Seong Ho Son

Design and Test of a Mobile Antenna System With Tri-Band Operation for Broadband Satellite Communications and DBS Reception

The tri-band mobile antenna system for broadband multimedia services in the Ka/K-band and a simultaneous direct broadcast service in the Ku-band has been developed. The radiating part of the antenna consists of a shaped dual reflector system and a tri-band feed of circular polarization. The low-profile offset main reflector has an aperture size of 60 cm 20 cm; therefore, a fan beam is formed which is sharp in azimuth and wide in elevation. The feed system contains a protruding dielectric rod for Ka/K-band and 2 2 active phased array. The latter provides the beam for Ku-band along with four fast electrical beams for stable satellite tracking. The new tracking algorithm based on a fusion technique involving various sensors and fast electrical beams has also been introduced. The antenna system was mounted on a large ship and a test car, and successfully operated via Koreasat-3.

Theoretical analysis for beam pointing accuracy of stair-planar phased array antenna with tracking beam

In this paper, the stair-planar phased array antenna with tracking beam is introduced and the beam pointing characteristics are theoretically analyzed from numerical simulation. The vehicular antenna was developed as engineering model to provide the satellite internet services in mobile environment. Especially, radiator subarrays of this antenna are arranged like stairs to design for low profile. This structure brings about the pointing error between the main and tracking beams due to the difference of each frequency. To solve this trouble, two treatments using RF cable length compensation and 2/sup nd/ level phase compensation are investigated. In addition, this analytic model of beam steering will help to design and evaluate the antenna beam pattern.

A sectorized object matching approach for breast magnetic resonance image similarity study

In this paper, we propose a new image retrieval method consisting of shape feature data. In this approach we assume the images are classified into single objects through other known classification methods such as K-means and SVM algorithms. From collected binary object images, we develop a new algorithm that has less computation but equal efficiency as using shape feature - the curvature of the contour. We have experimented with classified binary object image from actual breast medical images used in real medical diagnosis. Actual experimental results show that the proposed algorithm achieves equal results against traditional image retrieval using curvature of the contour with higher efficiency.

Low sidelobe design by position perturbation in mobile array antenna for satellite communications

This work presents the low sidelobe pattern design of a mobile terminal antenna for satellite communications. This is important to protect interferences from neighboring satellites, which is regularized by international unions such as ITU. To synthesize the array beam pattern with low sidelobes, the position of the array elements is perturbed and optimized by a genetic algorithm. In addition, the characteristics of the optimized array beam patterns are investigated according to the number of array elements. Furthermore, the illustrative antenna, which is a mobile array antenna with a stair-planar configuration for satellite multimedia communication, is introduced to verify this optimizing technique.

Mobile antenna system for Ku-band satellite Internet service

This paper introduces the mobile antenna system for Ku-band satellite Internet service as well as DBS reception on the vehicle. The antenna type is an active phased array with mechanical rotation of which scanning range is the electronic /spl plusmn/10deg in elevation and the mechanical 360deg unlimited in azimuth. The G/T and EIRP are 7 dB/K and 34 dBW, respectively, and the number of subarrays is 24 with 1 by 8 microstrip patch. The distance of each subarray is perturbed with genetic optimization technique in order to reduce the sidelobes of the antenna. The pointing error is within 0.2deg in automatic tracking. In addition, the tracking performance of this mobile antenna system via satellite is verified by road test through highway and downtown.

A novel tracking control realization of phased array antenna for mobile satellite communications

This paper describes a realization of a tracking phased array antenna system for the shipboard station in X-band satellite communication. This antenna system is necessary to accurately track a target satellite due to the narrow beamwidth. This developed antenna has Tx and Rx antenna beams as well as tracking beam. These beams are independently steered electronically in elevation and mechanically in azimuth. However, a Korean target satellite will be launched into geostationary orbit after a few years. We verified the real-time tracking performance by means of transmission and reception of television broadcasting signal between the developed antenna system and satellite transponder simulator under operation of ship motion simulator. In addition, the measured results of antenna system are presented and discussed.

Computer aided breast cancer diagnosis system with fuzzy multiple-parameter support vector machine

Computer Aided Diagnosis (CAD) system has been proven that it can be utilized as the secondary option for physicians for early breast cancer detection. A typical CAD system consists of several phases like image segmentation, feature extraction and selection, classification. Among those phases, the classification phase is one of the important phases that directly affect the performance of the entire system. Therefore the main issue is to enhance the classification phase to construct better decision-making procedure comparing to conventional classification phase by assigning enhanced logic. In this paper, we propose a Fuzzy Multiple-parameter Support Vector Machine (SVM), which will be used in the CAD system. The proposed method uses fuzzy membership to tune up each training data points by assigning proper weight, corresponding to its feature, and adopts multiple parameters as a classifier for SVM, which further improves the machine-learning process to a more robust level. The experimental result shows that the proposed method is far more superior to the existing SVM in terms of performance, sensitivity and accuracy. Additionally, the result suggests for more sophisticated and complex approach to the current classification for CAD system.

Enhanced Breast Cancer Classification with Automatic Thresholding Using SVM and Harris Corner Detection

Image classification and extracting the characteristics of a tumor are the powerful tools in medical science. In case of breast cancer medical treatment, the breast cancer classification methods can be used to classify input images as normal and abnormal classes for better diagnoses and earlier detection with breast tumors. However, classification process can be challenging because of the existence of noise in the images, and complicated structures of the image. Manual classification of the images is time-consuming, and need to be done only by medical experts. Hence using an automated medical image classification tool is useful and necessary. In addition, having a better training data set directly affect the quality of classification process. In this paper, a method is proposed based on supervised learning and automatic thresholding for both generating better training data set, and more accurate classification of the mammogram images into Normal/Abnormal classes. The procedure consists of preprocessing, removing noise, elimination of unwanted objects, features extraction, and classification. A Support Vector Machine (SVM) is used as the supervised model in two phases which are testing and training. Intensity value, auto-correlation matrix value of detected corners, and, energy, are three extracted features used to train the SVM. Experimental results show this method classify images with more accuracy and less execution time compared to existing method.

Similarity measurement with mesh distance fourier transform in 2D binary image

Shape based feature is a widely used method in Content based Image Retrieval (CBIR) for similarity measurements because contours of an image provide relevant information for similarity. In this paper, we propose a novel shape feature named Mesh Distance Fourier Descriptor (MDFD) which takes into account the contour information of each of the boundary points with respect to other contour points in the images such that the relationship of one boundary point is evaluated with respect to all other boundary points in 2D space. In this paper we have used binary images which are classified into single objects using known classification methods such as K-means and SVM algorithms. The proposed method has been compared with Sectorized Object Matching (SOM) and the result shows that the proposed algorithm outperforms SOM in terms of matching of similar images.

Magnetic resonance image retrieval based on contour to centroid triangulation with shape feature similarity

In this paper, we present an image retrieval method based on contour to centroid triangulation with shape feature similarity. We assume test images and database images used in this paper are all single objects that are segmented by known algorithms such as SVM and K-means algorithms. From these classified binary images, we propose novel Shape based image retrieval method integrating sectored characteristic points to the Contour to Centroid Triangulation (CTCT) method using Unique Representation Grid (URG) as shape feature that can perform as the filtering process. The experimental result shows proposed method has improved conventional CTCT in retrieving medical object image compared to conventional CTCT method with 79 percent match rate while CTCT showed 33 percent match.