-
公开(公告)号：KR1020090065218A
公开(公告)日：2009-06-22
申请号：KR1020070132695
申请日：2007-12-17
Applicant: 한국전자통신연구원
CPC classification number: G06F11/2028 , G06F11/2025
Abstract: A cluster system and a method for operating the same are provided to enable an active management in a normal node without the control for the management nodes in the case of node failure. A working list is registered at a working list of a board server(10), and an agent server(20) manages the working board. Plural general server nodes(30a-30n) perform a correspondent work based on the working list, and performs a specific work instead the general server node at which a failure occurs.
Abstract translation: 提供了一种群集系统及其操作方法，以便在节点故障的情况下，能够在普通节点中进行主动管理，而无需对管理节点进行控制。 在板服务器（10）的工作列表中登记工作列表，代理服务器（20）管理工作板。 多个通用服务器节点（30a-30n）基于工作列表执行对应工作，并执行特定工作，而不是发生故障的一般服务器节点。
-
公开(公告)号：KR1020090061522A
公开(公告)日：2009-06-16
申请号：KR1020070128553
申请日：2007-12-11
Applicant: 한국전자통신연구원
CPC classification number: G06F11/3495 , G06F11/2025 , H04L41/046 , H04L41/0672 , H04L41/0677 , H04L67/125 , H04L69/40
Abstract: A large-scale cluster monitoring system and a method for automatically constructing/restoring the same are provided to increase the availability by actively coping with configuration information in case of the node fail. A CM(Cluster Master) server(110) manages nodes in a system of a large-scale cluster environment. A DB server(120) stores monitoring information which is the status information of each nodes within a group. A plurality of GM(Group Master) nodes(130-1~130-n) collect the monitoring information. A plurality of NA(Node Agent) nodes(140-1~140-n) access the CM server to obtain GM node information. A DB agent(150) observes the monitoring information of each nodes to detect the failure of each node.
Abstract translation: 提供了大规模集群监控系统及其自动构建/恢复方法，通过在节点出现故障的情况下积极应对配置信息来提高可用性。 CM（Cluster Master）服务器（110）管理大规模集群环境的系统中的节点。 DB服务器（120）存储监视信息，该信息是组内每个节点的状态信息。 多个GM（组主）节点（130-1〜130-n）收集监控信息。 多个NA（节点代理）节点（140-1〜140-n）访问CM服务器以获得GM节点信息。 DB代理（150）观察每个节点的监视信息，以检测每个节点的故障。
-
公开(公告)号：KR100772195B1
公开(公告)日：2007-11-01
申请号：KR1020060079492
申请日：2006-08-22
Applicant: 한국전자통신연구원
Abstract: A multi-streaming service method and an apparatus therefor are provided to solve a problem that when contents are transmitted to a user terminal because the terminal should simultaneously decode and execute multi-source streams, a load of the user terminal greatly increases. Contents servers(400-402) store particular contents divided in conformity with a certain rule. A conversion server(300) receives multi-source streaming data from the contents servers(400-402) and converts them to a single streaming. A central server manages a streaming service. The conversion server(300) is connected with a repeater(200) that performs data transmission with a user terminal(100). The contents data stored in the contents servers(400-402) are hierarchically coded or multi-description coded.
Abstract translation: 提供了一种多流服务方法及其装置，用于解决由于终端同时解码并执行多源流而将内容发送到用户终端的问题，因此用户终端的负担大大增加。 内容服务器（400-402）存储按照某个规则划分的特定内容。 转换服务器（300）从内容服务器（400-402）接收多源流数据，并将它们转换成单个流。 中央服务器管理流服务。 转换服务器（300）与用户终端（100）执行数据传输的中继器（200）连接。 存储在内容服务器（400-402）中的内容数据被分层编码或多描述编码。
-
公开(公告)号：KR1020070104102A
公开(公告)日：2007-10-25
申请号：KR1020060036280
申请日：2006-04-21
Applicant: 한국전자통신연구원
IPC: H04N21/238 , G06Q50/10 , H04N21/432 , H04N21/44 , H04N21/234 , H04L29/06
CPC classification number: H04N21/238 , G06Q50/10 , H04L29/06 , H04N21/23418 , H04N21/4325 , H04N21/44004
Abstract: A cluster-based streaming system using a multiple description coding, and a method thereof are provided to provide a service continuously even if a part of servers is out of order and to supply the service applied to a client/network condition. A cluster-based streaming system using a multiple description coding comprises a metadata database(200), a dispatcher(100), and a streaming server(300). The metadata database is used to keep metadata including description information of each content, streaming server information, server load information, and presently-serviced stream information. The dispatcher is used to receive the metadata from the metadata database and to return the streaming server information decided on the basis of the received metadata, to a client. The streaming server is used to keep a description of the content and to transfer the applicable description by using the information of the client by receiving the client information and a description file name from the dispatcher.
