专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

KR102132394B1 메모리 장치 및 메모리 장치의 인코딩 방법 有权
公开(公告)号：KR102132394B1
公开(公告)日：2020-07-10
申请号：KR1020180025223
申请日：2018-03-02
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 한미선
IPC分类号： G06F3/06 , G11C5/14

2. 发明授权

KR102122821B1 결함 정보 저장 테이블을 이용한 리페어 분석 시스템 및 그의 리페어 분석 방법 有权
公开(公告)号：KR102122821B1
公开(公告)日：2020-06-15
申请号：KR1020180023124
申请日：2018-02-26
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 최규현 , 전재영
IPC分类号： G11C29/00 , G11C29/12 , G11C11/401

3. 发明授权

KR101805694B1 데이터의 선택적 시프트 연산을 이용한 데이터의 확장 및 축소를 위한 병렬 연산용 데이터 변환 장치 및 그 방법 有权
标题翻译：用于通过使用数据的选择性移位操作来扩展和减少数据的并行计算的数据转换装置和方法
公开(公告)号：KR101805694B1
公开(公告)日：2017-12-06
申请号：KR1020160107664
申请日：2016-08-24
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 김호권 , 전재영 , 최규현 , 한영선
IPC分类号： G06F9/30
CPC分类号： G06F9/30134
摘要： 본발명은데이터의선택적시프트연산을이용한데이터의확장및 축소를위한병렬연산용데이터변환장치및 그방법을개시한다. 즉, 본발명은입력데이터에서여러특정위치의데이터비트를선택적으로미리설정된제 1 방향으로시프트하여데이터를확장하거나또는상기입력데이터에서여러특정위치의데이터비트를선택적으로미리설정된제 2 방향으로시프트하여데이터를축소하는데있어병렬적으로데이터변환을수행함으로써, 데이터크기가커져도지연시간이늘지않으며, 결함이발생한시스템에서결함회피를위한여분의자원이이용가능할경우데이터변환기의확장기/축소기를적용하여결함감내시스템을구축함으로써입력데이터의특정위치를선택적으로빠르게추출하는정보처리시스템의데이터변환용도로사용할수 있다.
摘要翻译：公开了一种用于使用数据的选择性移位操作来扩展和减少数据的并行计算的数据转换设备和方法。即，本发明是在第二方向上，以在特定位置中的数据位的不同移位选择性地移位到所述第一预定方向上延伸的数据，或选择性预先设定与从输入数据中的特定位置的输入数据的数目的数据位如果发生故障的系统中有用于避免缺陷的冗余资源，则应用数据转换器的扩展器/减速器通过构建缺陷容忍系统，它可以用于快速选择性提取输入数据特定位置的信息处理系统的数据转换。

4. 发明公开

KR1020140088725A 읽기 및 쓰기 접근에 따른 선택적 리프레쉬 기능을 구비한 동적 메모리 장치 및 그 선택적 리프레쉬 방법 有权
标题翻译：具有读取和写入功能的选择性刷新功能的动态随机访问存储器件及其选择性刷新方法
公开(公告)号：KR1020140088725A
公开(公告)日：2014-07-11
申请号：KR1020130000616
申请日：2013-01-03
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 이호균 , 최규현
IPC分类号： G11C11/406 , G11C11/4063 , G11C11/401
CPC分类号： G11C11/40611 , G11C11/40622
摘要： The present invention relates to a dynamic memory device including a selective refresh function according to read and write access and a selective refresh method. More specifically, the said invention reduces the power consumption by minimizing refresh by considering replicated data correlation among the LLC; the physical storage medium and the dynamic memory by selectively refreshing a data cell based on the read access and write access of the physical storage medium; and the LLC to directly exchange the data with the dynamic memory. The present invention improves the performance of the read and the write, reduces memory access time by selectively designating an access block or a refresh line based on an updated bit value of a refresh bit cell by constructing the refresh bit cell which represents a plurality of refresh bit vector cells, and the refresh bit vector cell corresponding to the access block or the line of the data cell by increasing a substantive refresh efficiency and a low buffer hit rate.
摘要翻译：本发明涉及一种动态存储装置，其包括根据读写访问和选择刷新方法的选择刷新功能。更具体地，所述发明通过考虑LLC中的复制数据相关性来最小化刷新来降低功耗; 物理存储介质和动态存储器，通过基于物理存储介质的读取访问和写访问有选择地刷新数据单元; 和LLC直接与动态内存交换数据。本发明通过构造表示多个刷新的刷新位单元，通过基于刷新位单元的更新的位值有选择地指定访问块或刷新线来提高读取和写入的性能，从而减少存储器访问时间通过增加实质的刷新效率和低的缓冲命中率，对应于访问块或数据单元的行的刷新位向量单元。

