Rozprawy doktorskie na temat „TIME-DOMAIN CHANNEL”

Complex-valued representation of a two-component real-valued time series yields additional physical insights that are lost otherwise. The spectral representation theorem allows us to study covariance stationary complex-valued random sequences in the frequency domain, and this is known as rotary spectral analysis. It is a widely-used technique for studying elliptical motions in ocean currents, wind etc. An important and useful parameter in rotary spectral analysis of scalar complex-valued time series is the rotary coefficient. It measures the tendency of vectors to rotate in a clockwise or counter-clockwise manner. We derive the theoretical distribution of the rotary coefficient estimator and apply our results to ocean current speed and direction measurements at six depths in the Labrador Sea. Canonical correlation techniques are commonly employed in the analysis of a pair of vector-valued random variables. We introduce a framework to extend classical multivariate analysis techniques such as canonical correlation analysis, partial least squares, and multivariate linear regression, to define coherence - a measure of correlation in the frequency domain. In the statistical analysis of complex-valued time series, we refer to a time series as proper/improper according to whether it is uncorrelated/correlated with its complex conjugate. In earlier work, complex-valued signals were assumed to be proper for the simple reason that it led to a simpler algebra. However, the loss in performance caused by overlooking the potential impropriety of such data is realized to be significant, and therefore, when the data is improper, information contained in the complementary covariance structure must be considered. Since impropriety in the time domain may not necessarily correspond to impropriety at all frequencies, we propose a generalized likelihood ratio test which may be used to test propriety of a discrete time complex-valued process at a given frequency. Finally, the idea of vector circulant embedding is exploited to yield a frequency domain bootstrap methodology. With the help of three example parameters involved in the study of multi-channel complex-valued time series, we illustrate how our method allows us to draw statistical inference such as confidence intervals. Our method can prove useful in cases where no theoretical distributional results are available, or to check the effect of nuisance parameter estimates where theoretical results are available.

The Nanopore Detector using á-hemolysin channel transcribes kinetics of a single molecule along the nanometer-scale pore. The transcribed data is represented by electrical measurements. We present accurate and computationally inexpensive tools to analyze single molecule kinetics. The HMM-EM level projection method de-noises data, retaining the transitions with very high precision. This approach doesn't require input number of levels. Another advantage is the minimal tuning required. The levels are then identified using Finite State Automata (FSAs). Spike Detector algorithm analyzes spikes characterizing behavior of molecule in pore. No commercial tools available are capable of analyzing spikes in presence of noise. The formulation of HMM-EM, FSAs and Spike Detector together provides a robust method for analysis of channel current data. Application of these methods is described for Vercoutere channel blockade dataset which contains signals of radiated and non-radiated molecules. The tools developed were used successfully to differentiate between these two molecules.

This research is the first of its kind to investigate the utilisation of a multi-threading software-based countermeasure to mitigate Side Channel Analysis (SCA) attacks, with a particular focus on the AES-128 cryptographic algorithm. This investigation is novel, as there has not been a software-based countermeasure relying on multi-threading to our knowledge. The research has been tested on the Atmel microcontrollers, as well as a more fully featured system in the form of the popular Raspberry Pi that utilises the ARM7 processor. The main contributions of this research is the introduction of a multi-threading software based countermeasure used to mitigate SCA attacks on both an embedded device and a Raspberry Pi. These threads are comprised of various mathematical operations which are utilised to generate electromagnetic (EM) noise resulting in the obfuscation of the execution of the AES-128 algorithm. A novel EM noise generator known as the FRIES noise generator is implemented to obfuscate data captured in the EM field. FRIES comprises of hiding the execution of AES-128 algorithm within the EM noise generated by the 512 Secure Hash Algorithm (SHA) from the libcrypto++ and OpenSSL libraries. In order to evaluate the proposed countermeasure, a novel attack methodology was developed where the entire secret AES-128 encryption key was recovered from a Raspberry Pi, which has not been achieved before. The FRIES noise generator was pitted against this new attack vector and other known noise generators. The results exhibited that the FRIES noise generator withstood this attack whilst other existing techniques still leaked out secret information. The visual location of the AES-128 encryption algorithm in the EM spectrum and key recovery was prevented. These results demonstrated that the proposed multi-threading software based countermeasure was able to be resistant to existing and new forms of attacks, thus verifying that a multi-threading software based countermeasure can serve to mitigate SCA attacks.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2004.
Includes bibliographical references (leaves 98-103). Also available in electronic version. Access restricted to campus users.

Thesis (M.S. in Electrical Engineering) Naval Postgraduate School, December 1998.
"December 1998." Thesis advisor(s): Jovan Lebaric, Clark Robertson, David C. Jenn. Includes bibliographical references (p. 127). Also available online.

Ultra-wideband (UWB) communication has been the subject of extensive research in recent years due to its unique capabilities and potential applications, particularly in short-range multiple access wireless communications. However, many important aspects of UWB-based communication systems have not yet been thoroughly investigated. The propagation of UWB signals in indoor environments is the single most important issue with significant impacts on the future direction, scope, and generally the extent of the success of UWB technology. The objective of this dissertation is to obtain a more thorough and comprehensive understanding of the potentials of UWB technology by characterizing the UWB communication channels. Channel characterization refers to extracting the channel parameters from measured data. The extracted parameters are used to quantify the effect of the channel on communication UWB systems using this channel as signal transmission medium. Data are measured in different ways using a variety of time-domain and frequency-domain techniques. The experimental setups used in channel characterization effort also include pulse generators and antennas as integral parts of the channel, since the pulse shape and antenna characteristics have significant impact on channel parameters. At a fundamental level, the propagation of UWB signals, as any electromagnetic wave, is governed, among other things, by the properties of materials in the propagation medium. One of the objectives of this research is to examine propagation through walls made of typical building materials and thereby acquire ultra-wideband characterization of these materials. The loss and the dielectric constant of each material are measured over a frequency range of 1 to 15 GHz. Ten commonly used building materials are chosen for this investigation. These include, dry wall, wallboard, structure wood, glass sheet, bricks, concrete blocks, reinforced concrete (as pillar), cloth office partition, wooden door, and styrofoam slab. The work on ultra-wideband characterization of building materials resulted in an additional interesting contribution. A new formulation for evaluating the complex dielectric constant of low-loss materials, which involves solving real equations and thus requiring only one-dimensional root searching techniques, was found. The results derived from the exact complex equation and from the new formulation are in excellent agreement. Following the characterization of building materials, an indoor UWB measurement campaign is undertaken. Typical indoor scenarios, including line-of-sight (LOS), non-line-of-sight (NLOS), room-to-room, within-the-room, and hallways, are considered. Results for indoor propagation measurements are presented for local power delay profiles (local-PDP) and small-scale averaged power delay profiles (SSA-PDP). Site-specific trends and general observations are discussed. The results for pathloss exponent and time dispersion parameters are presented. The analyses results indicate the immunity of UWB signals to multipath fading. The results also clearly show that UWB signals, unlike narrowband signals, do not suffer from small scale fading, unless the receiver is too close to walls. Multipath components are further studies by employing a deconvolution technique. The application of deconvolution results in resolving multipath components with waveforms different from those of the sounding pulse. Resolving more components can improve the design of the rake receiver. The final part of this research elaborates on the nature of multiple access interference and illustrates the application of multi-user detection to improve the performance of impulse radio systems. Measured dispersion parameters and their effects on the multiple access parameters are discussed.
Ph. D.

