CIRCUITS

Author: LOPEZ NAVARRO JUAN MANUEL.
Year: 2003.
University: POLITÉCNICA DE MADRID [www.upm.es].
Place of defense: ESCUELA TÉCNICA SUPERIOR DE INGENIEROS DE TELECOMUNICACIÓN.
Place of preparation: INSIA (E.T.S.I. INDUSTRIALES UPM).

Summary: The objective of this dissertation is the presentation of a novel architecture for the measurement of environmental noise. That architecture makes it possible to integrate into one system analysis of environmental parameters with the measure itself noise, achieving perform actions according to the existing rules. Specifically, the system complies with paragraphs of interest in this type of equipment standards UNE60651, UNE61260, UNE60804 and ISO 1996, the proposed architecture has the potential to be controlled and accessed via the latest mobile technologies, GPRS. This permitrá exploitation via the Internet using Web browsers, which has been implemented a stack TCO / IP protocols. It has succeeded in increasing the capacity of the process and benefits of this architecture compared to those existing at present. This has been achieved with the use of new digital signal processors, which have capacities of calculating the order of Gigaflops. It has conducted a detailed study of the execution times, the DSP of the proposed algorithms. Establishing, thus, execution speed required minimum in a DSP to implement in real time for these algorithms. The system has the advantage of not doing any spectral estimation in action, thanks to the implementation of digital type that has been done, by reference to the model of an analog sound meter and an analyzer into thirds of the eighth. Likewise, this type of implementation is in accordance with those specified in the rules, which are generally taken as a reference analog equipment. Another of the key points of this architecture and who has been given much attention, because it is one of the most critical elements in building a sound meter, is the microphone preamplifier. This has led to a thorough investigation of different topologies implementation. As a result, there has obtained a preamplifier with very low noise and high dynamic range (6.9 uV across the audio bandwidth and 100 dB). It can be concluded, in view of the results obtained in this investigation, which exposed the architecture here is feasible for its implementation and improves performance of existing equipment at present. It may represent an important contribution in the field of acoustic instrumentation.

Author: SERRA SERRA MOISÉS.
Year: 2004.
University: AUTÓNOMA DE BARCELONA [www.uab.es].
Place of defense: ETSE.
Place of preparation: ESCUELA DE POSTGRADO.

Summary: This thesis presents the application of a new methodology for rapid prototyping based platforms for communication subsystems, namely the digital physical layer protocol wireless HiperLAN / 2. The main contributions of this research are: * Development and validation of a design methodology that transforms directly specifications to the level of system platforms physical maintaining intensive verification heterogenia multilevel. * Exploration and implementation of new hardware architectures to improve complex algorithms in the chain HiperLAN / 2, gaining models sintetizables for physical layer's digital transmitter and receiver of HiperLAN / 2. * Prototyping part of the digital physical layer issuer's HiperLAN / 2 with a transmission rate Dice 12MBits / s, in an appropriate platform for WLAN. The purpose of this prototype is to demonstrate the high productivity of the methodology just implementing the physical design platforms. The work is divided into four parts: fundamental concepts, rapid prototyping methodology, design of the physical layer and prototyping of the issuer. In key concepts, provides an introduction to the model explaining OFDM signal properties and advantages of this modulation. Following is the standard wireless HiperLAN / 2 that uses OFDM. In particular, present working environments, the topology, the physical layer, the complexity of the model and type of bursts of standard HiperLAN / 2. In the methodology of rapid prototype of a WLAN system, presents the methodology used in the prototype transmitter and receiver of the digital part of the physical layer of HiperLAN / 2. The following is a quick introduction to the prototype, outlines the requirements of the platforms for prototyping of a WLAN system, shows the state of the art environments and platforms for WLAN architectures and explains the FPGAs of Virtes. Following is the methodology WLAN sync with the methods and design flows using. Specifically, there are two different design flow, a flow classic and the other more adapted to specific WLAN systems, which offer several alternatives very interesting verification heterogenia multilevel. Finally, discusses the concepts of functional abstraction, d'ortogonalitat and heterogeneity of methodology. The design of the physical layer of sender and receiver are shown in detail its structures and computational components. Significant contributions have been proposed for modules IFFT / FFT, equalizer canla, Synchronizer, interleaver / deinterleaver and Viterbi. The results of synthesis showed improvement compared with existing design alternatives. The models obtained from the issuer and the receiver show its functionality and performance at system level. The prototyping of emsior, described changes to the model of the issuer for prototyping in a specific platform. In the end, we present the results obtained through the spectrum analyzer, along with an estimate of the power consumption of the issuer to validate the proposed method.

Author: SALAS MERINO VICENTE.
Year: 2004.
University: CARLOS III DE MADRID [www.uc3m.es].
Place of defense: ESCUELA POLITECNICA SUPERIOR.
Place of preparation: UNIVERSIDAD CARLOS III DE MADRID.

Summary: The objective of this thesis is to propose a new control strategy, using a single variable extent, the flow of photovoltaic generator, as well as an algorithm, in order to find, at any moment, the maximum photovoltaic generator. The above strategy is included in the methods called "tracking real" peak. Both the strategy as the method have been applied to PV systems autonomous 24 V. For verification theoretical simulations have been performed static, in principle, for the topology reductive. Subsequently, it has also been found that it is also useful for other topologies: lift and reductora-elevadora, and their derivatives. Subsequently provides experimental results obtained in both open loop and closed-loop implementation for a converter DC / DC topology with reductive with digital control. This control is based on the use of a microcontroller Risc, the PIC 16F873. The results obtained confirm the above assumptions. Subsequently, the implementation has been compared with other configurations: configuration parallel 12V, parallel 24 V configuration settings series 12 V.

Author: MÉNDEZ PATIÑO ARTURO.
Year: 2006.
University: POLITÉCNICA DE VALENCIA [www.upv.es].
Place of defense: UNIVERSIDAD POLITÉCNICA DE VALENCIA.
Place of preparation: UNIVERSIDAD POLITÉCNICA DE VALENCIA.

Summary: The vast amount of information that currently is handled on a digital, the need for storage and transmission of compression makes an indispensable process. One of the steps in the compression process is the transformation, or changes shape fits the data manipulated by a mathematical equation, concentrating energy or essence of information, then reduce or compress the amount of data. The transformed exploits the features used in time and / or frequency of data compression to achieve an optimal concentration of information. Different mathematical transformations with characteristics suitable for various types of data. For data from images and video, some of the transformations with suitable properties are: the Cosine Transform; transform Wavelets and transform Karhumen-Loève (KLT). Because of the complexity of the mathematical KLT and its dependence with the data set is not generally used, despite being offered the best results for the compression. The transformations are applied on scanner data using discrete version: transformed Discreta of DCT and Cosine Transform Discreta of Wavelets DWT. The DCT emerged in the seventies and has been adopted by a number of standard image compression and video. For three decades have developed algorithms for simplifying the calculation and architectures have been proposed for implementation. By definition both the DCT as DWT require the use of fractional numbers for its implementation, which always involves a compromise between the complexity or the amount of hardware used and the precision or the number of errors introduced into the calculation. Changes have also been used in which only uses whole numbers. These variants some authors have given the name of Whole Transforms ((ICT; R) -