Search

The characterization of the wireless propagation channel has always been an important issue in radio communications. However, in recent years, given the dramatic increase of demand in terms of capabilities of wireless systems, e.g. data rate, quality of service etc., the study of propagation has become of crucial importance. As measurements are generally costly and time consuming, channel models are widely used for this purpose. The modeling of propagation may rely on different types of models: empirical, statistical and deterministic. Typically, the latter, based on the electromagnetic properties of the propagating waves, relies on Ray-Tracing (RT), a geometrical optic technique used to evaluate the paths followed by rays as they interact with the environment. This technique has been widely used to predict the behavior of propagation both in outdoor and indoor scenarios. Classic RT tools take into account line of sight propagation, specular reflection and diffraction, with penetration included when indoor propagation is considered. With the advent of diversity techniques and MIMO systems, whose goal is to exploit the multi-dimensional properties of a channel, it is required that the prediction models are capable to analyze propagation in depth, i.e. including polarization, delay and angular behavior of a channel. A 3-D RT tool has the great advantage to be able to provide inherently that kind of information, but to achieve a sufficient accuracy, it is necessary to develop an advanced prediction tool. To this goal, a pre-existent RT tool has been improved by implementing in it propagation mechanisms as penetration, diffuse scattering and contributions by vegetation., (FSA 3) -- UCL, 2012

In this thesis the multi-polarized Cognitive Radios are studied. Cognitive Radios are proposed as an interesting way to more efficiently use the frequency resources. A Cognitive Radio secondary user finds the frequency bands which are not utilized by primary users and communicates on them without interfering with the primary users. In order to achieve this goal the secondary user must be able to detect reliably and quickly the presence of a primary user in a frequency band. In this thesis, the impact of polarization on the spectrum sensing performances of cognitive radio systems is studied. First the depolarization occurring in the wireless channel is studied for two cognitive radio scenarios. This is done through an extensive measurement campaign in two outdoor-to-indoor and indoor-to-indoor scenarios where the parameters characterizing the radiowaves polarization are characterized at three different spatial scales: small-scale variation, large-scale variation and distance variation. Second, a new approach is proposed in modeling of multi-polarized channels. The polarization of received fields is characterized from an electromagnetic point of view by modeling the polarization ellipse. Theoretical formulations are proposed in order to obtain the parameters characterizing the polarization ellipse based on the signals received on three cross-polarized antennas. A system-based statistical model of the time-dynamics of polarization is proposed based on an indoor-to-indoor measurement campaign. The analytical formulations needed in order to project the polarization ellipse onto a polarized multi-antenna system are given and it is shown how the model can be generated. Third, the impact of polarization on the spectrum sensing performances of energy detection method is presented and its importance is highlighted. The performance of spectrum sensing with multi-polarized antennas is compared with unipolar single and multi-antenna systems. This analysis is based on an analytic, (FSA 3) -- UCL, 2012

As MEMS devices become ubiquitous in modern technologies, a proper knowledge of the mechanical and material properties of the structural layers of these devices becomes critical. The need for accurate micro-scale characterization of material properties has triggered a flood of characterization techniques. Motivated by the practical and scientific importance of thin films, the work presented in this dissertation, aims to study the dynamics of MEMS tensile stages and provide a method for estimating stress that is based on the out-of-plane flexural resonance of the test structures. More precisely, a one dimensional (1D) dynamic analysis is performed based on two different techniques: (i) the Euler-Bernoulli beam equation and (ii) the Rayleigh-Ritz approach. Two types of structures were fabricated and tested for this work, the first having a 330 nm thick Palladium film as the test material, and the other having sc Si as the test material. This thin film testing approach was extended in the case of sc Si to study the impact of the sample’s thickness on the apparent extensional Young’s modulus. Indeed a dependence of the Young’s modulus of the sc Si on the sample thickness was observed, the variation in the extensional elasticity reached almost 50% for ultrathin sc Si beams., (FSA 3) -- UCL, 2012