In automatic speech recognition systems (ASRs), training is a critical phase to the system?s success. Communication media, either analog (such as analog landline phones) or digital (VoIP) distort the speaker?s speech signal often in very complex ways: linear distortion occurs in all channels, either in the magnitude or phase spectrum. Non-linear but time-invariant distortion will always appear in all real systems. In digital systems we also have network effects which will produce packet losses and delays and repeated packets. Finally, one cannot really assert what path a signal will take, and so having error or distortion in between is almost a certainty. The channel introduces an acoustical mismatch between the speaker's signal and the trained data in the ASR, which results in poor recognition performance. The approach so far, has been to try to undo the havoc produced by the channels, i.e. compensate for the channel's behavior. In this thesis, we try to characterize the effects of different transmission media and use that as an inexpensive and repeatable way to train ASR systems.

Finite-difference models have been used for nearly 40 years to solve electromagnetic problems of heterogeneous nature. Further, these techniques are well known for being computationally expensive, as well as subject to various numerical artifacts. However, little is yet understood about the errors arising in the simulation of wideband sources with the finitedifference time-domain (FDTD) method. Within this context, the focus of this thesis is on two different problems. On the one hand, the speed and accuracy of current FDTD implementations is analysed and increased. On the other hand, the distortion of numerical pulses is characterised and mitigation techniques proposed. In addition, recent developments in general-purpose computing on graphics processing units (GPGPU) have unveiled new methods for the efficient implementation of FDTD algorithms. Therefore, this thesis proposes specific GPU-based guidelines for the implementation of the standard FDTD. Then, metaheuristics are used for the calibration of a FDTD-based narrowband simulator. Regarding the simulation of wideband sources, this thesis uses first Lagrange multipliers to characterise the extrema of the numerical group velocity. Then, the spread of numerical Gaussian pulses is characterised analytically in terms of the FDTD grid parameters. The usefulness of the proposed solutions to the previously described problems is illustrated in this thesis using coverage and wideband predictions in large-scale scenarios. In particular, the indoor-to-outdoor radio channel in residential areas is studied. Furthermore, coverage and wideband measurements have also been used to validate the predictions. As a result of all the above, this thesis introduces first an efficient and accurate FDTD simulator. Then, it characterises analytically the propagation of numerical pulses. Finally, the narrowband and wideband indoorto-outdoor channels are modeled using the developed techniques.

Cette thèse doctorale aborde divers aspects méthodologiques de l‟analyse de levés électromagnétiques aéroportés en domaine temporel (TDEM) pour une interprétation détaillée à finalités géologique et hydrogéologique. Ce travail s‟est appuyé sur un levé réalisé dans la région de Courtenay (Nord-Est de la région Centre) caractérisée par un plateau de craie karstifié (karst des Trois Fontaines) recouvert par des argiles d‟altération et des alluvions. Tout d‟abord, une méthode de filtrage des données TDEM utilisant la Décomposition en Valeurs Singulières (SVD) a été développée. L‟adaptation rigoureuse de cette technique aux mesures TDEM a permis de séparer avec succès les bruits, qui ont pu être cartographiés, et le " signal géologique ", diminuant grandement le temps nécessaire à leur traitement. De plus, la méthode s‟est avérée efficace pour obtenir, rapidement, des informations géologiques préliminaires sur la zone. Ensuite, une analyse croisée entre le modèle de résistivité obtenu en inversant les données filtrées et les forages disponibles a été effectuée. Celle-ci a mené à une amélioration de la connaissance géologique et hydrogéologique de la zone. Une figure d‟ondulation, séparant deux dépôts de craie, et le réseau de failles en subsurface ont pu être imagés, apportant un cadre géologique au karst des Trois Fontaines. Enfin, une nouvelle méthode combinant l‟information aux forages et les pentes issues du modèle de résistivité EM a permis d‟obtenir un modèle d‟une précision inégalée du toit de la craie. L‟ensemble de ces travaux fournit un cadre solide pour de futures études géo-environnementales utilisant des données TDEM aéroportées, et ce, même en zone anthropisée.

Telemetry system, Optimisation, Sensoric networks, Smart Grid, Internet of Things, Sensors, Information security, Cryptography, Cryptography algorithms, Cryptosystem, Confidentiality, Integrity, Authentication, Data freshness, Non-Repudiation.

A comparative analysis of time-domain (TD) solution, based on an established UTD diffraction model, are presented for modeling of Ultra-wideband (UWB) signal for lossy dielectric obstacles which gives accurate result to any arbitrary position of transmitter and receiver in a complex channel environment. Obstacles considered in the work include dielectric Wedge with homogenous, isotropic, low-loss dielectric characteristics. Different UTD-TD Diffraction Coefficients are proposed and compared with the IFFT of rigorous (i.e., Maliuzhinets) diffraction coefficient (RDC). The proposed work provides an in-depth analysis of the UTD model and presents an accurate and computationally more efficient TD solution for the available UTD diffraction coefficients for lossy dielectric medium, for both soft and hard polarizations. Moreover the reciprocity and symmetry for the diffraction coefficient in the time-domain have been proven for different position of transmitter and receiver. The time-domain modeling for transmission and reflection of UWB signals for 2-D & 3-D multi-modeled obstacle is also done. The obstacle is called multi-modeled since the obstacle consists of two entirely different structure i.e. dielectric wedge followed by dielectric slab. The comparison between the TD solution and the numerical inverse fast Fourier transform of the FD (IFFT-FD) solution proves the accuracy of the proposed solution. The significant gain in the computational speed achieved through the proposed TD solution is demonstrated by comparing its computation time with that of the exact IFFT-FD solution.

碩士
國立成功大學
電腦與通信工程研究所
98
In wireless communications, receiver needs the information of the fading channel to equalize or mitigate the distortion of the received signal. We need frequency response of the channel (CFR) to determine coefficients of the equalizer. In this thesis we estimate the channel response of OFDM systems by inserting among data symbols some pilot signals that is known to the receiver. We consider both the pilot signals in time and frequency domains. We use time-domain (TD) and frequency-domain (FD) estimations. For the TD estimation approach, we directly transmit one pulse whose amplitude is known to the receiver in time domain before every data block, and we can easily obtain the whole instant CIR and decide the path number at the receiver. For the FD estimation approaches, we insert pilots in some subcarriers (pilot subcarriers). The CFR at other sub-channels are obtained by linear interpolation. Then we propose a two-dimension method by allowing channel estimation processed after receiving several OFDM blocks, and estimate the channels jointly to obtain better performance. Finally, we study the effect of the channel delays that are not on sampling instant exactly, and use the oversampling to compensate the distortions.

