The Orthogonal frequency division multiplexing (OFDM) technique has been intensively used in many wireless communication systems to achieve higher data rate transmissions. Due to the fact that the OFDM technique entails redundant block transmissions; the transmitted blocks suffer from the inter-symbol interference (ISI) and inter-block interference (IBI). To compensate this serious effect, in many literatures redundant symbols (or guard interval) with adequate length are inserted in the transmitted symbols to prevent the IBI. Also, in the receiver the equalizer can be employed to deal with ISI. In this thesis, we present a new pseudo random cyclic-postfix (PRCP-) OFDM associated with the multiple-input multiple-output (MIMO) antenna system configuration to further improve the system performance. In fact, the MIMO system can enhance channel capacity and achieve high data-rate. The above-mentioned PRCP-OFDM technique combines with the MIMO antennas system, through the appropriate model design can be used to combat the multi-path effect or the inter-block interference. As evident from the simulation results, the proposed ST-BC MIMO PRCP-OFDM system can avoid the interference of transmitted signals during the estimation of channel impulse response (CIR) with proposed cyclic-postfix sequences. In addition, to further improve and eliminate the residual IBI and ICI, the equalizer with the framework of the generalized sidelobe canceller (GSC) is considered. Specifically, when SNR grows, the proposed ST-BC MIMO PRCP-OFDM system can perform successfully in terms of symbol-error rate and semi-blind channel estimation. This is verified via the computer simulations.

The main topic of this research is the simulation of internal wave interaction by a two-dimensional numerical model developed by Lynett & Liu (2002) of Cornell University, then modified by Cheng et al. (2005). The governing equation includes two-dimensional momentum and continuity equation. The model uses constant upper and lower layer densities; hence, these factors as well as the upper layer thickness. Should be determined before the simulation. This study discusses the interface depth and the density according to the buoyancy frequency distribution, the EOF, and the eigen-value based on the measured density profile. Besides, a method based on the two-layer KdV equation and the KdV of continuously-stratified fluid. By minimize the difference of linear celeriy, nonlinear and dispersion terms, the upper layer thicknes can also be determined. However, the interface will be much deeper than the depth of max temperature drop in the KdV method if the total water depth is bigger than 500 meters. Thus, the idealization buoyancy frequency formula proposed by Vlasenko et al. (2005) or Xie et al. (2010) are used to modify the buoyancy frequency. The internal wave in the Luzon Strait and the South China Sea are famous and deserves detailed study. We use the KdV method to find the parameters in the two fluid model to speed up the simulation of internal wave phenomena found in the satellite image.

In this thesis, we study the noise power estimation in the LTE system. Two approaches, the weight method and the subspace method, are considered. The performance of noise power estimation using the weight method highly depends on the accuracy of the channel estimation. The channel estimation usually gets poor results under scenarios with long delay spreads. Therefore, the weight method also gets poor result. To overcome the mentioned drawback, we propose the subspace method which is independent from the channel estimation. From simulation results, we observe that the subspace method gets bias results. However, the bias depends on the length of the observation window and only gets a little influence from the channel conditions. Therefore, we can correct the bias using a simple look-up-table approach. Computer simulations show that the subspace method gets the more accurate result than the weight method.

In multi-cell downlink OFDMA radio network system, users in one cell would suffer from the inter-cell interference caused by frequency reuse, namely co-channel interference. For a practical system, the inter-cell interference seriously decreases the quality of communication, especially for cell-edge users. Therefore, some interference management techniques, such as resources allocation, beamforming¡Ketc., will become an important issue in this system. Therefore, how to allocate resources to enhance cell-edge user performance and total system throughput is the major problem of our research. In this thesis, for management the inter-cell interference in multi-cell downlink OFDMA radio network system, a power allocation method based on the Cross-Entropy algorithm is proposed to find the sub-optimal solution and corresponding subcarriers allocation. In the system, it is considered that a sum-rate maximization problem while satisfying the target rate of both cell-edge users and cell-interior users. The simulation results show that the proposed method can effectively reduce interference between cells, and increases the transmission performance of cell-edge users and overall system throughput.

In orthogonal frequency division multiplexing (OFDM) system, blind estimator was proposed which can obtain high bandwidth efficiently. There is a serious ambiguity problem in blind data detector structure. Solution methods can divide into three cases: pilot signal, superimposed training, and channel coding. In order to achieve totally blind estimate, we use channel coding to solve ambiguity in this thesis. In previous study, it had been use low-density-parity-check code (LDPC) to solve ambiguity, and proposed an encoding method to avoid ambiguity for BPSK. However, we consider generic linear block code (LBC) and want to extend BPSK modulation to higher modulation scheme, including QPSK, 16QAM, and 64QAM. For any constellation follows grey coding, we induct a difference of inner product for ambiguity and derive some sufficient conditions for LBC. If LBC satisfy some conditions, then it could avoid ambiguity between valid code words and it can achieve totally blind estimate. In simulation section, for data estimate, we respectively use two LBC cases, which exist ambiguity or not. In order to be fair, we insert a pilot to solve ambiguity in LBC, which exist ambiguity. In simulation results, the performance of two cases is similar in high signal to noise ratio (SNR). In other words, if we use proper channel code which it satisfy sufficient conditions, then we can increase bandwidth efficiently.

