Search Results (1 - 5 of 5 Results)

Sort By  
Sort Dir
 
Results per page  

Imbulgoda Liyangahawatte, Gihan Janith MendisHardware Implementation and Applications of Deep Belief Networks
Master of Science in Engineering, University of Akron, 2016, Electrical Engineering
Deep learning is a subset of machine learning that contributes widely to the contemporary success of artificial intelligence. The essential idea of deep learning is to process complex data by abstracting hierarchical features via deep neural network structure. As one type of deep learning technique, deep belief network (DBN) has been widely used in various application fields. This thesis proposes an approximation based hardware realization of DBNs that requires low hardware complexity. This thesis also explores a set of novel applications of the DBN-based classifier that will benefit from a fast implementation of DBN. In my work, I have explored the application of DBN in the fields of automatic modulation classification method for cognitive radio, Doppler radar sensor for detection and classification of micro unmanned aerial systems, cyber security applications to detect false data injection (FDI) attacks and localize flooding attacks, and applications in social networking for prediction of link properties. The work in this thesis paves the way for further investigation and realization of deep learning techniques to address critical issues in various novel application fields.

Committee:

Jin Wei (Advisor); Arjuna Madanayaka (Committee Co-Chair); Subramaniya Hariharan (Committee Member)

Subjects:

Artificial Intelligence; Computer Engineering; Electrical Engineering; Engineering; Experiments; Information Technology

Keywords:

deep belief networks; multiplierless digital architecture; Xilinx FPGA implementations; low-complexity; applications of deep belief networks; spectral correlation function; modulation classification; drone detection; doppler radar; cyber security

Chae, Chun SikStudies of the Interferometric Phase and Doppler Spectra of Sea Surface Backscattering Using Numerically Simulated Low Grazing Angle Backscatter Data
Doctor of Philosophy, The Ohio State University, 2012, Electrical and Computer Engineering

Range-resolved interferometric phase and Doppler spectra are two subjects of interest with regard to the retrieval of sea surface height profiles from coherent marine radar measurements. The studies of this dissertation attempt to improve understanding of the properties and associated measurement errors of these quantities through the use of numerically simulated low-grazing-angle backscatter data.

In the first part of the dissertation, studies of the interferometric phase are described. Backscattered fields computed using the method of moments for one dimensional ocean-like surface profiles are used to examine statistical properties of the single-look interferometric phase estimator, in order to investigate the applicability of standard expectations for height retrieval accuracy in this problem. The results show that shadowing and multipath propagation effects cause errors in interferometric phase estimation beyond those caused by speckle effects alone. In addition, the decorrelation between the fields received at two antennas is found to be impacted by shadowing and multipath propagation effects, making standard models for this quantity less applicable as well. These results show that modeling the expected performance of interferometric sea surface height retrieval approaches at low grazing angles is difficult.

The second part of the dissertation involves studies of the range-resolved Doppler spectra at low-grazing-angles. Backscattered fields are computed for a single realization of a one-dimensional ocean-like surface profile as the realization evolves in time. Transformation into the range-Doppler domain enables examination of properties of the resulting Doppler spectra (for both HH and VV polarizations) and their relationship to properties of the surface profile. In general, a strong correspondence between the long wave orbital velocity of the surface and the Doppler centroid frequency is observed for visible portions of the surface, as well as some evidence of relationships between the width of the Doppler spectrum and variations of the orbital velocity in time at a given range point. Evidence of similar relationships even in some shadowed portions of the surface is also provided. Doppler spectra from HH and VV polarizations are qualitatively similar in most respects, although the portion of shadowed surface points from which Doppler information is available is somewhat larger in VV polarization. A further examination is conducted using backscattered fields computed with a "single scattering" method that neglects shadowing and any multiple scattering effects. The remarkable similarities observed in Doppler spectra for the complete and single scattering models even in some shadowed portions of the surface suggests that non-line-of-sight propagation effects do not significantly in fluence Doppler properties in such regions.

The studies in this dissertation provide improved understandings of range-resolved interferometric phase and Doppler spectra at low grazing angles. These results provide new information for the design of coherent marine radars for the retrieval of sea surface profiles.