In this thesis the multi-polarized Cognitive Radios are studied. Cognitive Radios are proposed as an interesting way to more efficiently use the frequency resources. A Cognitive Radio secondary user finds the frequency bands which are not utilized by primary users and communicates on them without interfering with the primary users. In order to achieve this goal the secondary user must be able to detect reliably and quickly the presence of a primary user in a frequency band. In this thesis, the impact of polarization on the spectrum sensing performances of cognitive radio systems is studied.First the depolarization occurring in the wireless channel is studied for two cognitive radio scenarios. This is done through an extensive measurement campaign in two outdoor-to-indoor and indoor-to-indoor scenarios where the parameters characterizing the radiowaves polarization are characterized at three different spatial scales: small-scale variation, large-scale variation and distance variation. Second, a new approach is proposed in modeling of multi-polarized channels. The polarization of received fields is characterized from an electromagnetic point of view by modeling the polarization ellipse. Theoretical formulations are proposed in order to obtain the parameters characterizing the polarization ellipse based on the signals received on three cross-polarized antennas. A system-based statistical model of the time-dynamics of polarization is proposed based on an indoor-to-indoor measurement campaign. The analytical formulations needed in order to project the polarization ellipse onto a polarized multi-antenna system are given and it is shown how the model can be generated. Third, the impact of polarization on the spectrum sensing performances of energy detection method is presented and its importance is highlighted. The performance of spectrum sensing with multi-polarized antennas is compared with unipolar single and multi-antenna systems. This analysis is bas, Doctorat en Sciences de l'ingénieur, info:eu-repo/semantics/nonPublished

Within the Microelectronics Industry, the core research is focused on the realization of the Moore's law, which states that circuit density doubles every 24 months, shaping the framework of the More of Moore paradigm. However, Moore's law is expected to end, as devices are reaching limitations inherent to the approach of the atomical dimensions. Alternative research paths emerged, forming the new More than Moore paradigm. This consists in using the microfabrication technological knowhow towards the realization of alternative devices and applications, among others the miniaturization and integration of Radio Frequency (RF) devices by Micro Electromechanical Systems (MEMS), i.e. RF-MEMS. The RF-MEMS devices offer high performance, tuning by movable parts and open new perspectives at extra high frequency i.e. 30 to 300 GHz. In this thesis, cavity resonators' design and characterization are introduced, as a preliminary discussion. Their integration in the bulk of High Resistivity Silicon (HR-Si) wafers by micromachining techniques is realized at 60 and 75 GHz. Further, the tuning performance induced by internal volumes of perturbation is thoroughly investigated. Furthermore, the integration of a voltage controlled tuning system for air filled cavity resonators is realized at 60 GHz, using a MEMS based Faraday cage. Additionally, a new miniaturization concept is demonstrated using High Impedance Surfaces (HIS). A seven-pole Chebyshev bandpass filter is realized in Low Temperature Co-fired Ceramic (LTCC). Finally, ultra-low phase-noise oscillators at 60 GHz are realized using cavity resonators integrated in HR-Si and LTCC. These oscillators improve the state of the art for integrated oscillators in the frequency band from 40 to 80 GHz, demonstrating the highest factor of merit, to our best knowledge and to date, FoM = -199 dBcHz @ 1MHz offset from the carrier frequency, fosc = 59.98 GHz., (FSA 3) -- UCL, 2011

The trade-off between insertion loss and isolation in RF antenna switches is critical for RF communication systems. Possible improvement through circuit design is limited. The implementation of RF antenna switch using an optimized technology or an innovative device architecture could be the solution to achieve better insertion loss and isolation. The dynamic threshold MOSFET is presented here as one possible solution.

Although scaling down the dimensions of transistors is a main requirement for digital applications, it could be problematic for RF applications. This paper shows the importance of relaxing the scaling rules for the gate oxide thickness in order to optimize the benefits of scaling.

A Zero Temperature Coefficient characteristic is presented and characterized for the first time for current gain cutoff frequency and maximum oscillation frequency of different MOSFET structures. The benefits of these points on the design of RF circuits are of first importance especially for harsh environment applications and high density RF circuits.

The effect of elevated temperature on the harmonic distortion in Graded-Channel MOSFETs is presented in this work. The Graded-Channel devices show interesting advantages in terms of nonlinear behavior compared to classical devices especially at higher temperatures up to 200°C.