碩士
國立交通大學
電信工程系
90
In a discrete multitone (DMT) receiver, a time-domain equalizer (TEQ) is used to shorten the channel impulse response. The general criterion for the TEQ design is to maximize bit loading. To do that, all noise sources have to be identified and taken into signal to noise ration (SNR) calculation. There are mainly two noise sources in a DMT system; one is the channel noise and the other is the residual intersymbol interference (ISI). Recently, the residual ISI has been identified as a noise source and a method to achieve maximum bit loading was proposed. Unfortunately, this method does not consider the special effect of channel noise. In this thesis, we intend to solve this problem. We first show that noise can be more damaging than it looks. This can be explained by the fact that the transmitted signal has a cyclic prefix (CP) while the noise signal does not. Thus, the SNR (for each tone) in the TEQ output is lower than that in its input. We then propose a TEQ design method that can simultaneously take the residual ISI and the channel noise into account. Simulations show that our method outperforms all existing TEQ design algorithms.

碩士
國立中央大學
通訊工程研究所
95
To deal with the channel estimation problem, the mobile orthogonal frequency division multiplexing (OFDM) system usually requires pilots to be inserted in transmitted signals in which the bandwidth efficiency is reduced. For the purpose of saving pilots, this paper studies a new least squares (LS) time-domain channel estimator using a decision-feedback loop. However, the drawback of the decision-feedback system is decision-error propagation. We propose error detectors for reducing the error probability and some compensation methods for enhancing channel estimation. In simulation results, the performances of the proposed decision-error detectors and compensation methods are compared. We show that the acceptable bit error rate can be achieved without pilots used for channel tracking and thus, the loss of bandwidth efficiency due to pilots is significantly reduced.

碩士
國立成功大學
電腦與通信工程研究所
97
In wireless communication environments, the channel must be estimated for equalization in the receiver. The accuracy of channel information affects the system performance. In this thesis, we will introduce and analyze a simple channel impulse response (CIR) estimation method in which a unit pulse followed by a series of consecutive zeros is inserted ahead of each time domain OFDM block signal. Another more popular method is to estimate the channel frequency response (CFR) in which some pilot symbols are inserted in the frequency-domain signal, and we refer to it as “FD estimation”. For the CIR estimation method, we can directly transmit one pulse before every data block. In this thesis, we call it “TD estimation”. At the receiver, we can easily obtain the whole instant CIR and decide the path number. Unlike the TD estimation, FD estimation just only estimates the CFR at pilot sub-channels. The CFR at other sub-channels are obtained by interpolation or least-squares fitting, linear minimum mean-squares error and so on. In this thesis, we will compare the performance between TD and FD estimation. Finally, we study the performance of TD estimation in a slow time-variant multipath channel. We use the linear interpolation method to approximate the variation of CIR. And then a simple frequency-domain equalizer technique is also introduced to compensate the effect of interchannel interference (ICI) due to time-variant channel.

An accurate time-domain (TD) modeling for transmission of UWB signal through a 3- dimensional building structure made of low loss dielectric materials is presented. Transmission through building is analyzed for arbitrary position of transmitter and receiver. Time-domain formulation are presented for transmission coefficient for both soft and hard polarizations, for propagation loss and then for the transmitted field at the receiver. An analytical time domain (TD) solution, based on an established FD diffraction model, is also presented for diffraction of UWB signal through same dielectric obstacle. The comparison of transmitted and reflected field proves that the transmitted field component is very significant in case of non line of sight (NLOS) communication in deep shadow regions, where the reflected and diffracted components are weak. The TD results have been validated with the inverse fast Fourier transform (IFFT) of the corresponding exact frequency domain results. The significant gain in the computational speed achieved through the proposed TD method is demonstrated by comparing its computation time with that of the exact IFFT-FD solution. Finally comparative analysis of normalized mean and mean square error between both methods is performed for different loss tangent values. It is shown that the proposed TD solution outperforms the exact IFFT-FD solution in terms of computational efficiency. Also it is observed that normalized error tends to increase for higher loss tangent values.

碩士
國立成功大學
電腦與通信工程研究所
100
he single-carrier block transmission system can use the low-complexity frequency-domain equalization, and it does not have the problem of high Peak-to-Average Power Ratio (PAPR), like that in the orthogonal frequency-division multiplexing (OFDM) system. To equalize the multipath channel, the receiver needs to estimate the channel responses. In this thesis, we study some channel estimation algorithms and schemes for the single-carrier block transmission system. We consider 1D and 2D channel estimation schemes based on frequency-domain and time-domain pilot signals, respectively. For the time-domain channel estimation, pilot signal is an impulse ahead of each transmitted block, while for the frequency-domain channel estimation, pilot signals are composed of pilot symbols across subchannels. The Least Square Fitting (LSF) principle is applied in the channel estimation. We also study the 2D schemes for channel estimation based on time-domain and frequency-domain pilot signals. The 2D schemes results in better performance at the price of longer delay and higher complexity.

碩士
國立交通大學
電信工程系所
95
In time-variant OFDM systems, channel estimation usually relies on pilot subcarriers. However, the number of pilot subcarriers is usually limited. Channel estimation is then a critical task for receiver design. In this thesis, we first propose a two-dimension slide-window channel estimator for IEEE802.16e systems. The estimator can outperform conventional approaches and requires low-complexity. To further improve the performance, we then propose a high-performance least-squares (LS) channel estimator, joint operated in the time and frequency domains. The main idea is use the channel response, estimated in the frequency domain, to locate significant time-domain channel taps, and then use the LS method to estimate the responses in those taps. With an iterative algorithm, we can then obtain accurate channel estimate with computational complexity much lower than the conventional time domain LS estimator.

碩士
國立成功大學
電腦與通信工程研究所
97
For the multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) system, the receiver needs accurate channel estimation to achieve good performance. In this thesis, we first compare the frequency domain and time domain channel estimation. For the frequency domain channel estimation, the linear minimum mean-squares error (LMMSE) algorithm can attain the optimal performance, but its complexity is quite high. For the time domain channel estimation, a unit pulse is put ahead every OFDM symbol block, and the receiver can readily obtain the time domain channel response. The simulation shows that the time domain algorithm attains better performance under the same data rate and pilot signal energy. Implied by the time domain approach, we also propose a frequency domain channel estimation algorithm based on the time domain channel model and the least-squares fitting. The simulation shows that the proposed algorithm results mean-squares error (MSE) and bit-error rate (BER) near those of the LMMSE algorithm.