Committee:

Joel Johnson (Advisor); Robert Burkholder (Committee Member); Fernando Teixeira (Committee Member)

Subjects:

Electrical Engineering

Keywords:

electromagnetic scattering; surface scattering; interferometric phase; sea doppler spectra; retrieval of sea surface height; microwave remote sensing; method of moments; interferometry; doppler radar; low grazing angle; ocean remote sensing

Randeny, Tharindu DMulti-Dimensional Digital Signal Processing in Radar Signature Extraction
Master of Science in Engineering, University of Akron, 2015, Electrical Engineering
Aperture-arrays combined with multi-dimensional (MD) digital signal processing (DSP) techniques provide the capability of synthesizing arbitrarily steered beams enabling directional enhancement of radio frequency (RF) signals. This is particularly desirable in radar applications, where electronically-scanned beams are needed to obtain measurements over range, angle/direction, polarization and doppler domains. Digital aperture arrays employing two-dimensional (2-D) infinite impulse response (IIR) beam filters combined with radar systems can lead to rapidly steerable RF beams with modulated radar signatures which carry essential information for critical real time electromagnetic (EM) sensing requirements. This thesis introduces novel MD DSP algorithms for radar signal extraction that find applications in emerging RF technologies. Directional sensing and remote localization architectures comprising of 2-D IIR digital filters , continuous wave (CW) doppler radars and frequency modulated continuous wave (FMCW) radars are proposed along with simulation results. Detection and classification of miniaturized unmanned aerial systems (UAS) using doppler radar signatures is presented with experimental results. Radar signal extraction algorithms in applications of 3-D body tracking and remote vital sign detection is also reviewed through a system study on recent advances of radar systems.

Committee:

Arjuna Madanayake, Dr. (Advisor); Ryan Toonen, Dr. (Committee Member); Jin Kocsis, Dr. (Committee Member)

Subjects:

Computer Engineering; Electrical Engineering

Keywords:

Digital Signal Processing, Radar Systems, Localization, 2-D IIR SBP, FMCW, Doppler Radar

Yang, WuTraffic Surveillance Using Low Cost Continuous Wave (CW) Doppler Radars
Doctor of Philosophy (PhD), Wright State University, 2012, Engineering PhD
Low cost un-modulated continuous wave (CW) radar (CW Doppler radar) can be used to measure the speed of a vehicle. Traditionally, a radar gun, a lidar gun or a speed camera is used to capture a speeding vehicle. A radar gun can either measure the fastest vehicle or the vehicle with the strongest reflection. If a radar gun is used, a police officer must determine which vehicle has the speed shown on the screen of the radar gun. A lidar gun can precisely detect a speeding vehicle, but it requires precise aiming. When a camera is used, a picture will be taken at a fixed location. For the first case, human error is unavoidable, for the second case, the aiming requirement makes it unsuitable for automated surveillance, and in the third case, the surveillance region is very limited. In order to solve these problems, we have invented an automatic traffic surveillance system (ATSS) using two CW Doppler radars (forward radar and side radar) and a video camera. An algorithm to balance in-phase and quadrature channel of directional CW Doppler radar based on spectrogram has been developed and tested on real highway data. A detailed architecture for Doppler speed tracking has been designed. Doppler speed tracks are initialized and extended by the side radar and further extended by the forward radar. Three algorithms have been developed for Doppler speed tracking. All algorithms have been tested on real highway data and simulated data. The results show that all three algorithms can successfully extract the Doppler speed tracks from CW radar signals.

Committee:

Lang Hong, PhD (Committee Chair); Kefu Xue, PhD (Committee Member); Zhiqiang Wu, PhD (Committee Member); Xiaodong Zhang, PhD (Committee Member); Raj Bhatnagar, PhD (Committee Member); Ramana Grandhi, PhD (Other); Andrew Hsu, PhD (Other)

Subjects:

Electrical Engineering

Keywords:

CW Doppler Radar; Traffic Surveillance; Kalman Filter; Spectrogram

Roy, AruneshFusion of Video and Doppler Radar for Traffic Surveillance
Doctor of Philosophy (PhD), Wright State University, 2010, Engineering PhD

Current Continuous Wave (CW) Doppler radar speed measurement systems lack the ability to distinguish multiple targets. Most systems can only identify the strongest (closest) target or the fastest target.

This dissertation is related to a fusion algorithm for a VIdeo-Doppler-radAR (Vidar) traffic surveillance system. The Vidar systems uses a robust matching algorithm which iteratively matches the information from a video camera and multiple Doppler radars corresponding to the same moving vehicle, and a stochastic algorithm which fuses the matched information from the video camera and Doppler radars to derive the vehicle velocity and angle information.

We use two heterogeneous sensors of very different modalities, the first a high resolution (1024x768 pixels) video camera operating at 30 Hz with a 1/3″ sony CCD fitted with a narrow field-of-view lens and the other a CW Doppler radar operating in the unlicensed Ka band (35 GHz) with a maximum detection range of 3000 ft. First, a high resolution Time-Frequency representation of the radar signal is obtained by employing the method of Time-Frequency reassignment. Then, the angle information obtained from the video camera is fused with the information from the Doppler radar to produce a velocity and angle track of the targets within the surveillance region.

Committee:

Lang Hong, PhD (Advisor); Kefu Xue, PhD (Committee Member); Fred Garber, PhD (Committee Member); Arthur Goshtasby, PhD (Committee Member); Michael Temple, PhD (Committee Member)

Subjects:

Electrical Engineering

Keywords:

traffic surveillance; fusion; video; doppler radar