Compact antennas for wireless communication devices have gained very high interest. Antenna miniaturization and their performance enhancement remains a challenge, due to the fundamental limitations that bound the realizable performance of an antenna of given size. Radiation efficiency and impedance bandwidth, which are primary considerations in such designs, are critically affected by the antenna size and near-field environment. The main objective of this thesis is to contribute insights and design methodologies into antenna miniaturization and performance enhancement for both wire and slot-type antennas. First, formulae for the radiation resistance of short planar wire antennas are derived. This work shows that an electrically-short planar monopole exhibits a peculiar radiation resistance, which increases linearly with the monopole electrical length. This leads to an appreciably high radiation resistance, even for a short monopole height, and to a convenient aspect ratio of the antenna. A miniaturized bent planar monopole is proposed to illustrate how one can exploit the high radiation resistance of such short monopoles, while reducing the antenna overall size. The miniaturization approach takes advantage of the linear element topology of a planar wire radiator to yield an intrinsic capacitive loading. Its effectiveness in miniaturization is demonstrated. Second, a compact and low-profile antenna design with a very stable impedance bandwidth versus antenna supporting ground plane is proposed. The antenna essentially corresponds to a slot antenna backed with a very shallow box containing the slotline feed. The slot miniaturization concept is dual to the approach proposed for miniaturization of planar monopoles. Finally, a new design procedure is proposed for dual-linear and circular polarization., (FSA 3) -- UCL, 2010

In this brief, it is clearly demonstrated that a two-parameter noise model is sufficient to accurately extract the MOSFET high-frequency noise performance, as long as channel uniformity is ensured (which corresponds to mainstream CMOS technology). Nevertheless, in the case of asymmetric channel-based MOSFETs, it is shown that a three-parameter noise model is required.

In this paper, measurements from various narrowband and wideband Land Mobile Satellite (LMS) and Satellite Aeronautical experimental campaigns that have been homogeneously pre-processed and re-analysed in order to derive concise statistical results for submission to ITU-R statistical databanks. A number of problems in the processing steps were identified and various methods were applied to minimise the effect on the results, for instance, the calibration of the Line of Sight (LOS) level, the accuracy of position and attitude of transmitter and receiver or problems with bandwidth and frequency synchronisation/drifts for wideband measurements., Anglais

Wireless, portable, and performance are becoming the keywords for both the consumer electronics and the telecommunication markets. These key requirements are very demanding on the circuit design and on the technology levels. In fact, the circuit design is no more separated from the technology. The passage from micro- to nano-electronics was accompanied by significant revolutionary enhancements and modifications of device modeling. In order to meet the new challenges of advanced applications, it is indispensable for the device model to approach from the device physics, as opposed to the empirical models known earlier. This leads, de facto, to a more complicated device model. As a result, a good circuit designer should coercively comprehend the device physics. A profound research is indeed compulsory in this gray area between the circuit design and the device characterization conventional areas. In this context, the work of this PhD thesis is devoted to provide the link between the device engineering/characterization domain and the circuit design domain, pronouncedly for RF applications. This is achieved through an elaborated research of already existing transistor structures (floating-body and body-tied) as well as new structures (graded channel). Different technologies such as bulk, partially- and fully-depleted SOI are also considered. Based on a complete flow (design of layout to on-wafer measurements), a detailed comparison of all transistors (in terms of dc, ac, RF, non-linear, RF noise, low voltage low power, and high temperature performances) is presented in order to render the RF circuit designer with adequate information for a successful RF product., Sans fils, portable, et performance sont devenus les mots-clés dans les marchés des électroniques et télécommunications. Ces mots clés sont très exigeants aux niveaux des conceptions de circuits et de la technologie. En effet, la conception de circuit n’est plus loin de la technologie. Le passage de micro- au nano-électronique a été accompagné par des améliorations révolutionnaires dans la modélisation de composant. Afin de répondre aux défis des applications avancées, il est indispensable que le model s’approche de la physique du composant, au contraire du cas précédent où des modèles empiriques étaient utilisés. Le concepteur de circuit est donc obligé de bien comprendre la physique du composant. Une recherche approfondie est fortement demandé dans cette zone peu servie entre la conception de circuit et la caractérisation conventionnelle des composants. Dans ce contexte, cette thèse est dévoué à établir le lien entre la caractérisation des composants d’un côté et la conception de circuit de l’autre côté, en particulier pour les applications haute fréquence. Cela est atteint en appliquant une recherche élaborée de différentes structures des transistors. Des technologies différentes sont aussi considérées comme les technologies bulk, SOI partiellement- et complètement- déserté. Sur base d’un cycle complet de recherche (en commençant par le layout jusqu’aux mesures sous pointes), une comparaison détaillé des tous les transistors (en terme des performances de dc, ac, RF, non-linaires, bruit RF, faible tension faible puissance et haute température) est présentée afin de fournir le concepteur de circuit RF avec un ensemble d’information suffisant pour une conception réussite., (FSA 3) -- UCL, 2010