5. 发明公开

KR1020120019949A 수동 통신 시스템에서의 데이터 전송 방법 有权
标题翻译：无源通信系统的数据传输方法
公开(公告)号：KR1020120019949A
公开(公告)日：2012-03-07
申请号：KR1020100083583
申请日：2010-08-27
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 김종옥 , 황석중
IPC分类号： H04B5/02 , H04W56/00 , H02J17/00
CPC分类号： H04L1/0007 , H04B5/0031 , H04B5/0037 , H04L1/1809 , H04W4/12
摘要： PURPOSE: A data transmission method in a passive communication system is provided to variably set the number of synchronizing signals considering the current link frequency properties to optimize the transmission efficiency. CONSTITUTION: An active device transmits a request message with a number of synchronizing signals which are determined by the communication environment information. A passive device transmits a response message with a message fraction bundle consisting of a synchronizing signal and one or more message fractions to the active device(S134,S140). The active device determines the number of message fractions located in between the synchronizing signals or the length of the message fractions based on the communication environment information. The active device includes determined information in the request message.
摘要翻译：目的：提供无源通信系统中的数据传输方法，以便考虑当前的链路频率特性来可变地设置同步信号的数量，以优化传输效率。构成：活动装置发送具有由通信环境信息确定的多个同步信号的请求消息。无源设备将具有由同步信号和一个或多个消息分数组成的消息分组的响应消息发送到活动设备（S134，S140）。有源设备基于通信环境信息确定位于同步信号之间的消息分数的数量或消息分数的长度。活动设备包括请求消息中确定的信息。

6. 发明授权

KR100975085B1 복수개의 기억장치를 구비한 시스템에서 사용되는 전력관리장치 有权
标题翻译：具有多个存储介质的系统的电源管理单元
公开(公告)号：KR100975085B1
公开(公告)日：2010-08-11
申请号：KR1020080092371
申请日：2008-09-19
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 오용수 , 이광도
IPC分类号： G06F3/06 , G06F1/26 , G06F1/00
摘要： 복수개의 기억장치를 구비한 시스템에서 사용되는 전력관리장치는, 복수개의 기억장치 각각에 대응되어 전원을 연결하거나 또는 차단하는 복수개의 스위치와, 시스템의 상태에 따라 복수개의 기억장치를 모두 구동시킬지, 또는 특정의 기억장치 만을 구동시킬지 결정하고, 이러한 결정에 따라 구동시키고자 하는 기억장치에 대응하는 스위치에는 온 신호를 보내고, 구동시킬 필요가 없는 기억장치에 대응하는 스위치에는 오프 신호를 보내는 제어부를 구비한다.

7. 发明公开

KR1020100022743A 영상관리 시스템 및 방법 有权
标题翻译：管理视频数据的系统及其方法
公开(公告)号：KR1020100022743A
公开(公告)日：2010-03-03
申请号：KR1020080081412
申请日：2008-08-20
申请人： 고려대학교 산학협력단
发明人： 김선욱 , 조명진
IPC分类号： H04N5/765 , H04N5/76 , G11B27/02
CPC分类号： G06F17/3007 , G06F17/30321 , G06F17/30781
摘要： PURPOSE: An image management system and a method thereof are provided to record various information of an image to a file name, thereby recovering an index file only with a file name without opening the file when the index file is damaged. CONSTITUTION: An image management system includes an input/output unit, an image process unit(102), a timer unit(103), a memory unit(107), and a control unit(104). The input/output unit exchanges an event signals with an external hardware device such as a camera and a sensor. The image process unit generates an image by processing a camera signal provided from the input/output unit. The timer unit detects time of event generation. The memory unit stores images processed by the image process unit, file names of the images, and index files. The control unit generates recording information of image generation processes as file names and generates index files by indexing file names.
摘要翻译：目的：提供一种图像管理系统及其方法，用于将图像的各种信息记录到文件名，从而在索引文件被损坏时不打开文件而仅用文件名恢复索引文件。构成：图像管理系统包括输入/输出单元，图像处理单元（102），定时器单元（103），存储单元（107）和控制单元（104）。输入/输出单元与诸如照相机和传感器的外部硬件设备交换事件信号。图像处理单元通过处理从输入/输出单元提供的相机信号来生成图像。定时器单元检测事件生成的时间。存储单元存储由图像处理单元处理的图像，图像的文件名和索引文件。控制单元生成图像生成处理的记录信息作为文件名，并通过索引文件名来生成索引文件。

