-
[show abstract]
[hide abstract]
ABSTRACT: This paper describes the synthesis of linear antenna arrays using the cross entropy (CE) method to produce array responses with minimum peak sidelobe levels (SLL) and position nulls at specific angular locations. The CE method is a very simple, yet highly efficient approach to solving complex multi-objective and multi-extremal optimization problems. CE seeks to adaptively estimate an optimal sampling distribution which generates solutions in the neighborhood of the globally optimal solution by minimizing the cross entropy (or Kullback-Leibler divergence) between current best estimates and the overall estimate of the optimal distribution. The performance of CE is comparable to other popular optimization techniques such as the genetic algorithm, simulated annealing and particle swarm optimization.
Industrial Electronics, 2008. IECON 2008. 34th Annual Conference of IEEE; 12/2008
-
[show abstract]
[hide abstract]
ABSTRACT: Resonant clocks are new design techniques for multi-gigahertz clock distributions that are gaining prominence in the design of low power and ultrahigh frequency microprocessors. In the radiofrequency range, new challenges with respect to skews and jitters become greatly pronounced rendering many conventional clocking techniques inadequate. In this work, we present a comparative study of the power dissipation of three resonant clocking techniques: standing wave, rotary wave and dasiaresonant-loadpsila global clock distributions. Specifically, we generated non-overlapping clock signals resonantly to drive transmission gates in the design of our new binary counter. We used a simplified Sunpsilas SwaP Metric to determine the power efficiency of each resonant technique. All of our designs were simulated using Agilent ADS 2006A. Furthermore, our analysis has revealed that the rotary clock design can achieve a power efficiency of a magnitude of two compared to the other resonant clock techniques.
Computer and Communication Engineering, 2008. ICCCE 2008. International Conference on; 06/2008
-
[show abstract]
[hide abstract]
ABSTRACT: Stock forecasting is a major component of any finance institution because predictions of future prices, indices, volumes and many more values are often incorporated into the economic decision-making process. Although there are many different approaches out there, this paper will compare unsupervised classification techniques such as k-means clustering with supervised learning algorithms such as support vector machines (SVMs). In our study, a list of stock prices taken from historical data of the S&P 500 is used as our testbed. These prices will be categorized as increasing or decreasing in price on a weekly basis. The goal of this study is to determine the best method for forecasting the trend of stock prices.
System Theory, 2008. SSST 2008. 40th Southeastern Symposium on; 04/2008
-
[show abstract]
[hide abstract]
ABSTRACT: This paper presents a comprehensive approach to accurate and efficient modeling of microwave active devices such as metal semiconductor field effect transistors (MESFETs) using artificial neural networks (ANNs). A radial basis function (RBF)-ANN model is developed for S-parameters and equivalent circuit parameters (ECPs) of MESFETs. The training and testing data for these models are obtained from the measured two-port scattering parameters and extracted ECPs of a 0.25 times 200 mum (4 times 50 mum) gallium arsenide MESFET. A four- input eight-output ANN is used to model the S-parameters of a microwave MESFET versus bias, temperature, and frequency, and a three-input eight-output ANN is used to model the ECPs of a microwave MESFET versus bias and temperature. Comparisons of measured and modeled data are presented, and the results show very good agreement. The average relative errors using the RBF-ANN models for the S-parameters and ECPs were 0.81% and 0.77%, respectively, which both represent about 60% reduction in error when compared to backpropagation ANN models of similar parameters of the same device.
IEEE Transactions on Instrumentation and Measurement 11/2007; · 1.21 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Kernel methods are a mathematical tool that provides a generally higher dimensional representation of given data set in feature space for feature recognition and image analysis problems. Typically, the kernel trick is thought of as a method for converting a linear classification learning algorithm into non-linear one, by mapping the original observations into a higher-dimensional non-linear space so that linear classification in the new space is equivalent to non-linear classification in the original space. Moreover, optimal kernels can be designed to capture the natural variation present in the data. In this paper we present the performance results of fifteen novel kernel functions and their respective performance for kernel principal component analysis on five select databases. Empirical results show that our kernels perform as well and better than existing kernels on these databases.
Computational Intelligence in Robotics and Automation, 2007. CIRA 2007. International Symposium on; 07/2007
-
[show abstract]
[hide abstract]
ABSTRACT: Summary form only given. This study presents a performance comparison analysis of Fourier transform (FT), discrete cosine transform (DCT), wavelet transform (WT), and wavelet packets (WP) on a 1024×1024, 12-bit mammogram and a 512×256, 8-bit, fingerprint image. In the multi-resolution analysis methods, three to five level decompositions and different entropy models at any decomposition levels will be used. An adaptable signal decomposition algorithm for minimizing the decomposition tree will also be introduced. The images are first segmented into two regions: region of interest (ex. micro-calcification in the mammograms), and the background region. The two regions are then compressed at two different levels, to better preserve the information in the image, but most importantly in the region of interest. The quality of the resultant compressed images is subjected to visual analysis by a group of 30 non-experts students, as well as analyzed objectively based on the peak signal-to-noise ratio (PSNR), mean square error (MSE), and reconstruction error. This study could potentially help radiologists and fingerprint experts better detect the important details in the images. Furthermore, the results will save storage space, reduce access time, and improve the accuracy of diagnosis - in other words, cost savings. The compressed images are also better suited for remote access and transfer, for tele-diagnostic, and tele-medicine research and training.