碩士
國立中正大學
電機工程研究所
91
While signal passes through multipath fading channel, the signal will suffer a certain degree of distortion and time delay. There are some important issues about the synchronization for the signal and how to compensate the signal in receiver while the signal passes through the multipath fading channel. In this thesis, we add pilots in time domain to evaluate the issues in timing synchronization and channel estimation for OFDM systems. Here, the pilots we added are the PN code. First, we use the characteristics of PN code autocorrelation to achieve the timing synchronization and delay time estimation for each delay path. Then, we compute each channel tap gain by maximum likelihood algorithm. Besides, the linear regression model is applied to implement synchronization and channel estimation. Through the data matrix we design, we can estimate each channel tap gain in time domain. Because the pilots we add are the PN code, we can synchronize and estimate channel impulse response over multipath fading channels not only with integer time delays but also with fractional time delays through spreading the PN code. After the channel impulse response is estimated in time domain, we compensate the signal in time domain with our estimated channel response. Because the issues of computation complexity and error propagation, we transfer our estimated channel in time domain to channel transfer function in frequency domain. Then, we compensate the signal in frequency domain through one-tap equalizers. By the simulation results, we have better performance than traditional systems with pilots that are added in frequency domain.

碩士
國立交通大學
電信工程系所
97
In DVB-T systems, the system performance heavily relies on that of the channel estimation. Recently, a join time-and-frequency-domain channel estimator has been proposed. It has been shown that its performance is significantly better than the conventional frequency domain estimator. However, the performance of the join time-and-frequency-domain channel estimator depends on the number of pilot subcarrier it uses. In the first part of this thesis, we propose an improved time-and-frequency-domain channel estimation method that can have high performance even when the number of pilot subcarriers is low. It has been well-known that the intercarrier-interference (ICI), induced when OFDM systems operated in high-mobility environments, can degrade the system performance seriously. Existing ICI cancellation methods often require high computational complexity. In the second part of the thesis, we propose a number of low-complexity ICI cancellation methods. Simulations show that the proposed algorithms can have the performance similar to existing method, while the computational complexity is significantly lower.

碩士
國立交通大學
電機學院通訊與網路科技產業專班
96
Compared with traditional analog video broadcast, digital video broadcast (DVB) has many advantages such as high definition, reduced bandwidth, and high interference immunity. However, the actual performance greatly depends on the receiver design. The purpose of this thesis is to design a high-performance DVB-T receiver. First, we built a complete platform for the DVB-T system, which includes the transmitter and the receiver. The receiver consists of the inner and outer receivers; the former conducts synchronization, and the later forward error decoding. Then, we focus on the design of the high-performance channel estimator. Conventional approaches for channel estimation is conducted in the frequency domain, and the performance is not optimal. We use a newly developed joint time-and-frequency- domain channel estimator to solve the problem. Simulations show that the channel estimator has better performance that conventional methods, specially when the channel delay spread is large.

碩士
國立東華大學
電機工程學系
104
Due to the rapid growth of wireless networks and the requirement on the quality of service (QoS), the improvement of the spectrum utilization efficiency under busy and congested urban environment has become an important issue. Cognitive radio (CR) is the best technique to fulfill this task. Without affecting the transmission of the primary users (PUs), the secondary users (SUs) can detect the idle channels and send their information via these channels. The theories of economics were used in some previous research to improve the spectrum utilization efficiency. In this thesis, the auction theory is incorporated into CR for the allocation of channels so that the number of users can be increased and the efficiency can be enhanced. The results are further compared with those from the general allocation of channels in CR. Keywords : Cognitive radio ; Auction theory

Being one of the most spectacular events in nature, lightning is basically a transient high current electric discharge in the atmosphere, which extends up to kilometres. Cloud to ground discharge is the most hazardous one as far as ground based structures are considered. Among the different phases of a lightning flash, return stroke is considered to be the most energetic phase and is basically responsible for most of the damages. Hence, much emphasis has been given to return stroke modeling. A more realistic modeling of return stroke is very essential to accurately study the interaction of return stroke with the structures on ground. As return stroke is dominated by electromagnetic phenomenon, an electromagnetic model will be the most suitable one. It does not call for any assumption on the mode of wave propagation, as well as, electromagnetic coupling between the different channel portions. There are mainly two approaches adopted for electromagnetic models i.e. frequency domain and time domain approach. Time domain approach is more reliable as it can handle, in principle, the nonlinear processes in the lightning channel. It is also free of numerical frequency domain to time domain inversion problem, which are found to be quite severe. However, most of the previous works on time domain electromagnetic models suffered from following two serious limitations - (i) the initial charge on the channel, which forms the true excitation for the problem, is not considered and (ii) instead of the non-linearly rising conductivity of the channel, a constant resistance is employed. For a realistic simulation of the interaction between the channel and any intercepting system, a time domain model with the above two major aspects being fully represented is very essential. In an earlier work, all these aspects have been fully considered but a domain based numerical modelling was employed. Consequently, it was difficult to consider the down conductor and further the number of unknowns was considerably large. In view of this, the present work is taken up and its scope is defined as to develop a boundary based numerical time domain electromagnetic model in which the initial charge on the channel and the non-linearly evolving channel conductance are fully considered. For the electrical engineering applications, electromagnetic aspects of the lightning phenomena is more important than the other associated physical processes and hence, importance is given only to the electromagnetic aspects. In other words, the light emission, thunder, chemical reactions at the channel etc. are not considered. Also, for most of the electrical engineering applications, the critical portion of current would be the region up to and around the peak and hence, modeling for this regime will be given prime importance. Owing to the complexity of the problem, some simplifying assumptions would be very essential. The literature indicates that these assumptions do not affect the adequate representation of the phenomena. Lightning channel is considered to be vertically straight without any branches. Earth is considered to be perfectly conducting. Explicit reference to dynamically varying channel radius, temperature and the air density is not made. However, it is assumed that the arc equation employed to describe the temporal changes in conductivity would adequately take care of these parameters. Lightning channel is represented by a highly conducting small core, which is surrounded by a weakly conducting corona sheath. The initial charge on the channel is deduced by solving for electrostatic field, with leader portion set to possess an axial gradient of 60 V/cm and the streamer portion to 5 – 10 kV/cm. The radius of the corona sheath is set iteratively by enforcing a gradient of 24 kV/cm up to its radial boundary. As analytical solution for the problem is impractical, suitable numerical solution is sought. Since the spatial extension of this time marching problem is virtually unbound and that the significant conduction is rather solely confined to an extremely small cross section of the channel core, a boundary-based method is selected. Amongst the numerical methods, the present work employs the moment method for the solution of the fields associated with the return strokes. A numerical solution of the Electric Field Integral Equation (EFIE) for thin structures has been developed in the literature. The same approach has been employed in the present work, however, with suitable modifications to suit the lightning problem. The code was written in MATLAB and integrations involved in the EFIE were solved using MATLAB symbolic computation. Before introducing the channel dynamic conductance and the initial charge on the channel, the code developed is validated by comparing the results for a center fed dipole antenna with that given in the literature. Also, NEC (Numeric Electromagnetic Code) simulations for various cases of monopole and dipole antenna were performed. The results from the code developed are shown to have good matching with that obtained from NEC based time domain results. In an earlier work, the dynamic conductance of the return stroke channel core, which is a high current electric arc, was represented by a first order arc equation. The same approach is employed in the present work also. Similarly, the transition from streamer to leader was modeled by Braginskii’s spark law and the same has been considered in the present work. A value of 10-5 S/m was used for minimum value of streamer conductance. For numerical stability, upper (Gmax = 3 S/m) and lower bounds (Gmin = 0.0083 S/m) for the channel conductance are forced. Preliminary simulations were run with and without dynamic channel conductance. The initial charge distribution along the channel formed the excitation. Results clearly show that without the dynamically varying channel conductance, no streamer to leader transition and hence, no return stroke evolution can occur. In other words, the non-linearly evolving channel conductance is mainly responsible for the evolution of the return stroke. In order to consider the charge neutralization by the return stroke, the charge deposited by it is diffused into the corona sheath. A fixed value of the corona sheath conductance is employed and the diffusion process is modeled by an equation derived from the continuity equation. To study the effect of corona sheath, simulations were run with and without corona. From the simulation results it was observed that the corona sheath causes increase in peak value of the stroke current, as well as, time to front and a decrease in the velocity of propagation. For the validation of the model, the basic characteristics of the return stroke current like the current wave shape, temporal variation of stroke current at different heights, velocity of propagation and the vertical electric fields at various radial distances were compared with available field/experimental data. A good agreement was seen and based on this, it is concluded that the present work has successfully developed a boundary based time domain numerical model for the lightning return stroke. Natural lightning being a stochastic process, the values of the parameters associated with it would differ in every event. On other hand, any deterministic model like the one developed in the present work predicts a fixed pattern of the simulated quantities. Therefore, it was felt that some of the model parameters must be permitted to vary so that a range of results could be obtained rather than a single pattern of results. Incidentally, the model parameters like arc time constant, settling value of arc conductivity/gradient, bounds for channel conductivity, streamer gradient, radius of the core etc. are not precisely known for the natural lightning environment. Further, some of them are known to vary within an event. Considering these and that simplicity is very important in already complex model, the above-mentioned parameters are taken as tunable parameters (of course to be varied within the prescribed range) for deducing the return stroke currents with some desired characteristics. A study on the influence of these parameters is made and suggestions are provided. Simulations for the nominal range of stroke currents are made and results are presented. These simulations clearly show the role of cloud potential, which in turn dictates the length of final bridging streamer, on the return stroke currents. The spatio-temporal variation of the current, charge deposited by the return stroke and the channel conductivity are presented which, reveal the dynamic processes leading to the evolution of return stroke current. Subsequently, simulations for two cases of stroke to elevated strike object are attempted. The upward leader was modeled quite similar to the descending one. Many interesting findings are made. In summary, the present work has successfully developed a boundary-based time domain numerical electromagnetic model for the lightning return stroke, wherein, the initial charge deposited on the channel and the non-linearly rising channel conductance have been appropriately considered. Simulation using the model clearly depicts the dynamic evolution of the return stroke. The characteristics of the simulated return strokes are in good agreement with the field data. Some of the parameters of the model are suggested as tunable parameters, which permit simulation of strokes with different characteristics.

