
http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
양희재,강상윤 대한미용성형외과학회 2020 Archives of Aesthetic Plastic Surgery Vol.26 No.1
Giant congenital melanocytic nevi are rare, and it is recommended to excise due to the possibility of malignancy. Although tissue expansion for nevus removal involves multiple stages, the advantages of tissue expansion are known well. We report a case of a giant congenital melanocytic nevus of the upper extremity that extended from the elbow to the shoulder. This case was challenging due to the circumferential nature of the lesion, the lack of healthy skin, the need to ensure minimal scarring, and the desire to impact the patient’s daily life as little as possible. The use of a pre-expanded thoracodorsal artery perforator (TDAP) flap and re-expansion of the transferred flap were considered most appropriate. A pre-expanded pedicled TDAP flap can be transferred to any anatomical region with its versatility. The re-expansion of the transferred flap ensures the health of the skin and minimizes donor site scarring. Here, we achieved the optimal outcome through the combined use of a pre-expanded pedicled TDAP flap and sequential re-expansion of the transferred flap in the treatment of a giant nevus of the upper extremity.
양희재 慶星大學校 1998 論文集 Vol.19 No.2
WebCheck, a WWW log analyzer, was developed to understand user surfing behavior by exploring the Web sever log files efficiently. As the popularity of the Web has exploded, there is a strong desire to analyze the usage of one's own Web server. Although many log analyzers have been already developed and used in real world, little is known about the internal structure of a log analyzer and few research parers on it have been published yet. Furthermore, as the size of Web server log files increases rapidly, efficiency is becoming the most important issue for which a log analyzer must have. In this paper, we describe the design of WebCheck and point out what is important to build an efficient log analyzer. Specifically we propose some hash functions to be applied to efficiently analyze the domain names of client hosts as well as the path names the clients requested. WebCheck is implemented on a Windows 95 system and the experiment says that it outperforms than many other commercial analyzers because of its proprietary efficient hash functions.
내장형 시스템을 위한 자바 클래스 파일에서 상수풀 항목에 대한 통계
양희재,권중장 경성대학교 공학기술연구소 2002 공학기술연구지 Vol.9 No.-
Constant pool is known to occupy the biggest part in size in a typical Java class file. It is very important for any embedded system to restrict the usage of memory since memory is one of the most vital resource of the system. Because constant pool holds the majority of Java class file in size, we are interested in knowing how to reduce its effect. In this paper we have analyzed all entries in constant pool and gathered the statistics in detail. Specifically, we have investigated a hundred of Java class files for embedded system, one from the J2ME/CLDC classes of Sun Microsystems, and others from the simpleRTJ classes of RTJ Computing. The statistics said that there are forty four constants on the average in the pool and only six percents of them are actually used to execute the bytecode instructions. Another seventy eight percents of the constant are used merely for type-checking and linking classes. It suggests the possibility of reducing memory usage by restricting runtime type-checking and dynamic class loading, which is not an extraordinary condition for embedded system. The result would be applied to build a more memory-efficient embedded Java system.
논리적 URL 사용을 위한 포워딩 서버의 설계 및 구현
양희재 한국정보통신학회 2003 한국정보통신학회논문지 Vol.7 No.2
웹이 인터넷 정보도구의 표준으로 자리잡게 됨에 따라 점점 더 많은 웹 사이트들이 만들어졌고, 동시에 더 많은 웹 주소 공간이 필요하게 되었다. 그러나 웹 주소를 나타내는 물리적 URL은 기억하기에 너무 길뿐 아니라 웹 사이트 내용을 직관적으로 나타내기 어려운 점이 많았다. 본 논문에서는 물리적 URL 대신 논리적 URL을 사용하여 이런 어려움을 개선할 수 있도록 하는 URL 전달 서버의 구현에 대해 고찰하였다. URL 전달서버는 HTTP 재방향 트랜잭션을 이용하는 도메인 매퍼와 BIND에 기반한 네임서버로 구성되어지며, 본 논문에서는 이들 상호간의 작용 및 동작에 대해 상세히 연구하였다. 실험 결과 URL 전달에 따른 시간은 현재의 인터넷 트래픽 조건에 비해 무시할 만한 수준인 것으로 나타났다. A lot many WWW sites are come into the world more and more as Web is considered as the unified Internet information tool. The location of each site or resource is usually specified by a physical URL, which is often too long to remember and tends to raise difficulty to show the aim of the site intuitively by seeing it. Since any person or organization can get his/her own domain name easily, it is more desirable to use a logical URL with the domain name which can be chosen more compact to remember and meaningful to represent the ultimate intention of the site. This paper presents an implementation of a URL forwarding server which forwards a URL to another, so that a WWW site can use a logical URL instead of a physical one. The server consists of a domain mapper which uses the redirection transaction of the HTTP protocol, and a name server based on the HIND. The paper shows how the interaction between the domain mapper and the name sever can make forwarding possible and describes its implementation in detail. Experimental results shows that the overhead incurred by URL forwarding is negligible compared to the typical delay of current Internet traffic condition.
양희재 경성대학교 공학기술연구소 2002 공학기술연구지 Vol.9 No.-
A lot many WWW sites are come into the world more and more as Web is considered as the unified Internet information tool. The location of each site or resource is usually specified by a physical URL, which is often too long to remember and tends to raise difficulty to show the aim of the site intuitively by seeing it. Since any person or organization can get his/her own domain name easily, it is more desirable to use a logical URL with the domain name which can be chosen more compact to remember and meaningful to represent the ultimate intension of the site. This paper presents an implementation of a URL forwarding server which forwards a URL to another, so that a WWW site can use a logical URL instead of a physical one. The server consists of a domain mapper which uses the redirection transaction of the HTTP protocol, and a name server based on the BIND. The paper shows how the interaction between the domain mapper and the name server can make forwarding possible and describes its implmentation in detail. Experimental results shows that the overhead incurred by URL forwarding is negligible compared to the typical delay of current Internet traffic condition.
다중 컴퓨터를 이용한 고속 시뮬레이션 기법에 관한 연구
양희재 慶星大學校 1992 論文集 Vol.13 No.2
A fast simulation method, called the parallel batch means method, is presented and analyzed in this paper. Contrary to the conventional parallel simulation methods, it assumes a large grain-size parallelism and thus it is less dependent on the speed of communications between the computers. The method is based on the theory of batch means, which is a well-known simulation output analysis technique. The method achieves a linear speedup characteristic and an experimental simulation using the method shows that it produces even accurate simulation results with great saving in simulation time. As inexpensive computers are increasingly available nowadays, the method will be indispensable for fast and accurate computer simulation.
다중켬퓨터 상호연결망의 구조가 일의 할당에 미치는 영향
양희재 慶星大學校 1992 論文集 Vol.13 No.3
In this paper we examine the effects of multicomputer interconnection networks on the task assignment scheduling as a means of evaluating various types of networks. For the sake of task assignment, i.e. in order to easily find the optimal assignment scheduling, the size of the configuration space for the assignment should be as small as possible. This paper introduces a new concept, called the equivalent assignment, and then apply the idea to analyzing the configuration space. The analysis indicates that the size of configuration space is greatly influenced by the structure of given interconnection network. The configuration spaces for some typical interconnection networks are analyzed and the result says that the hypercube structure has minimal size of the space. It means that the hypercube structure is better than other kinds of network in respect of task assignment scheduling.