It is well known that the Preisach model is useful to approximate the effect of hysteresis behavior in smart materials, such as piezoactuators and Shape Memory Alloy(SMA) wire actuators. For tracking control, many researchers estimate a Preisach model and then compute its inverse model for hysteresis compensation. However, the inverse of its hysteresis behavior also shows hysteresis behavior. From this idea, the inverse model with Kransnoselskii-Pokrovskii(KP) model, a developed version of Preisach model, can be used directly for SMA position control and avoid the inverse operation. Also, we propose another method for the tracking control by approximating the inverse model using an orthogonal polynomial network. To estimate and update the weight parameters in both inverse models, a gradient-based learning algorithm is used. Finally, for the SMA position control, PID controller, adaptive controllers with KP model and adaptive nonlinear inverse model controller are compared experimentally.

In this study we develop a new method for testing the null hypothesis that the predictor function in a canonical link regression model has a prescribed linear form. The class of models, which we will refer to as canonical link regression models, constitutes arguably the most important subclass of generalized linear models and includes several of the most popular generalized linear models. In addition to the primary contribution of this study, we will revisit several other tests in the existing literature. The common feature among the proposed test, as well as the existing tests, is that they are all based on orthogonal series estimators and used to detect departures from a null model. Our proposal for a new lack-of-fit test is inspired by the recent contribution of Hart and is based on a Laplace approximation to the posterior probability of the null hypothesis. Despite having a Bayesian construction, the resulting statistic is implemented in a frequentist fashion. The formulation of the statistic is based on characterizing departures from the predictor function in terms of Fourier coefficients, and subsequent testing that all of these coefficients are 0. The resulting test statistic can be characterized as a weighted sum of exponentiated squared Fourier coefficient estimators, whereas the weights depend on user-specified prior probabilities. The prior probabilities provide the investigator the flexibility to examine specific departures from the prescribed model. Alternatively, the use of noninformative priors produces a new omnibus lack-of-fit statistic. We present a thorough numerical study of the proposed test and the various existing orthogonal series-based tests in the context of the logistic regression model. Simulation studies demonstrate that the test statistics under consideration possess desirable power properties against alternatives that have been identified in the existing literature as being important.

In recent years, the development of OFDM system has received a lot of attention. Some examples of existing systems where OFDM system is used are digital audio broadcasting, high-definition television terrestrial broadcasting, asymmetric digital subcarrier lines and so on. There are several reasons for using OFDM systems. First, OFDM system is an efficient way to deal with multipath effect. Under a fixed amount of delay spread, the implementation complexity of OFDM system is much less than that of single-carrier system. The reason is that OFDM system can simply use guard time to process delay spread without a complex equalizer. Second, OFDM system can achieve high data rate to transmit by using large number of subcarriers. Third, OFDM system can also efficiently combat with narrow band interference. On the other hand, OFDM system also has two main drawbacks. One is more sensitive to frequency offset, the other is higher PAPR. This thesis focuses on the PAPR problem. Pulse shaping method is an effective way to solve this problem. It can be used for any number of subcarriers of OFDM systems, so it is very flexible. It doesn¡¦t have any additional IFFTs in comparison to the selected mapping or partial transmit sequence method. Its implementation is simpler. And because it also doesn¡¦t distort the OFDM symbols, its bit error performance should be better than the clipping method. According to the pulse shaping method, we get a better waveform that can make the PAPR of OFDM symbols do not exceed about 2.

The selection and learning of kernel functions is a very important but rarely studied problem in the field of support vector learning. However, the kernel function of a support vector regression has great influence on its performance. The kernel function projects the dataset from the original data space into the feature space, and therefore the problems which can not be done in low dimensions could be done in a higher dimension through the transform of the kernel function. In this paper, there are two main contributions. Firstly, we introduce the gradient descent method to the learning of kernel functions. Using the gradient descent method, we can conduct learning rules of the parameters which indicate the shape and distribution of the kernel functions. Therefore, we can obtain better kernel functions by training their parameters with respect to the risk minimization principle. Secondly, In order to reduce the number of support vectors, we use the orthogonal least squares method. By choosing the representative support vectors, we may remove the less important support vectors in the support vector regression model. The experimental results have shown that our approach can derive better kernel functions than others and has better generalization ability. Also, the number of support vectors can be effectively reduced.

All digital communication channels are subject to inter-symbol interference (ISI). To achieve the desired system performance, at receiver end, the effect of ISI must be compensated and the task of the equalizer is to combat the degrading effects of ISI on the transmission. Due to the demand of high data transmission rate, the multicarrier modulation (MCM) technique implemented with the orthogonal frequency division multiplexing (OFDM) has been adopted in many modern communications systems for block transmission. In block transmission systems, transmitter-included redundancy using finite-impulse response (FIR) filterbanks can be utilized to suppress inter-block-interference (IBI). However, the length of redundancy will affect the system performance, which is highly dependent on the length of channel impulse response. To deal with the effect of ISI, many equalizing schemes have been proposed, among them the FIR zero-forcing (ZF) equalizer with the non-orthogonal projector provides a useful transceiver design structure for suppressing the IBI and ISI, simultaneously. In this thesis, we propose a new equalizing scheme; it combines the FIR-ZF equalizer with non-orthogonal projector as well as the Bayesian decision feedback equalizer (DFE) for IBI and ISI suppression. The Bayesian DFE is known to be one of the best schemes to achieve the desired performance for eliminating ISI. It can be employed to achieve the full potential of symbol-by-symbol equalizer. That is, after removing the effect of IBI with the non-orthogonal projector, the Bayesian DFE is employed for eliminating the ISI, simultaneously. For comparison, the system performance, in term of bit error rate (BER) is investigated, and compared with the minimum mean square error (MMSE)-IBI-DFE. The advantage of the new proposed equalizing scheme is verified via computer simulation under condition of insufficient redundancy.