Experiments were carried out to determine the characteristics of a reflection-type modulator. The modulator consisted of an E-plane T-junction that was terminated on the side-arm by a reflex klystron operating in the passive region. Switching was accomplished by applying a pulse to the repeller of the passive klystron. It was found that the switching characteristics were strongly dependent on the power level at the input to the modulator. Good switching characteristics were obtained at low power levels hut these deteriorated as the input power became comparable to that produced by the klystron when operating as an oscillator. For the particular klystron used, satisfactory switching was possible for input power levels approximately 5 to 10 dB less than the output of the switching klystron. It was also found that the modulator characteristics depended on the operating mode of the passive klystron. The desired operating mode was a compromise between power-handling capability and the switching rate. An attempt was made to explain the decrease in switching range at high input power levels by relating the observed results to an increase in the bunching parameter at the boundaries of the passive region. However, this did not fully account for the observed results. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

Two types of devices employing ferroelectric modulation of a semiconductor thin-film have been realized and studied. The first consists of a cadmium selenide film with electrodes deposited on a barium titanate substrate together with a switching electrode on the other side of the substrate. This gives a two-valued resistor; in effect, a nondestructive readout of the state of the ferroelectric crystal which is regarded as a storage element. The second device is a thin-film transistor (TFT) deposited on a barium titanate crystal. A fourth counterelectrode - on the other side of the crystal allows changing between two opposite polarization directions in the crystal, thus giving a TFT with two sets of characteristics, roughly equivalent to a two-valued built-in gate bias. The read-in, or switching time, of the device is substantially determined by the barium titanate crystal and can be in the microsecond range for high switching fields. Readout of the devices can be continuous or not, as desired. Characteristics of the TFT, which is considered equivalent to a two-gate device, are analyzed in terms of the gradual channel approximation. Experimental results of the two devices are presented and discussed in relation to the predicted behaviour. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

In a transistorized push-pull amplifier, third order intermodulation distortion was effectively reduced at VHF frequencies with an output power of two watt. The non-linear distortion of the amplifier is modeled using a power series. The resulting expression is used as the basis for choosing the push-pull configuration to reduce the third order intermodulation distortion. The amplifier was built and tested, and the experimental results compare favorably with the theoretical results. The level of the third order intermodulation distortion is found to be at least 30 dB below the interfering signal level.

This report summarizes the design of a 25kW peak power pulse modulator for an airborne medium--range weather radar. It is the purpose of the modulator to collect and store energy over a certain time period and to form this energy into a short, high-power pulse. The modulator is required to drive a coaxial magnetron with a pulse of 5kV at 5A. System considerations make a pulse width of 3.5us and repetition rate of 99Hz necessary. The pulse is generated in a line-type pulse circuit which utilizes an SCR as a switching device. It is shown that a solid-state modulator can use a commercial grade SCR for the pulse generation. Although currents of 100A are switched, the instantaneous power dissipation in the SCR is reduced significantly through the use of a saturable delay reactor. A pulse transformer is used to achieve maximum power transfer from the modulator to the magnetron. The pulse transformer is insulated with semi-rigid epoxy. Corona generation is avoided by limited the voltage gradients in the insulation to 80V/mil.

Mel-frequency cepstral coefficients (MFCCs) are widely adopted in speech recognition as well as speaker recognition applications. They are extracted to primarily characterize the spectral envelope of a quasi-stationary speech segment. It was shown that cepstral features are closely related to the linguistic content of speech. Besides the magnitude-based cepstral features, there are resources in speech, e.g, the phase and excitation source, are believed to contain useful properties for speaker discrimination. Moreover, in real situations, there are large variations exist between the development and application scenarios for a speaker recognition system. These include channel mismatch, recording apparatus mismatch, environmental variation, or even change of emotional/healthy state of speakers. As a consequence, the magnitude-based features are insufficient to provide satisfactory and robust speaker recognition accuracy. Therefore, the exploitation of complementary features with MFCCs may provide one solution to alleviate the deficiency, from a feature-based perspective. / Speaker recognition (SR) refers to the process of automatically determining or verifying the identity of a person based on his or her voice characteristics. In practical applications, a voice can be used as one of the modalities in a multimodal biometric system, or be the sole medium for identity authentication. The general area of speaker recognition encompasses two fundamental tasks: speaker identification and speaker verification. / Wang, Ning. / Adviser: Pak-Chung Ching. / Source: Dissertation Abstracts International, Volume: 73-06, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (leaves 177-193). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