Being one of the most spectacular events in nature, lightning is basically a transient high current electric discharge in the atmosphere, which extends up to kilometres. Cloud to ground discharge is the most hazardous one as far as ground based structures are considered. Among the different phases of a lightning flash, return stroke is considered to be the most energetic phase and is basically responsible for most of the damages. Hence, much emphasis has been given to return stroke modeling. A more realistic modeling of return stroke is very essential to accurately study the interaction of return stroke with the structures on ground. As return stroke is dominated by electromagnetic phenomenon, an electromagnetic model will be the most suitable one. It does not call for any assumption on the mode of wave propagation, as well as, electromagnetic coupling between the different channel portions. There are mainly two approaches adopted for electromagnetic models i.e. frequency domain and time domain approach. Time domain approach is more reliable as it can handle, in principle, the nonlinear processes in the lightning channel. It is also free of numerical frequency domain to time domain inversion problem, which are found to be quite severe. However, most of the previous works on time domain electromagnetic models suffered from following two serious limitations - (i) the initial charge on the channel, which forms the true excitation for the problem, is not considered and (ii) instead of the non-linearly rising conductivity of the channel, a constant resistance is employed. For a realistic simulation of the interaction between the channel and any intercepting system, a time domain model with the above two major aspects being fully represented is very essential. In an earlier work, all these aspects have been fully considered but a domain based numerical modelling was employed. Consequently, it was difficult to consider the down conductor and further the number of unknowns was considerably large. In view of this, the present work is taken up and its scope is defined as to develop a boundary based numerical time domain electromagnetic model in which the initial charge on the channel and the non-linearly evolving channel conductance are fully considered. For the electrical engineering applications, electromagnetic aspects of the lightning phenomena is more important than the other associated physical processes and hence, importance is given only to the electromagnetic aspects. In other words, the light emission, thunder, chemical reactions at the channel etc. are not considered. Also, for most of the electrical engineering applications, the critical portion of current would be the region up to and around the peak and hence, modeling for this regime will be given prime importance. Owing to the complexity of the problem, some simplifying assumptions would be very essential. The literature indicates that these assumptions do not affect the adequate representation of the phenomena. Lightning channel is considered to be vertically straight without any branches. Earth is considered to be perfectly conducting. Explicit reference to dynamically varying channel radius, temperature and the air density is not made. However, it is assumed that the arc equation employed to describe the temporal changes in conductivity would adequately take care of these parameters. Lightning channel is represented by a highly conducting small core, which is surrounded by a weakly conducting corona sheath. The initial charge on the channel is deduced by solving for electrostatic field, with leader portion set to possess an axial gradient of 60 V/cm and the streamer portion to 5 – 10 kV/cm. The radius of the corona sheath is set iteratively by enforcing a gradient of 24 kV/cm up to its radial boundary. As analytical solution for the problem is impractical, suitable numerical solution is sought. Since the spatial extension of this time marching problem is virtually unbound and that the significant conduction is rather solely confined to an extremely small cross section of the channel core, a boundary-based method is selected. Amongst the numerical methods, the present work employs the moment method for the solution of the fields associated with the return strokes. A numerical solution of the Electric Field Integral Equation (EFIE) for thin structures has been developed in the literature. The same approach has been employed in the present work, however, with suitable modifications to suit the lightning problem. The code was written in MATLAB and integrations involved in the EFIE were solved using MATLAB symbolic computation. Before introducing the channel dynamic conductance and the initial charge on the channel, the code developed is validated by comparing the results for a center fed dipole antenna with that given in the literature. Also, NEC (Numeric Electromagnetic Code) simulations for various cases of monopole and dipole antenna were performed. The results from the code developed are shown to have good matching with that obtained from NEC based time domain results. In an earlier work, the dynamic conductance of the return stroke channel core, which is a high current electric arc, was represented by a first order arc equation. The same approach is employed in the present work also. Similarly, the transition from streamer to leader was modeled by Braginskii’s spark law and the same has been considered in the present work. A value of 10-5 S/m was used for minimum value of streamer conductance. For numerical stability, upper (Gmax = 3 S/m) and lower bounds (Gmin = 0.0083 S/m) for the channel conductance are forced. Preliminary simulations were run with and without dynamic channel conductance. The initial charge distribution along the channel formed the excitation. Results clearly show that without the dynamically varying channel conductance, no streamer to leader transition and hence, no return stroke evolution can occur. In other words, the non-linearly evolving channel conductance is mainly responsible for the evolution of the return stroke. In order to consider the charge neutralization by the return stroke, the charge deposited by it is diffused into the corona sheath. A fixed value of the corona sheath conductance is employed and the diffusion process is modeled by an equation derived from the continuity equation. To study the effect of corona sheath, simulations were run with and without corona. From the simulation results it was observed that the corona sheath causes increase in peak value of the stroke current, as well as, time to front and a decrease in the velocity of propagation. For the validation of the model, the basic characteristics of the return stroke current like the current wave shape, temporal variation of stroke current at different heights, velocity of propagation and the vertical electric fields at various radial distances were compared with available field/experimental data. A good agreement was seen and based on this, it is concluded that the present work has successfully developed a boundary based time domain numerical model for the lightning return stroke. Natural lightning being a stochastic process, the values of the parameters associated with it would differ in every event. On other hand, any deterministic model like the one developed in the present work predicts a fixed pattern of the simulated quantities. Therefore, it was felt that some of the model parameters must be permitted to vary so that a range of results could be obtained rather than a single pattern of results. Incidentally, the model parameters like arc time constant, settling value of arc conductivity/gradient, bounds for channel conductivity, streamer gradient, radius of the core etc. are not precisely known for the natural lightning environment. Further, some of them are known to vary within an event. Considering these and that simplicity is very important in already complex model, the above-mentioned parameters are taken as tunable parameters (of course to be varied within the prescribed range) for deducing the return stroke currents with some desired characteristics. A study on the influence of these parameters is made and suggestions are provided. Simulations for the nominal range of stroke currents are made and results are presented. These simulations clearly show the role of cloud potential, which in turn dictates the length of final bridging streamer, on the return stroke currents. The spatio-temporal variation of the current, charge deposited by the return stroke and the channel conductivity are presented which, reveal the dynamic processes leading to the evolution of return stroke current. Subsequently, simulations for two cases of stroke to elevated strike object are attempted. The upward leader was modeled quite similar to the descending one. Many interesting findings are made. In summary, the present work has successfully developed a boundary-based time domain numerical electromagnetic model for the lightning return stroke, wherein, the initial charge deposited on the channel and the non-linearly rising channel conductance have been appropriately considered. Simulation using the model clearly depicts the dynamic evolution of the return stroke. The characteristics of the simulated return strokes are in good agreement with the field data. Some of the parameters of the model are suggested as tunable parameters, which permit simulation of strokes with different characteristics.