This paper presents a new approach to optimize the RF performance at high temperatures for low power low voltage applications. It is shown that the correct choice of the bias point can result in an improvement of the RF behavior of SOI transistors with increasing the temperature, which is opposite to the traditional degradation of RF behavior with increasing temperature. This approach is confirmed by RF measurements for both floating-body and body-tied SOI MOSFET transistors.

This paper describes the results achieved by an experimental campaign whose goal was to characterize the directional propagation channel at 1.9 GHz in forested areas for moderate antenna heights. These experimental results are useful in the context of pedestrian radio-localization systems relying on cellphone signals. Wideband experimental results, which are recorded with transmit-to-receive distances ranging from 40 to 110 m, indicate that the path loss is well modeled by ITU-R Recommendation 833-4. The temporal fading and multipath dispersion are also analyzed. The delay spread, which ranges from 60 to 120 ns, is strongly anticorrelated with the temporal coherence and increases with transmit-to-receive distance. The azimuth spread remains limited, with average values of about 15 degrees.

In this paper, measured RF noise performance of graded-channel metal-oxide-semiconductor (MOS) transistors (GCMOS-also named laterally asymmetric channel transistors) shows impressive reduction in minimum noise figure (NFmin) as compared to classical MOSFET transistors (with the same gate length L-g = 0.5 mu m). The reason is proven to be because of the higher noise correlation coefficient (C). GCMOS also shows lower sensitivity to extrinsic thermal noise as compared to classical MOSFET. Moreover, it is demonstrated that the use of 0.5-mu m-gate-length GCMOS gives a competitive RF noise performance as compared to 0.25-mu m-gate-length classical nMOS transistors.

The superior performance of Graded-Channel MOSFET over classical MOSFET transistors is demonstrated to extend to down scaled channel lengths. While keeping a normal down scaling trend, Graded-Channel devices continue to show favoured static and analog performances in comparison to classical devices. Graded-Channel devices are also characterized in high temperature and high frequency regimes of operation., Anglais

This paper is presenting a model for the prediction of cumulative statistics of amplitude scintillation at microwave frequencies, on an earth-space communication link. This model is using one year of pre-processed radiosonde data for which correction algorithms are applied in order to improve the vertical humidity profiles. The statistics of the vertical profile of the structure constant C/sup 2//sub n/ of the troposphere is then calculated and used for the evaluation of the mean value of the variance of the scintillation amplitude ( sigma /sup 2//sub chi / ). Assuming a Gaussian distribution, the cumulative distribution of sigma /sup 2//sub chi / is then calculated. The statistics have been evaluated for 1990, using radiosonde launches from Uccle, Belgium, and compared to statistics of scintillation variance measured the same year at Louvain-la-Neuve, using the Olympus beacons., Anglais

The introduction of deep n-well protection for bulk MOS transistors can highly enhance their DC and RF performance. It also gives the advantage of having floating-body and body-tied structures in bulk MOSFETs while eliminating the disadvantages related to the shift in performance between these two structures. We demonstrate that the high temperature, DC and RF performance of n-well isolated bulk MOSFETs are really competitive compared to the state-of-the-art Partially Depleted SOI MOS technology., Anglais

This paper presents a complete study of the impact of mechanical stress on the performance of SOI MOSFETs. This investigation includes dc, analog and RF characteristics. Parameters of a small-signal equivalent circuit are also ex- tracted as a function of applied mechanical stress. Piezore- sistance coefficient is shown to be a key element in describing the enhancement in the characteristics of the device due to mechanical stress.

