<p> Memory and learning in the brain are realized by a collection of mechanisms that interact, resulting in learning and subsequent recognition of input patterns. While these mechanisms are complex and span different areas of the brain, we hypothesize that we can construct electronic circuits that mimic these mechanisms, demonstrating some aspects of pattern recognition that mimic the brains ability to learn and recognize patterns.</p><p>

<p> Precise frequency and phase estimation are important aspects of a telecommunication system. They are the important sub-system of communication, navigation, radar, and various other engineering systems. Systems and methods provide fast and precise frequency estimation by keeping minimal sampling and a high tolerance to noise. Methods used to estimate the frequency include multiple phase shift keying and multiple quadrature amplitude modulation scheme. This project utilizes a frequency estimation technique to keep sampling minimal and to have a high endurance to noise for better performance. Recently, fast and precise frequency estimation have become increasingly important in the evolution of the Orthogonal Frequency Division Multiplexing. Previous techniques of frequency estimation are based on the Extended Kalman Filter and Fast Fourier Transform. This project has developed a fast and precise carrier frequency estimation scheme. The Kalman Filter approach, Classen approach, and Moose approach are used for the carrier frequency estimation.</p>

<p> In Optical Wireless Communication (OWC) systems, the technique of Flip-Orthogonal Frequency Division Multiplexing (Flip-OFDM) is applied, where the positive segment and negative segment of the signal are transmitted on two continuous OFDM sub-frames. Conventionally in Flip-OFDM, data is recovered by simply subtracting the negative sub-frame from the positive sub-frame. But the signal survey confirms that the information in both sub-frames gets disrupted during signal propagation and an incorrect detection of the received values takes place at the detector. </p><p> An iterative receiver is proposed that utilizes the input signal values along with the received signals sub-frames to improve the efficacy of the Flip-OFDM. The Matlab software is used to design the transmitter and iterative receiver. Features such as cyclic prefixing, frame multiplexing, and channel characterization are included in the design. </p><p> The results of the simulation depict that the suggested receiver delivers a superior Bit Error Rate (BER) and Signal to Noise Ratio (SNR) gain when compared to the conservative receiver.</p>

This thesis studies the feasibility of developing a smart shipboard sensor network. The objective of the thesis is to prove that sensors can be made smart by keeping calibration constants and other relevant data such as network information stored on the sensor and a server computer. Study will focus on the design and implementation of an Ipsil IP(micro)8930 microcontroller, which is then connected, by the standard TCP/IP implementation, to a network where the sensor information can be seen using a web page. The information to make the sensor "smart" will be stored on the Ipsil chip and server computer and can by accessed by a HTML based program. By taking pre-computed calibration constants that minimize the measurement errors and writing them through the web page stored in the Ipsil chip's EEPROM, the calibrated sensor reading can be calculated. The expected contribution from the research effort would be a reduction in manpower, increased efficiency, and a greater awareness of plant and equipment operation among naval vessels, specifically the DDX. Hardware is relatively inexpensive, reliable, and COTS (Commercial Off the Shelf) available. If implemented, a Smart Shipboard Sensor Network would allow the watch standers, CHENG, OOD, and CO, to all see the same information about the ship.s engineering plant and equipment. A prototype sensor test bed was constructed in the laboratory, which consists of an Ipsil IP(micro)8930 microcontroller, a Linksys LAN router, and a Dell Inspiron 9300 laptop. The newly developed smart sensor was successfully demonstrated.

The realization of today's telecommunication networks is a challenging task. Network architectures are constantly changing to meet new requirements for many new and exciting services and applications. As a result of these added new requirements, new types and mixes of traffic profiles are being introduced into these networks. To facilitate these needs, there are many tools which have been developed to aid in the planning, development, optimization and traffic prediction process. One such tool is PLANITU 3.0. PLANITU is designed to handle many types of circuit network systems and features a powerful graphics capability. The software uses wellestablished, iterative prediction concepts, such as Erlang-B loss equation and the Wilkinson ERT method. Two types of network systems were modeled using real data supplied by Siemens Indonesia. Target networks for study included a fixed switched networks and a GSM (Global System for Mobile Communications) network. PLANITU 3.0 performed well for the fixed switched network systems demonstrating reasonable results within an acceptable degree of accuracy, but performed poorly for GSM systems yielding inoperable simulation features, numerous bugs and software instability.