碩士
國立交通大學
電機與控制工程系所
95
We propose an instructive derivation for the generalized block-level orthogonal space-time block encoder, capable of achieving full spatial diversity via frequency- selective fading environment provided that channel order is known. Instead of dealing with special case and then extending the results intuitively, we provide an alternative by starting with the general signal model with multiple transmit and multiple receive antennas, from which a general form of block-level orthogonality is established. In particular, transmit diversity with more than two transmit antennas can be achieved without compromise by means of frequency-domain equalization, in contrast to the QO-STBC-based approach. Pairwise error probability analysis is derived, under certain assumption which is numerically supported by simulation results, for analytical verifications of our claim on full diversity, inclusive of transmit-receive diversity and the multipath one. Moreover, the encoder structure enables us to generalize a training-based channel estimation technique, originally proposed for flat-fading scenario, to the frequency-selective fading scenario. Surprisingly we even obtain similar optimality criteria for optimal training block design which in our case, the signal block are fixed as OSTBC-based and the design derivation reduces to derive optimal power constraint over the training blocks. The optimality criteria for the training blocks are easy to satisfy when randomness of signal constellation is not a concern. Simulation results validate our discussion of the behaviors of the least-squares and linear MMSE channel estimates.

碩士
國立臺灣大學
電信工程學研究所
98
Increasing demand for higher data rates under high mobility has catalyzed sev- eral new techniques and systems. One of the most popular techniques is orthogonal frequency-division multiplexing (OFDM). The OFDM system is widely adopted in many modern communication systems because of its spectral e±ciency and its ro- bustness over multipath channels. For many mobile applications, the end users might be moving and this will introduce the doppler eRect. Due to the doppler eRect, chan- nel is not time-invariant anymore and the orthogonality among the subcarriers of the OFDM systems will be destroyed. As a result intercarrier interference (ICI) is intro- duced. ICI will degrade the performance of the system signi¯cantly. Thus, in order to be a reliable communication system, some ICI mitigation schemes are important to the system. There are many ICI mitigation schemes. One technique is to design an equalizer. By having an equalizer at the receiver, the ICI could be eRective suppressed. But the computational complexity of the conventional zero-forcing (ZF) or minimum mean square error (MMSE) equalizer will be high when the channel is time-varying. Sev- eral low cost equalizers have been proposed in the literature. In particular, it was shown [16] that most ICI comes from neighbor subcarriers. From this conclusion, a reduced complexity equalizer, Q-tap equalizer, is proposed without sacri¯cing much performance. To reduce the complexity of equalizer, another method is to add a time domain window at the receiver [17]. The purpose of adding the time domain window is to prone the channel response into a desired form. However, this idea does not reduce additive channel noise, nor does it reduces interference from other subcarriers. The result of performance seems to be less than satisfactory. Presented in this thesis is a novel method of channel equalization, and ICI reduc- tion based on the time domain window and Q-tap equalizer. By jointly designing the window and Q-tap equalizer, the system can better reduce ICI and its performance is greatly improved without a large computation. Besides, we also extend the time domain window using banded matrix. Finally, Simulation results are provided to demonstrate the performance of the proposed methods.