by Kingsley, King-chi, Kwan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1994. / Includes bibliographical references (leaves 85-88). / Tables of Contents / Lists of Figures & Tables / Acknowledgments / Abstract / Chapter Chapter 1 --- Introduction / Chapter 1.1 --- Digital Communication System --- p.P. 1 / Chapter 1.2 --- Channel Coding --- p.P. 1 / Chapter 1.3 --- Convolution Encoder --- p.P. 4 / Chapter 1.4 --- Additive White Gaussian Noise (AWGN) Channel --- p.P. 7 / Chapter 1.5 --- Trellis Diagram --- p.P. 8 / Chapter 1.6 --- Error Event and Free Distance --- p.P. 8 / Chapter 1.7 --- Euclidean Distance --- p.P. 10 / Chapter 1.8 --- Organization of the Thesis --- p.P. 11 / Chapter Chapter 2 --- QAM and MTCM / Chapter 2.1 --- Introduction --- p.P. 13 / Chapter 2.2 --- M-ary Quadrature Amplitude Modulation (QAM) / Chapter 2.2.1 --- M-ary Digital Modulation --- p.P. 13 / Chapter 2.2.2 --- Quadrature Amplitude Modulation (QAM) --- p.P. 14 / Chapter 2.2.3 --- Probability of Bit Error of M-ary QAM --- p.P. 16 / Chapter 2.3 --- Trellis Coded Modulation (TCM) --- p.P. 17 / Chapter 2.4 --- Multiple Trellis Coded Modulation (MTCM) --- p.P. 19 / Chapter Chapter 3 --- Set Partitioning of Signal Sets / Chapter 3.1 --- Introduction --- p.P. 21 / Chapter 3.2 --- Traditional Set Partitioning Methods / Chapter 3.2.1 --- Ungerboeck's Set Partitioning Method --- p.P. 21 / Chapter 3.22 --- Set Partitioning by M.K. Simon and D. Divsalvar --- p.P. 23 / Chapter 3.3 --- The new Set Partitioning Method / Chapter 3.3.1 --- Nomenclature of the Signal Points in the Signal Constellations --- p.P. 24 / Chapter 3.3.2 --- Generation of the Signal Sets --- p.P. 26 / Chapter 3.3.3 --- Partitioning of the Signal Sets / Chapter 3.3.3.1 --- Input Constraints of the Partitioning Method --- p.P. 30 / Chapter 3.3.3.2 --- The Set Partitioning Method --- p.P. 30 / Chapter 3.3.4 --- Distance Properties of the Partitioned Signal Sets --- p.P. 36 / Chapter 3.3.5 --- The Selection Scheme --- p.P. 39 / Chapter 3.3.6 --- Assignment of Signal Subsets into Trellis --- p.P. 42 / Chapter Chapter 4 --- Performance Evaluation / Chapter 4.1 --- Introduction --- p.P. 46 / Chapter 4.2 --- Upper Bound of Error Probability / Chapter 4.2.1 --- Probability of Symbol Error --- p.P. 46 / Chapter 4.2.1.1 --- Upper Bound on Probability of Symbol Error --- p.P. 48 / Chapter 4.2.1.2 --- Computation of the Transfer Function --- p.P. 49 / Chapter 4.2.2 --- Probability of Bit Error --- p.P. 51 / Chapter 4.3 --- Computation of the Free Distance --- p.P. 53 / Chapter Chapter 5 --- Results Presentation and Discussions / Chapter 5.1 --- Introduction --- p.P. 58 / Chapter 5.2 --- Results Presentations / Chapter 5.2.1 --- Normalized Square Free Euclidean Distance --- p.P. 58 / Chapter 5.2.2 --- Error Probability --- p.P. 71 / Chapter 5.3 --- Discussions --- p.P. 77 / Chapter Chapter 6 --- Conclusions --- p.P. 83 / Bibliography --- p.P. 85 / Chapter Appendix A - --- Flowchart of the Program --- p.P. 89 / Chapter Appendix B - --- Tabulated Results of d2free --- p.P. 104

Estimating the carrier frequency from a modulated waveform is one of the most important functions of a coherent signal receiver. Good performance and low bit error rates are obtained by coherent demodulation. Therefore, exact knowledge of the received signal carrier frequency is critical for communication systems. Also due to the spectral crowding, a high probability of channel interference can be observed. Under moderate carrier frequency offsets, data-aided estimators have been developed which have a high accuracy of estimation. However, for high frequency offsets the frequency estimator does not have the information of data or timing. In this thesis we propose a parametric based carrier frequency estimation of GMSK, which has improved performance over ad-hoc methods (delay and multiply) and has high resolution capability. In this thesis three methods are implemented over GMSK data to improve the performance and their results compared with the standard delay and multiply method. Two of these methods are parametric based estimators and one is a fast frequency estimator. Parametric based estimators were chosen partly due to their high resolution capabilities and mainly for their proven performance. Parametric based estimators were seen to have high computational load, and hence an alternate fast frequency estimator was implemented. The tradeoffs involved with respect to computational load and performance were shown. The contributions of this thesis include the verification of the validity of applying a parametric based approach on GMSK data, and compare the performances of parametric methods and fast frequency estimator. It is showii that such an approach has a better performance compared to non-data aided ad-hoc delay. and multiply methods. A closed loop configuration of the open loop parametric methods is suggested in the end. / Graduation date: 2004 / Best scan available. Figures are light on the original.

Conducted Electromagnetic Interference (EMI) is a major cause of concern in switchmode power supplies (SMPS) which commonly use standard pulsewidth modulation (PWM). In this thesis, Sigma-Delta (����) modulation is proposed as an alternative switching technique to reduce conducted EMI from SMPS. The result of using ���� modulation is a spread in the spectrum of the conducted emissions so that large concentrations of power at discrete frequencies are avoided. Experimental time-domain waveforms and spectra of the switching function from first-order and second-order ���� modulators are presented to prove the viability of the scheme in EMI mitigation. These modulators are then applied to an off-the-shelf computer power supply and experimental results using FCC-specified methods show a reduction of roughly 5-10dB in EMI emissions over standard PWM modulators. / Graduation date: 1999

Thesis (M.S. in Electrical Engineering)--Naval Postgraduate School, December 2009. / Thesis Advisor(s): Jenn, David C. Second Reader: Pace, Phillip E. "December 2009." Description based on title screen as viewed on January 28, 2010. Author(s) subject terms: Direct conversion, homodyne, IQ mismatch, DC offsets, IQ circle, phase error, modulator, demodulator calibration, LabVIEW, Directional Finding (DF), digital beamforming, Robust Symmetrical Number System (RSNS). Includes bibliographical references (p. 73-75). Also available in print.

Thesis (Ph. D.)--Brigham Young University. Dept. of Electrical and Computer Engineering, 2004. / Includes bibliographical references (p. 169-181).