Abstract translation: 提供使用多描述编码的基于群集的流系统及其方法，用于连续地提供服务，即使服务器的一部分出现故障并提供应用于客户/网络状况的服务。 使用多描述编码的基于群集的流系统包括元数据数据库（200），调度器（100）和流服务器（300）。 元数据数据库用于保存包括每个内容的描述信息，流服务器信息，服务器负载信息和当前服务的流信息的元数据。 调度器用于从元数据数据库接收元数据，并将基于接收的元数据确定的流服务器信息返回给客户端。 流服务器用于保存内容的描述，并通过从分派器接收客户端信息和描述文件名称，通过使用客户端的信息来传送适用的描述。
-
公开(公告)号：KR101694306B1
公开(公告)日：2017-01-09
申请号：KR1020120146900
申请日：2012-12-14
Applicant: 한국전자통신연구원
Inventor： 정진환
CPC classification number: G06F9/4887 , G06F9/5027 , G06F2209/483 , G06F2209/5017 , G06F2209/5019 , G06F2209/509
Abstract: 설계중간단계에서하드웨어가속장치를이용한병렬화처리에따른성능향상을예측하도록한 하드웨어가속장치의병렬화에따른성능예측장치및 방법이제시된다. 제시된하드웨어가속장치의병렬화에따른성능예측장치는작업의수행시간을복수의작업단계들및 처리단계들로구분하고, 병렬화지수및 목표성능중에하나를설정하고, 복수의작업단계들및 처리단계들각각의수행시간을근거로작업중에병렬화대상단계의비율및 속도향상값중에적어도하나를산출하고, 산출한복수의작업단계들및 처리단계들의수행시간들과산출한병렬화대상단계의비율과속도향상값및 설정된목표성능중에적어도하나를근거로예상성능값또는병렬화지수를산출한다.
Abstract translation: 这里公开了一种用于预测归因于硬件加速装置的并行化的性能的装置和方法。 该装置包括设置单元，操作单元和预测单元。 设置单元将执行任务所需的时间分成多个任务阶段和处理阶段，并且设置并行化索引和目标性能之一。 操作单元计算执行阶段所需的时间，并且计算任务中的目标并行化阶段的比率和速度改善值中的至少一个。 预测单元基于计算出的执行阶段所需的时间，计算出的目标并行化阶段的比例，计算的速度改善值和设定的目标性能中的至少一个来计算期望的性能值或并行化指标。
-
Patent Agency Ranking
公开(公告)号：KR101378939B1
公开(公告)日：2014-04-02
申请号：KR1020100081723
申请日：2010-08-24
Applicant: 한국전자통신연구원
Abstract: 본 발명은 다차원 공간에서 스카이라인 객체가 되기 위한 최소 비용 벡터 획득 방법 및 장치에 관한 것으로, 본 발명에 따른 다차원 공간에서 스카이라인 객체가 되기 위한 최소 비용 벡터 획득 방법은, 복수개의 좌표 축으로 이루어진 다차원 공간에서, 질의 점 및 상기 질의 점을 이동시킨 복수개의 이동 점들 각각을 양 끝점으로 하는 각각의 벡터 값들을 계산하는 단계; 및 상기 벡터 값들 중에서 이동 점이 스카이라인(Skyline)에 포함되되 상기 질의 점과의 거리 값이 최소인 벡터 값을 최소 벡터 값으로 선택하는 단계를 포함한다.
-
公开(公告)号：KR1020110040657A
公开(公告)日：2011-04-20
申请号：KR1020100081723
申请日：2010-08-24
Applicant: 한국전자통신연구원
CPC classification number: G06F17/30979 , G06F17/16
Abstract: PURPOSE: A minimum cost vector acquiring method and apparatus thereof are provided to supply a query processing method which is useful for decision making by supplying an efficient method in order to include an object in a skyline. CONSTITUTION: A vector value calculator(711) calculates a vector value in multidimensional space. A query point and a moving point are determined as both ends of a vector. A minimum vector value processor(712) selects a minimum value which is included in a skyline. The minimum vector selects a vector in which the distance with the query point is minimum.