碩士
國立中山大學
電機工程學系研究所
94
In this thesis, an algorithm based on frequency-space domain MUSIC method is presented for estimating the propagation delay of a wireless multipath channel.For indoor geolocation systems, the time-of-arrival (TOA) is the most popular technique for accurate positioning system. The basic idea in TOA-based techniques is to accurately estimate the propagation delay of the radio signal arriving from the direct line-of-sight (DLOS) path. However, dense multipath environments may cause unresolved paths, and yield an error in the estimation of the DLOS path. UWB (Ultra-wideband) technology provides an excellent means for wireless positioning due to its high resolution capability in the time domain. Its ability to resolving multipath components makes it possible to obtain accurate location estimates. In this thesis, we investigate the use of UWB signals in positioning and combine frequency-domain MUSIC algorithm. At the same time, the structure of time-space-time method is studied. In addition, we propose a frequency-space domain MUSIC algorithm, called FSF-MUSIC algorithm, and use the spatial smoothing technique to improve the performance of the algorithm. For a two-multipath case, analysis and simulation results of multipath resolvability and the variance of estimation errors of signal arrival time are discussed.

碩士
國立臺灣大學
光電工程學研究所
104
In this thesis, bio-chemical reaction and protein complexes interaction are detected by an IGZO thin film transistor-based (TFT-based) biosensor. These two issues are discussed separately in two parts. In the first part, a bio-chemical reaction is detected. This reaction is a part of the metabolic pathway and it is a step in malate-aspartate shuttle, which is one of the most important bio-chemical system in human body. The reaction system involving different concentrations of composition are investigated. By measuring the electrical signals of TFT, the relation between current change and complete/incomplete reactions is revealed. The second part is about the detection of Streptavidin-Biotin protein complexes. The TFT-based biosensor is integrated with microfluidic channels in order to investigate the diffusion behaviors of target analytes. The diffusion experiments of Streptavidin and Biotin were first conducted separately to understand the diffusion properties. Then the binding interaction of Streptavidin-Biotin complexes is detected and analyzed in time domain. We also verified the binding among Streptavidin and Biotin by using the fluorescence microscope.

In this thesis we investigate the problem of image restoration. The main focus of our research is to come up with novel algorithms and enhance existing techniques in order to deliver efficient and effective methodologies, applicable in real-time image restoration scenarios. Our research starts with a literature review, which identifies the gaps in existing techniques and helps us to come up with a novel classification on image restoration, which integrates and discusses more recent developments in the area of image restoration. With this novel classification, we identified three major areas which need our attention. The first developments relate to non-blind image restoration. The two mostly used techniques, namely deterministic linear algorithms and stochastic nonlinear algorithms are compared and contrasted. Under deterministic linear algorithms, we develop a class of more effective novel quadratic linear regularization models, which outperform the existing linear regularization models. In addition, by looking in a new perspective, we evaluate and compare the performance of deterministic and stochastic restoration algorithms and explore the validity of the performance claims made so far on those algorithms. Further, we critically challenge the ne- cessity of some complex mechanisms in Maximum A Posteriori (MAP) technique under stochastic image deconvolution algorithms. The next developments are focussed in blind image restoration, which is claimed to be more challenging. Constant Modulus Algorithm (CMA) is one of the most popular, computationally simple, tested and best performing blind equalization algorithms in the signal processing domain. In our research, we extend the use of CMA in image restoration and develop a broad class of blind image deconvolution algorithms, in particular algorithms for blurring kernels with a separable property. These algorithms show significantly faster convergence than conventional algorithms. Although CMA method has a proven record in signal processing applications related to data communications systems, no research has been carried out to the investigation of the applicability of CMA for image restoration in practice. In filling this gap and taking into account the differences of signal processing in im- age processing and data communications contexts, we extend our research on the applicability of CMA deconvolution under the assumptions on the ground truth image properties. Through analyzing the main assumptions of ground truth image properties being zero-mean, independent and uniformly distributed, which char- acterize the convergence of CMA deconvolution, we develop a novel technique to overcome the effects of image source correlation based on segmentation and higher order moments of the source. Multichannel image restoration techniques recently gained much attention over the single channel image restoration due to the benefits of diversity and redundancy of the information between the channels. Exploiting these benefits in real time applications is often restricted due to the unavailability of multiple copies of the same image. In order to overcome this limitation, as the last area of our research, we develop a novel multichannel blind restoration model with a single image, which eliminates the constraint of the necessity of multiple copies of the blurred image. We consider this as a major contribution which could be extended to wider areas of research integrated with multiple disciplines such as demosaicing.