Systems Engineering, 2005. ICSEng 2005. 18th International Conference on; 09/2005
-
[show abstract]
[hide abstract]
ABSTRACT: In the age of the Internet, computer intrusion detections are at best a black art. The science to perceive, track, and understand intrusions is still at its infancy. Many different technologies and topologies are under investigation to see which model provides adequate data for intrusion detection. In this paper, we present a platform independent Java-based mobile agent intrusion detection system (IDS). In our prototype mobile agent IDS, the focus is on port scanning and file integrity checking. The mobile agents are implemented using the Concordia mobile agent development kit. The performance and memory resources required to run these mobile agents are discussed.
System Theory, 2004. Proceedings of the Thirty-Sixth Southeastern Symposium on; 02/2004
-
[show abstract]
[hide abstract]
ABSTRACT: This paper explores a hardware implementation of the wavelet transform using field programmable gate arrays (FPGA). Many signal processing applications use wavelets for more efficient signal decomposition than the traditional FFT technique. The parallel architectures of the FPGA are most suited to the practical filter bank implementations of the wavelets. Comparative hardware analyses show that wavelet transform outperforms the Fourier transform.
System Theory, 2003. Proceedings of the 35th Southeastern Symposium on; 04/2003
-
[show abstract]
[hide abstract]
ABSTRACT: Using a genetic algorithm (GA) to evolve ant colony systems (ACS),
we have succeeded at producing evolvable random number generators (RNG)
that can be written to hardware. Although the simulated behavior of
individual ants is limited to a small number of choices, "fit" colonies
pass many stringent tests for randomness
Evolutionary Computation, 2002. CEC '02. Proceedings of the 2002 Congress on; 02/2002
-
S.Y. Foo
[show abstract]
[hide abstract]
ABSTRACT: In this paper, a fuzzy logic approach is applied to detect
hydrocarbon fires in aircraft dry bays and engine compartments. The
inputs to the fuzzy system consist of a set of statistical measures
derived from the histogram and image subtraction analyses of successive
image frames. Specifically, fuzzy rules based on the median, standard
deviation, and normalized first-order moment statistical measures of
histogram data and the mean statistical measure of image subtraction
data of successive frames are used to compute the probability of a fire
event. This fuzzy logic approach is also tested for false alarms such as
those due to flashlights and high-power halogen lights. It is shown that
image subtraction analysis can be used to accurately distinguish fires
from false alarms
IEEE Transactions on Industrial Electronics 11/2000; · 5.16 Impact Factor
-
S.Y. Foo
[show abstract]
[hide abstract]
ABSTRACT: In this paper, a machine approach is applied to detect hydrocarbon
fires in aircraft dry bays and engine compartments. The inputs to the
machine vision system consist of a set of statistical measures derived
from the histogram and image subtraction analyses of successive image
frames. Specifically, heuristic rules based on the median, standard
deviation and normalized first-order moment statistical measures of
histogram data and the mean statistical measure of image subtraction
data of successive frames are used to compute the likelihood of a fire
event. This machine vision system is also tested for false alarms such
as those due to flashlights and high-power halogen lights
IEEE Transactions on Industry Applications 04/2000; · 1.66 Impact Factor
-
S.Y. Foo
[show abstract]
[hide abstract]
ABSTRACT: In this paper, a machine vision approach is applied to detect and
categorize hydrocarbon fires in aircraft dry bays and engine
compartments. Images for computer analysis are provided by
charge-coupled device imaging sensors placed inside dry bays and engine
compartments. Using a set of heuristics based on statistical measures
derived from the histogram and image subtraction analyses of successive
image frames, we showed that it is possible to detect and categorize
life-threatening fires from nonfire/nonlethal events accurately in
submillisecond response time. Specifically, the median, standard
deviation, and 1st-order moment statistical measures of the histogram
data of each image frame are used to confirm the presence or absence of
fire. Concurrently, another set of mean, median, and standard deviation
statistical measures from the image subtraction of two successive frames
are used to determine the growth and subsequently reaffirm the existence
of a fire. This approach is also tested for false alarms such as those
due to flashlights and high-power halogen lights
Industry Applications Conference, 1995. Thirtieth IAS Annual Meeting, IAS '95., Conference Record of the 1995 IEEE; 11/1995
-
[show abstract]
[hide abstract]
ABSTRACT: Principles of operation and basic building blocks of artificial neural networks are described. Deterministic components, comprising variable linear conductance devices and components used for processing elements, are discussed. These devices are analyzed using SPICE. The reasons for using simple analog circuits rather than digital circuits are examined.< >
IEEE Circuits and Devices Magazine 08/1990; · 1.18 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The issues that must be addressed before commercial database management systems can be used to manage VLSI CAD data are defined. A survey is presented of approaches addressing four of the defined issues: design hierarchies and multilevel representations, design alternatives and version control, common interface between cell libraries and efficient cell selection based on given design constraints. A frame-based model is considered as a case study of the special-purpose design database management system approach. This framework for capturing design data is based on semantic networks. It is well suited for application-specific ICs, yet general enough for other CAD/CAM environments. Benchmark results for the selection algorithms that run on top of the frame-based database system are presented.< >
Computer 03/1990; · 1.47 Impact Factor