Orthogonal Frequency Division Multiplexing (OFDM) is fast becoming the signal modulation technique of choice for many commercial and military wireless applications. Its resilience to cochannel interference and bandwidth efficiency make it ideal for many different applications. With its increasing popularity among disparate facets of society, it becomes likelier that enemy militaries and/or nonmilitary combatants will utilize the technique or a system that uses the technique. In light of this development, the need to develop techniques and algorithms to enable detection becomes apparent. This thesis will attempt to develop a model for OFDM and measure its performance in a multipath, outdoor environment with low signal-to-noise ratio, high noise and cochannel interference. Because of the unpredictability of the outdoor environment and the proliferation of various OFDM standards, the simulation will utilize only one algorithm for modeling outdoor environments and the IEEE 802.11a standard.

The ability to sense and monitor a variety of environmental conditions using un-tethered sensors offers a significant change over traditional sensing systems that need to be strategically positioned and have topologies engineered. As such, recent research into wireless sensor networks has attracted great interest due to its diversity of applications, ranging in areas such as home, health, environmental and military applications. In this thesis, the evaluation of a routing protocol developed by Crossbow Technologies called XMesh, is presented. The main components of the routing protocol are described and the routing algorithm explained. Experiments were conducted to determine the connectivity ranges of motes in different transmission power settings. The relationship of mote transmission power and network connectivity is presented. An energy efficiency study looked at the means of extending the lifespan of the network. Although, packet losses during the period of a node failure were significant, the routing protocol showed that it was able to adapt and reorganize to provide reliable and stable routing in a network.

In this thesis a method is developed to estimate and remove group delay caused by the transmission channel in the case of a burst signal. The signal of interest is modeled as burst white noise, and no information about any underlying symbol structure or rate is assumed. The regions of transition between the signal present and the signal not present are used to estimate the group delay of the transmission channel. Once estimated, the group delay can be removed using an all-pass lter with the proper phase characteristics.

This thesis investigates the fundamentals of multiple-input single-output (MISO) and multiple-input multiple-output (MIMO) radio communication systems with space-time codes. A MISO system and MIMO systems were designed using multicarrier delay diversity modulation (MDDM). MDDM was incorporated with orthogonal frequency division multiplexing (OFDM). The design was implemented with binary phase shift keying (BPSK). Matlab was used to simulate the design, which was tested in both an additive white Gaussian noise (AWGN) channel and in a slow fading frequency nonselective multipath channel with AWGN. The receiver design was incorporated with the maximal ratio combiner (MRC) receiving technique with perfect knowledge of channel state information (CSI). The theoretical performance was derived for both channels and was compared with the simulated results.

Integrated Power Systems will be used on future naval combatants. These systems will allow unprecedented control of shipboard power to propel the ship, sense the battle-space, and engage the enemy. One crucial enabling technology is robust power conversion modules like the hybrid dc to ac inverter. This thesis is a further exploration of the hybrid inverter scheme consisting of a six-step voltage-source inverter (VSI) and a hysteresis controlled current-source inverter (CSI). The six-step controller was redesigned to make it independent of the hysteresis controller. The hysteresis controller is fed a reference signal extracted from the total output current. The signal is filtered and modified by the closed-loop system such that the total output current approaches a perfect sine wave limited only by bandwidth. The modified closed-loop controller was compared to a previous Naval Postgraduate School effort and found to improve current total harmonic distortion from 3.2% to 1.8%. This thesis proves that existing power electronic technology can be used to produce high-fidelity waveforms for high-power Naval Propulsion Drives (50-100 MW).