They can also be used for the design of RF and microwave components with reconfiguration capabilities (i.e. tunable filters, antennas, impedance matching networks, phase shifters, etc.). This line of research is focused on the implementation of RF MEMS devices from the concept stage to fabrication.
Several of the microsystem fabrication processes are complex and their final results are difficult to predict. That is why part of our efforts are devoted to the modeling, characterization and efficient simulation through new hardware structures such as graphics processors (GPUs) of fabrication processes, especially anisotropic and reactive ion etching. Currently we have collaboratively developed CAD software for anisotropic etching processes, distributed by IntelliSense Corporation.

High Performance Interconnection Networks

Coordinator: Salvador Coll Arnau

In this line, high-speed interconnection networks are studied and used, based on standards such as Infiniband, Myrinet, QsNet, Hypertransport and VLSI architectures for switching between nodes in multiprocessor systems, multicomputers or computer clusters.

In addition, aspects such as the development of routing algorithms, new collective communication mechanisms and techniques to reduce power consumption in interconnection networks are studied.