The stationarity of Land Mobile Satellite (LMS) channels is investigated. It is shown that the evaluation of the time-varying channel stationarity is essential before data processing for parameter extraction. This is illustrated by several examples using real-world datasets.

Silicon-on-Insulator (SOI) CMOS technology constitutes a good candidate for mixed signal RF CMOS applications. Due to its low junction capacitance and reduced leakage current, SOI provides reduced static and dynamic power consumption of the digital logic combined with increased cut-off frequencies. Moreover, in terms of passive device integration the major benefit of SOI when compared to the conventional bulk is the possibility to use a high resistivity substrate which allows a drastic reduction of substrate losses allowing a high quality factor of the passive devices. Another issue is the harsh environment applications. Electronics capable of operating at high temperatures are required in several industrial applications, including the automobile industry, the aerospace industry, the electrical and nuclear power industries, and the well-logging industry. The capability of SOI circuits to expand the operating temperature range of integrated circuits up to 300°C has been demonstrated. SOI devices and circuits present advantages in this field over bulk counterparts such as the absence of thermally-activated latch up and reduced leakage current. In this context, various topologies of integrated transmission lines and spiral inductors implemented on standard and high substrate resistivities have been analyzed over a large temperature range. The temperature behavior of the SOI transistors is presented. The main figures-of-merit of the SOI MOSFETs are analyzed and the extraction of the extrinsic and intrinsic parameters of the small signal equivalent circuit is performed. Also, an example of RF circuit applications of the SOI technology, based on a fully integrated Low-Noise Amplifier for low-power and narrow-band applications, is investigated and characterized at high temperature. The main figures-of-merit of the designed circuit are extracted and discussed. The good results show that the SOI technology is now emerging as a good candidate for the realization of analog int, (FSA 3) -- UCL, 2008

This paper reviews and analyzes a compact model for integrated planar spiral inductors on standard and high resistivity substrates in silicon-on-insulator (SOI) technology. The inductors have been characterized over a temperature range from 25 to 200 degrees C. The temperature variation of each model parameter has been investigated. It demonstrates that only the variations of the metallic losses versus temperature have to be taken into account to model properly the high frequency behavior over a wide temperature range of a spiral inductor integrated on silicon high resistivity substrate. Based on these experimental and characterization results, guidelines for practical inductor designs in RFICs for high-temperature applications are drawn. (C) 2008 Elsevier Ltd. All rights reserved.

This paper presents designs and measurements of distributed amplifiers (DAs) processed on a 130-nm silicon-on-insulator CMOS technology on either standard-resistivity (10 Omega . cm) or high-resistivity (>1 k Omega . cm) substrates, and with either body-contacted (BC) or floating-body (FB) MOSFETs. Investigations have been carried out to assess the impact of active device performance and transmission line losses on circuit design by means of simulations, analytical calculations, and comparisons of the small-signal equivalent-circuit parameters. On standard-resistivity substrates, DAs with FB devices and lossy microstrip lines on thin film exhibit a measured gain of 7.1 dB and a unity-gain bandwidth (UGB) of 27 GHz for a dc power consumption of 57 mW. With the introduction of high-resistivity substrates, other DAs, with the same architecture and using lower loss coplanar waveguide lines, show a UGB of 51 GHz with FB devices and 47 GHz with BC devices. To the authors' knowledge, the designs presented in this paper achieve the best tradeoffis in terms of gain, bandwidth, and power consumption for CMOS-based circuits With comparable architecture.