Abstract translation: 目的：提供一种最小成本矢量获取方法及其装置，以通过提供一种有效的方法来提供一种对决策有用的查询处理方法，以便将对象包括在天际线中。 构成：向量值计算器（711）计算多维空间中的向量值。 查询点和移动点被确定为向量的两端。 最小向量值处理器（712）选择包括在天际线中的最小值。 最小向量选择与查询点的距离最小的向量。
-
公开(公告)号：KR1020110019182A
公开(公告)日：2011-02-25
申请号：KR1020090076792
申请日：2009-08-19
Applicant: 한국전자통신연구원
CPC classification number: H04W84/18 , H04M1/7253 , H04M2250/02 , H04M2250/04 , H04M2250/06 , H04W84/14
Abstract: PURPOSE: A mobile terminal and a method thereof are provided to transmit service data to a peripheral device through local area wireless network connection are provided to maximize the convenience of a user without deleting the transmitted data. CONSTITUTION: A mobile terminal(100) uses a local area wireless communication with a peripheral device. The mobile terminal configures a wideband wireless communication network connection by a wideband wireless communication method. The mobile terminal transmits service data to the peripheral device through the wideband wireless communication network.
Abstract translation: 目的：提供一种移动终端及其方法，用于通过局域无线网络连接向外围设备发送业务数据，以最大限度地提高用户的便利性，而不会删除所发送的数据。 规定：移动终端（100）使用与外围设备的局域无线通信。 移动终端通过宽带无线通信方式配置宽带无线通信网络连接。 移动终端通过宽带无线通信网络将服务数据发送到外围设备。
-
公开(公告)号：KR1020080051043A
公开(公告)日：2008-06-10
申请号：KR1020070107668
申请日：2007-10-25
Applicant: 한국전자통신연구원
IPC: H04L29/06 , H04N21/2662 , H04N21/238 , H04L12/851
CPC classification number: H04N21/2389 , H04L47/2416 , H04L69/22 , H04N21/2662
Abstract: A system and a method for transmitting contents are provided to reduce the cost according to the use of a streaming service and to offer contents to many users at the same time. A streaming service control manager(100) manages the contents streaming operation according to a contents streaming transmission environment. An inverse multiplexing manager(101) receives the input data and performs the inverse multiplexing work. A contents adaptive manager(102) converts a mode into an I Frame only mode or an Audio Only Play mode in case that the state of the network is bad and performs the adaptive conversion work of contents. A re-multiplexing manager(103) receives the adapted data through the contents adaptive manager and performs the re-multiplexing work. The stream service control manager includes a module. The module outputs the input data immediately in case that the state of the network is good, provides the streaming transmission service and also outputs the adaptively converted data in case that the state of the network is bad.
Abstract translation: 提供了一种用于发送内容的系统和方法，以根据流式传输服务的使用降低成本，并且同时向许多用户提供内容。 流服务控制管理器（100）根据内容流传输环境管理内容流操作。 逆多路复用管理器（101）接收输入数据并执行逆多路复用工作。 内容自适应管理器（102）在网络状态不良并执行内容的自适应转换工作的情况下将模式转换为仅I帧模式或仅音频播放模式。 再复用管理器（103）通过内容自适应管理器接收适配数据，并执行重新复用工作。 流服务控制管理器包括一个模块。 在网络状态良好的情况下，模块立即输出输入数据，提供流传输服务，并在网络状态不佳的情况下输出自适应转换的数据。
-
公开(公告)号：KR100734629B1
公开(公告)日：2007-07-03
申请号：KR1020050095554
申请日：2005-10-11
Applicant: 한국전자통신연구원
IPC: G06F15/16
CPC classification number: G06F12/12 , G06F12/0875 , G06F12/127
Abstract: 본 발명은 적응형 콘텐츠 전송에 있어서 객체의 분할을 통해 캐싱 공간을 최대한 확보하면서, 다양한 목적지 유형의 캐싱이 가능한 부분 캐싱 방법 및 이를 이용한 적응형 콘텐츠 전송 방법에 관한 것으로서, 그 구성은 캐시를 이용하여 임의 유형의 객체를 다수의 목적지 유형으로 트랜스코딩하는 트랜스코딩 프록시 스트리밍 서버에 있어서, 각 객체를 두 개로 분할하기 위한 분할 위치를 설정한 후, 캐시 내의 공간 확보를 위한 퇴출 요청이 수신되면, 캐시에 저장된 각 객체를 상기 설정된 분할위치를 기준으로 하여 선두에 위치하는 제1 부분와, 후미에 위치하는 제2 부분으로 분할한 후, 상기 제2 부분을 제거하고 제1 부분만을 캐시에 남기는 것을 기술적 요지로 한다.
적응형 콘텐츠 전송, 스트리밍, 트랜스코딩, 캐싱,
-
-
-
-
-
-
-
-
-
Search Results
30
records
(took 0.023 seconds)