Τα συστήματα πολλαπλών κεραιών στον πομπό και στο δέκτη (MIMO) αποτελούν βασικά μέτωπα ανάπτυξης των ασύρματων επικοινωνιών. Ωστόσο, η εφαρμογή της τεχνολογίας MIMO στα κινητά δίκτυα επικοινωνιών αντιμετωπίζει το πρακτικό πρόβλημα της ενσωμάτωσης πολλαπλών κεραιών σε μικρά κινητά τερματικά. Με σκοπό την αντιμετώπιση του εμποδίου αυτού, δημιουργήθηκε ένα άλλο σημαντικό μέτωπο έρευνας, αυτό των συνεργατικών επικοινωνιών. Στο πλαίσιο της παρούσας διδακτορικής διατριβής ασχοληθήκαμε με την ανάπτυξη και μελέτη αλγορίθμων επεξεργασίας σήματος για τα δύο παραπάνω συστήματα. Σχετικά με τα συστήματα MIMO η πρωτοποριακή έρευνα που πραγματοποιήθηκε στα Bell labs στα μέσα της δεκαετίας του ΄90, απέδειξε ότι η χρήση πολλαπλών κεραιών μπορεί να οδηγήσει σε σημαντική αύξηση της χωρητικότητας των ασύρματων συστημάτων βελτιώνοντας την αξιοπιστία της μετάδοσης. Προκειμένου να αξιοποιηθούν οι παραπάνω δυνατότητες απαιτείται η σχεδίαση σύνθετων δεκτών MIMO. Προς αυτήν την κατεύθυνση έχει στραφεί ένας μεγάλος αριθμός μεθόδων ισοστάθμισης του καναλιού και πιο συγκεκριμένα δεκτών ανατροφοδότησης αποφάσεων. Δεδομένου ότι σε ευρυζωνικά συστήματα επικοινωνιών το ασύρματο κανάλι είναι άγνωστο στο δέκτη και μεταβάλλεται χρονικά, στραφήκαμε προς τις προσαρμοστικές μεθόδους ισοστάθμισης. Στα πλαίσια της διαριβής αναζήτησαμε προσαρμοστικούς αλγόριθμους κατάλληλους για τη σχεδίαση προσαρμοστικών ισοσταθμιστών MIMO DFE με τα εξής χαρακτηριστικά: 1) να παρουσιάζουν απόδοση (ταχύτητα σύγκλισης) συγκρίσιμη με αυτή του RLS, 2) η υπολογιστική τους πολυπλοκότητα να είναι μικρότερη από αυτή του RLS και 3) να είναι αριθμητικά ευσταθείς. ΄Εχει αποδειχθεί ότι προσαρμοστικοί αλγόριθμοι που βασίζονται στη μέθοδο των συζυγών κλίσεων (conjugate gradient (CG)) πληρούν τις παραπάνω προϋποθέσεις. Αρχικά αναζητήσαμε τεχνικές που βασίζονται στη μέθοδο αυτή και χρησιμοποιούνται σε προβλήματα προσαρμοστικού φιλτραρίσματος και πιο ειδικά, στο πρόβλημα προσαρμοστικής ισοστάθμισης διαύλου στη περίπτωση SISO. Πιο συγκεκριμένα, υλοποιήσαμε έναν προσαρμοστικό αλγόριθμο στο πεδίο των συχνοτήτων που επεξεργάζεται τα δεδομένα κάθε φορά που λαμβάνεται ένα νέο εισερχόμενο πακέτο δεδομένων. Ο προτεινόμενος ισοσταθμιστής πετυχαίνει μια πολύ καλή απόδοση, ενώ οι υπολογιστικές του απαιτήσεις είναι πολύ χαμηλές. Στη συνέχεια αναπτύξαμε τρεις νέους αλγορίθμους προσαρμοστικής ισοστάθμισης συχνοτικά επιλεκτικών συστημάτων MIMO, που βασίζονται στη μέθοδο CG και στις προβολές Galerkin. Το πρόβλημα σχεδιασμού προσαρμοστικών MIMO DFE αντιμετωπίζεται ως ένα πρόβλημα επίλυσης γραμμικών εξισώσεων, με πολλαπλά δεξιά μέλη, που εξελίσσεται στο χρόνο. Επισημαίνουμε ότι τα σχήματα που προτείνουμε θα μπορούσαν να αποτελέσουν ένα γενικότερο πλαίσιο σχεδίασης προσαρμοστικών δεκτών για συχνοτικά επιλεκτικά συστήματα MIMO, με ιδιότητες σύγκλισης παρόμοιες με αυτές του RLS, έχοντας, ωστόσο, μικρότερες υπολογιστικές απαιτήσεις. Στα πλαίσια της παρούσας διδακτορικής διατριβής αναπτύξαμε τεχνικές εκτίμησης καναλιού για συνεργατικά δίκτυα με N αναμεταδότες που είτε ενισχύουν και αναμεταδίδουν ή αποκωδικοποιούν και αναμεταδίδουν το λαμβανόμενο σήμα. ΄Ολες οι τεχνικές εκτίμησης που προτείναμε υλοποιούνται εξ΄ ολοκλήρου στο πεδίο των συχνοτήτων. Αρχικά παρουσιάσαμε τεχνικές που βασίζονται στη μετάδοση πιλοτικών συμβόλων σε συγκεκριμένες συχνοτικές συνιστώσες. Στη συνέχεια αποδείξαμε ότι όλα τα κανάλια από την πηγή μέσω των αναμεταδοτών προς τον προορισμό μπορούν να εκτιμηθούν τυφλά εάν γνωρίζουμε τις φάσεις της απόκρισης συχνότητας του ασύρματου καναλιού μεταξύ πηγής και προορισμού.. Επιπρόσθετα, πραγματοποιήθηκε ϑεωρητική ανάλυση της απόδοσης των προτεινόμενων σχημάτων η οποία επαληθεύτηκε μέσω προσομοιώσεων σε υπολογιστή. Τέλος, αξιολογήσαμε πειραματικά διάφορα πρωτόκολλα συνεργατικής επικοινωνίας (AF, DF, SF) και τεχνικές κατανεμημένης χωροχρονικής επεξεργασίας DSTC για συνεργατικά δίκτυα σε μια πλατφόρμα υλοποίησης πραγματικού χρόνου που χρησιμοποιεί επεξεργαστές ψηφιακής επεξεργασίας σήματος. Διαπιστώσαμε ότι τα πειραματικά αποτελέσματα συμφωνούν πλήρως με τα θεωρητικά.
Systems employing multiple antennas at the transmitter and the receiver, known as MIMO (multiinput multioutput) systems, as well as space time coding techniques developed for such systems, are two of the main technologies employed for the evolution of wireless communications. However, the application of MIMO technology to mobile networks, often faces the practical implementation problem of having too many antennas on a small mobile terminal. In an attempt to overcome such a severe limitation, cooperative communication schemes have been proposed. This PhD dissertation, described our work on the design and analysis of signal processing algorithms for the two aforementioned systems, as is described in detail next. Concerning MIMO systems, the pioneering work performed at Bell Labs in the middle of the nineties, proved that the use of multiple antennas can lead to a significant increase in wireless systems capacity. To exploit this potential, sophisticated MIMO receivers should be designed. To this end, a large amount of channel equalizers and, more specifically, decision feedback equalizers has been proposed. Because these assumptions are difficult to meet in high rate single carrier systems, we have focused our attention on decision feedback equalizers. . Our main goal is to derive algorithms for updating the MIMO DFE filters with the following characteristics: 1) convergence properties similar to these of the RLS 2) more computationally efficient than RLS and 3) numerically stable. It is known that adaptive algorithms based on the CG (conjugate gradient) have the above characteristics We initially studied this method as an iterative method for solving linear equations and we pointed out the main differences with the steepest descent method, on which the LMS algorithm is based. An extended search of adaptive DFE algorithms, based on the CG method was carried out. More specifically, a new block adaptive CG algorithm was developed. In the resulting algorithm, one CG iteration per block update is executed. In order to reduce even more the complexity, the algorithm was implemented in the Frequency Domain. The proposed equalizer offers a good performance - complexity trade off. Three new adaptive equalization algorithms for wireless systems operating over frequency selective MIMO channels, based on the CG method and the Galerkin projection method, are proposed. The problem of MIMO decision feedback equalizer (DFE) design is formulated as a set of linear equations with multiple righthand sides (RHSs) evolving in time. These schemes provide a flexible framework in MIMO adaptive equalization design to implement schemes with convergence properties comparable to the RLS, but of lower computational cost. Furthermore, we worked on channel estimation for cooperative communication networks, where the nodes either simply amplify and forward the received signal, or they decode and transmit the signal (DF). We first propose efficient channel estimation techniques for relay networks with N relays. The new methods are implemented in the frequency domain (FD). Initially, training based techniques are presented, where the training pilots are multiplexed with the data in the frequency domain. It is then shown that all the channels in the network can be estimated blindly provided that we know the phases of the frequency response of the (Source → Destination) channel. Thus, by making use of a small number of pilots in only one link (the sourcetodestination link) we can estimate all the other channels (Source→Relay i→Destination) in the network. A theoretical performance study of the proposed algorithms is presented and closed form expressions for the mean squared channel estimation error are provided. The presented theoretical analysis is verified by extensive Monte Carlo simulations. The application of the derived schemes to the DF case, and the impact of erroneous detection to their performance are also studied. Finally, we investigated experimentally four cooperative relaying schemes: amplify and forward (AF), detect and forward (DF), cooperative maximum ratio combining (CMRC) and distributed spacetime coding (DSTC), and one novel selection relaying (SR) scheme on a realtime DSP based testbed. The experimental results are fairly close to the ones predicted by theory