At zero-temperature-coefficient bias points, transistors are known to have stable DC performance with temperature variation. In this work, the RF behavior at those specific bias points is presented in order to provide design guidelines for low-power low-voltage circuits featuring stable RF performance in variable temperature environments and applications. Fully- and partially depleted SOI MOSFETs with and without body contact are analyzed., Anglais

This paper describes the results achieved by various experimental campaigns aimed at characterizing the directional propagation channel in DCS/UMTS band (1.9 GHz) in indoor and forest areas for similar antenna heights at both ends. These experimental results are useful in the context of pedestrian radio-localization systems relying on cell phone signals. Wideband data, are used to model the channel through various parameters: pathloss, shadowing, fading statistics, delay and azimuth-spreads. For indoor scenarios, path-loss exponents range from 1.4 to 2.2, while the shadowing, temporal K-factor, delay- and azimuth-spreads are found to be lognormally distributed. The RMS delay-spread ranges from 9 ns to 95 ns, whereas the azimuth-spread ranges from 20 to 140 degrees. In forests, the delay-spread, ranging from 60 to 120 ns, is strongly anti-correlated with the temporal coherence, and increases with transmit-to-receive distance. The azimuth-spread remains limited, with average values around 15 degrees., Anglais

A low-noise amplifier (LNA) is presented in this paper. The LNA is designed and optimized for narrowband and wideband operations along with minimum noise figure around 24 GHz (K-band). The designed LNA employs microstrip input and output matching networks; it achieves a noise figure of 3.85 dB and 10.55 dB of gain along with 10.27 dB and 17.83 dB of input and output return losses, respectively for its narrow band. Where as for its wideband it accomplishes a gain of 6.5 dB with +/-0.3 dB flatness from 24 to 25.25 GHz wideband frequency range, the noise figure obtained for the presented LNA is in close proximity to the minimum noise figure over 23 - 24.5 GHz frequency range., Anglais

This work investigates the possibility to tune the zero-temperature-coefficient (ZTC) points in partially depleted (PD) SOI nMOSFET technology by controlling the body-source forward bias (V/sub BS/). Measured transconductance and drain current in the saturation region at temperatures between 25 and 200 degrees C were observed for various body-source forward bias conditions. It is found that the variation of threshold voltage (V/sub TH/) with body bias has an influence on ZTC points. The measurement results show wide voltage-range of gate-voltage giving either the transconductance ZTC point (V/sub GS/,/sub ZTC9m/) or the drain-current ZTC point (V/sub GS/,/sub ZTC1DS/) opening important opportunities in RF circuits design for nigh temperature applications., Anglais

At zero-temperature-coefficient bias points, transistors are known to have stable DC performance with temperature variation. In this work, the RF behavior at those specific bias points is presented in order to provide design guidelines for low-power low-voltage circuits featuring stable RF performance in variable temperature environments and applications. Fully- and partially depleted SOI MOSFETs with and without body contact are analyzed.

This work presents the high-temperature DC and RFbehaviorsofpartially-depletedSOI MOSFETs. DC and RF figures of merit are deeply investigated, both analytically and experimentally, to provide a complete study ofhigh-temperature performance and a comparison between floating-body and body-tied transistors. A highly stable RF performance is noticed especially for cutoff frequency and intrinsic elements for temperature as high as 250° C.

RF figures of merit are deeply investigated, both analytically and experimentally, to provide a complete study of high-temperature performance and a comparison between floating-body and body-tied transistors. A highly stable RF performance is noticed especially for cutoff frequency and intrinsic elements for temperature as high as 250 C.

High-quality silicon crystals provide the basis of many industrial technological advances, including computers and telecommunication devices. The increasing size and extremely high quality requirements of silicon wafers have made furnace design and crystal manufacturing a very challenging task. Numerical simulations have become an essential and powerful tool to overcome the difficulties of the experimental approach with a view to understanding the crystal growth process but also to finding an appropriate path to optimize the crystal pulling conditions in industry. This thesis deals with the use of alternating and steady transverse magnetic fields in silicon growth from the melt. The use of magnetic fields represents a powerful tool to damp out turbulence and control the melt flow. This technique can also be used to heat the system. We focus on the numerical modeling of (i) induction heating in the Floating Zone process and of (ii) melt convection under the effect of transverse magnetic fields in the Czochralski process. For each of these topics, our work is subdivided in two parts : firstly mathematical modeling, based on asymptotic or Fourier expansions, and secondly numerical implementation and simulation of the considered processes. First, a theoretical and numerical model of the alternating magnetic field distribution (as generated by induction heating) has been developed by means of an asymptotic expansion technique. Moreover, a new methodology has been developed to calculate the thermal and mechanical effects of alternating magnetic fields on the liquid conductor flow, leading to accurate expressions for the equivalent magnetic heat flux and surface stresses in the 2D and 3D cases. Second, investigation of the effect of a transverse magnetic field on the melt flow in semi-conductor crystal growth has been performed by the simplified FLET method ("Fourier Limited Expansion Technique".), (FSA 3)--UCL, 2007