8. 发明公开

KR1020090093620A 수동 ＲＦＩＤ의 대용량 데이터 전송 방법 및 시스템 有权
标题翻译：用于发送被动RFID的大量数据的RFID系统和方法
公开(公告)号：KR1020090093620A
公开(公告)日：2009-09-02
申请号：KR1020080019252
申请日：2008-02-29
申请人： 삼성전자주식회사 , 고려대학교 산학협력단
发明人： 구자남 , 김선욱 , 이준구
IPC分类号： G06K17/00 , G06K19/07
CPC分类号： G06K19/0723 , H04L69/08 , H04W28/06 , H04W84/18
摘要： An RFID system and a method for transmitting large data of a passive RFID are provided to use data without power consumption by transmitting data to an external device, which processes data, without a memory of a passive RFID tag. A data input unit(110) receives the input of the data that are transmitted to a passive RFID tag, and a controller(120) generates a transmission packet. The transmission packet includes the data and a command that commands data transmission. A communication unit(130) converts the generated transmission packet to an RF signal, and transmits the converted RF signal to the passive RFID tag.
摘要翻译：提供一种用于发送无源RFID的大数据的RFID系统和方法，以通过向没有无源RFID标签的存储器处理数据的外部设备发送数据而使用无功耗的数据。数据输入单元（110）接收被发送到无源RFID标签的数据的输入，控制器（120）生成发送分组。传输分组包括数据和命令数据传输命令。通信单元（130）将生成的发送分组转换为RF信号，并将转换的RF信号发送到无源RFID标签。

9. 发明公开

KR1020090032721A 모드 집합을 사용하는 명령어 처리기에서의 모드 명령어최적화 방법 失效
标题翻译：使用模式集的指令处理器中模式集优化的方法
公开(公告)号：KR1020090032721A
公开(公告)日：2009-04-01
申请号：KR1020070098184
申请日：2007-09-28
申请人： 고려대학교 산학협력단
发明人： 추지호 , 한영선 , 김선욱
IPC分类号： G06F9/44 , G06F9/30
CPC分类号： G06F9/4552 , G06F8/443 , G06F9/30003 , G06F9/45525
摘要： A method for optimizing a mode command in a command processor using a mode set is provided to optimize the mode command which is used in the command processor such as a digital signal processor even without modifying a control flow diagram. A mode flow diagram is generated by adding mode information about a mode set command to each node of a control flow diagram(S101). The control flow diagram processes a plurality of commands. The input mode and output mode states of each node which forms the generated mode flow diagram are calculated by a circulation data flow analysis. The mode set command is inserted in to each node depending on whether the input mode and output mode states are changed or not(S104). Each node of the mode flow diagram shows one command. The circulation data flow analysis is repeatedly performed until there is no change of the input mode and output mode states.
摘要翻译：提供了一种使用模式集来优化命令处理器中的模式命令的方法，以优化在诸如数字信号处理器的命令处理器中使用的模式命令，即使不修改控制流程图。通过将关于模式设置命令的模式信息添加到控制流程图的每个节点来生成模式流程图（S101）。控制流程图处理多个命令。通过循环数据流分析计算形成生成模式流程图的每个节点的输入模式和输出模式状态。根据输入模式和输出模式状态是否改变，将模式设置命令插入到每个节点（S104）。模式流程图的每个节点显示一个命令。重复执行循环数据流分析，直到输入模式和输出模式状态没有变化。

10. 发明授权

KR100769688B1 다중 임계값을 이용하는 태그 리드 방법, 그 기록 매체 및그 장치 失效
标题翻译：用多个阈值读取标签的方法，记录介质及其装置
公开(公告)号：KR100769688B1
公开(公告)日：2007-10-23
申请号：KR1020060046838
申请日：2006-05-25
申请人： 고려대학교 산학협력단
发明人： 이준구 , 황석중 , 김선욱
IPC分类号： G06K17/00
CPC分类号： G06K7/10029
摘要： A method for reading a tag with multi-threshold value, and a recording medium and a device thereof are provided to improve the efficiency even if the number of tags to be read and a frame size is big, and raise a tag recognition speed by using multi-threshold value instead of a complex method of estimating the number of tag. A ratio operator(610) operates a ratio of colliding slots and empty slots in a frame when tag recognition is tried. A frame differential increasing part(620) increases the frame size to the size multiplying the frame size by a first size if a colliding slot ratio is over a first threshold value, and increases the frame size to the size multiplying the frame size by a second size if the colliding slot ratio is over a second threshold value. A frame differential decreasing part(630) decreases the frame size to the size multiplying the frame size by a third size if an empty slot ratio is over a third threshold value, and decreases the frame size to the size multiplying the frame size by a fourth size if the empty slot ratio is over a fourth threshold value. A tag recognizer(640) tries the tag recognition by using the frame size.
摘要翻译：提供一种用于读取具有多阈值的标签的方法，以及记录介质及其装置，以便即使要读取的标签的数量和帧尺寸大，也提高效率，并且通过使用标签识别速度来提高标签识别速度多阈值，而不是估计标签数量的复杂方法。当尝试标签识别时，比例运算符（610）操作帧中的冲突槽和空槽的比率。如果碰撞时隙比超过第一阈值，则帧差分增加部分（620）将帧大小增加到将帧大小乘以第一大小的大小，并且将帧大小增大到将帧大小乘以第二如果碰撞槽隙比超过第二阈值，则尺寸。如果空隙率超过第三阈值，则帧差分减小部分（630）将帧大小减小到将帧大小乘以第三大小的大小，并且将帧大小减小到将帧大小乘以第四如果空时隙率超过第四阈值，则为大小。标签识别器（640）通过使用帧大小来尝试标签识别。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式