Space-time code (STC) designs classically rely on the assumption of independent and identically distributed (i.i.d.) Rayleigh channels. However, poor scattering conditions may have detrimental effects on the performance of STCs. In this letter, we derive code-design criteria leading to robust STCs in a large variety of slow-fading propagation conditions. No channel knowledge is assumed at the transmitter. Codes satisfying these criteria are shown to perform much better on real-world channels than codes designed only for i.i.d. channels. As examples, the robustness of various spatial multiplexing schemes, linear dispersion codes, and space-time trellis codes is discussed based on those criteria.

This paper presents improved methods for the prediction of cumulative distribution of scintillation variance, using radiosonde data. This method has the advantage of being usable all around the world and its accuracy is improved with respect to classical methods based only on ground measured data. The second part presents a two-layer depolarisation model separating rain and ice effects for better scalability., Anglais

In this paper, a physical-statistical approach is used in combination with time series simulations, which rely directly upon distributions of physical parameters and simple electromagnetic theory to predict fade dynamics and shadowing correlation statistics for navigation systems. In the former case, a UTD-based ray-tracing is used with a series of roadside buildings, randomly generated according to known statistical distributions for building height and spacing, street width, etc. When dealing with shadowing correlation statistics for a given satellite constellation, the distribution of the shadowing correlation coefficient is then obtained by deriving time series of first and second elevation angles as well as of azimuth separation, using an orbit generator as input., Anglais

This paper proposes to study the design of a novel high-Q fully integrated switched capacitor bandpass filter. This circuit, implemented in CMOS technology, allows a tunable high selectivity over a broad frequency band. The proposed architecture is intended to replace passive surface acoustic wave (SAW) filters in low-cost wireless radio-communication applications. To show the feasibility of the proposed filter a prototype has been fabricated and tested. Measurements show quality factors up to 300, and a tunable center frequency range of 290 MHz [240-530 MHz] with a bandwidth tuning., Anglais

This paper presents the design and the behavior vs. temperature of RF antenna switches in a 130 nm SOI technology. The design is implemented using two types of transistors; floating body and body tied transistors. It is shown that the floating body transistor is the best candidate for the design of RF antenna switches implemented in a fully integrated RF communication system. Outstanding high temperature behavior is also emphasized on a temperature range from 25 degrees C to., Anglais

This paper reviews and analyzes a fully integrated low-noise amplifier (LNA) for low-power and high temperature applications, in 130 nm partially depleted silicon-on-insulator (SOI) CMOS technology. The LNA has been characterized over a temperature range from 25 to 200 degrees C and designed using a cascode inductive source degeneration topology. Thanks to the SOI technology and the choice of the zero-temperature-coefficient (ZTC) bias point, the LNA measurements show a minor degradation of the gain due to the temperature variation for a power consumption of 2.3 mW under 1.2 V supply is applied. The effects of high temperature are observed on the gain of the LNA and on the SOI transistors in order to analyze the behavior of the LNA versus temperature effect., Anglais

A new compact DC and large-signal physics-based non-quasi-static (NQS) MOSFET model is proposed, suitable for partially-depleted body-tied silicon-on-insulator (SOI) MOSFETs. The developed four-terminal nonlinear model is comprised solely of fully-continuous explicit expressions that allow for quick simulation times and accurate intermodulation distortion analysis. Nonlinear depletion capacitances, as well as short-channel effects such as drain-induced barrier lowering, have been included, while the self-heating effects of SOI MOSFETs are modeled using a first-order thermal RC circuit. NQS distributed channel behaviour is accounted for by using a segmented channel model allowing the model in principle to operate up to high mm-wave frequencies. The model is verified for DC and large signal one-/multi-tone operation as well as through WCDMA measurements.