-
公开(公告)号：KR1020040050627A
公开(公告)日：2004-06-16
申请号：KR1020020078488
申请日：2002-12-10
Applicant: 한국전자통신연구원
Abstract: PURPOSE: A device for collecting IP packets including header information of a link layer is provided to define a common part for analysis, and to realize the defined common part in hardware system, then to make analysis software for configuring a monitoring system lighter, thereby constructing a network monitoring system which can measure in real time. CONSTITUTION: A network interface connection port(200) receives bit stream-type traffic. A PPS(Precise Positioning Service) port(206) generates a precise time stamp. A framer(201) makes the traffic inputted through the network interface connection port(200) into a PDU packet of a link layer. An FPGA(Field Programmable Gate Array)(202) records IP length and IP packets collected according to the designated length in an FIFO buffer(207) by detecting the generated time stamp information and the header of an IP frame and the length of the IP frame, and performs a designated filtering function. A booter(210) performs an initialization process. FIFO buffers(203,204) supply buffer functions for transmitting data between the FPGA(202) and a MPC(Micro Processor Controller)(205).
Abstract translation: 目的：提供一种用于收集包括链路层头信息的IP报文的设备，以定义用于分析的公共部分，并实现硬件系统中定义的公共部分，然后制作分析软件，使监控系统更轻，从而构建 可实时测量的网络监控系统。 构成：网络接口连接端口（200）接收位流类型流量。 PPS（精确定位服务）端口（206）生成精确的时间戳。 成帧器（201）使得通过网络接口连接端口（200）输入的流量成为链路层的PDU分组。 FPGA（现场可编程门阵列）（202）通过检测所生成的时间戳信息和IP帧的头部以及IP长度来记录根据指定长度收集的IP长度和IP数据包在FIFO缓冲器（207）中 帧，并执行指定的过滤功能。 启动器（210）执行初始化处理。 FIFO缓冲器（203,204）提供用于在FPGA（202）和MPC（微处理器控制器）（205）之间传送数据的缓冲器功能。
-
公开(公告)号：KR100411251B1
公开(公告)日：2003-12-18
申请号：KR1020010074723
申请日：2001-11-28
Applicant: 한국전자통신연구원
IPC: H04L12/28
CPC classification number: H04L45/50 , H04L45/12 , H04L45/122 , H04L45/24
Abstract: Disclosed herein is a constrained multipath routing method. In a Multi-Protocol Label Switching (MPLS) network, network conditions and traffic engineering requirements assigned to each link are set. The amount of assignments of each traffic requirement assigned to each link and a minimum value of each maximum link utilization α are calculated. A traffic rate (Xijkl) in which the degree of resource utilization is minimized in the minimized maximum link utilization α is calculated using the calculated amount of assignments of each traffic requirement and which is assigned to each link. Multiple paths corresponding to each traffic requirement and traffic division ratios of the multipath are calculated using the calculated traffic rate.
Abstract translation: 这里公开了一种约束多路径路由方法。 在多协议标签交换（MPLS）网络中，分配给每个链路的网络条件和流量工程要求被设置。 分配给每个链路的每个业务需求的分配量和每个最大链路利用率＆lt; 被计算。 业务速率（Xijkl），其中资源利用程度在最小化的最大链路利用率＆amp; 使用计算出的每个业务需求的分配量以及分配给每个链路的值来计算。 使用所计算的业务速率来计算与每个业务需求和多路径的业务划分比率相对应的多个路径。
-
公开(公告)号：KR100454684B1
公开(公告)日：2004-11-03
申请号：KR1020020079541
申请日：2002-12-13
Applicant: 한국전자통신연구원
IPC: H04L12/24
Abstract: PURPOSE: A method for performing TE(Traffic Engineering) by using simulation and optimization in multi-protocol label switching network, and a server therefor are provided to allow a network manager to effectively devise TE policy. CONSTITUTION: An input processor(31) receives every value inputted by a user. A router interface(32) obtains every data obtained from switches. A simulation logic(33) is a part where simulation techniques are actually implemented. An optimization logic(34) is a part where optimization techniques are actually implemented. A data manager(35) stores and loads every input and output data used for simulation and optimization and their meta data. A result output unit(36) reports the simulation and optimization result to a manager by a report or through a GUI(Graphical User Interface). A result application unit(37) generates a substantial TE command, applies it to a network and checks its result, in automatically reflecting the simulation and optimization result in the network through a server.
Abstract translation: 目的：提供一种通过在多协议标签交换网络中使用仿真和优化来执行TE（流量工程）的方法及其服务器，以允许网络管理员有效地设计TE策略。 构成：输入处理器（31）接收用户输入的每个值。 路由器接口（32）获得从交换机获得的每个数据。 模拟逻辑（33）是实际实现模拟技术的部分。 优化逻辑（34）是实际实现优化技术的部分。 数据管理器（35）存储和加载用于模拟和优化的每个输入和输出数据及其元数据。 结果输出单元（36）通过报告或通过GUI（图形用户界面）将模拟和优化结果报告给管理者。 结果应用单元（37）生成实质的TE命令，将其应用于网络并检查其结果，以通过服务器自动反映网络中的模拟和优化结果。
-
公开(公告)号：KR1020030044134A
公开(公告)日：2003-06-09
申请号：KR1020010074723
申请日：2001-11-28
Applicant: 한국전자통신연구원
IPC: H04L12/28
CPC classification number: H04L45/50 , H04L45/12 , H04L45/122 , H04L45/24
Abstract: PURPOSE: A multiple path allocating method satisfying a limit condition is provided to optimize performance of an Internet backbone network by dividing an optimum traffic of multipath in an MPLS(Multi-Protocol Label Switching) network. CONSTITUTION: A traffic request matrix that given network link connection information, link capacity information, information on nodes, and traffic quantity required between input/output nodes are recorded at every node is recorded(S201). An NTI(Network Topology Information) on general situations of a network is set(S202). A traffic engineering problem for minimizing a maximum link utilization by using the traffic request matrix and the NTI is mathematically modeled by using an MIP(Mixed integer Programming)(S203). A minimum value of the maximum link utilization(a) and a corresponding traffic rate(Xijkl) satisfying it are obtained by performing the mathematically using the MIP(S204). The minimum value of the maximum link utilization(A) is fixed as a constant 'C' and the sum of every resource used for a network, a new network optimization reference, is minimized(S205). A mathematical modeling using the MIP is performed by using the traffic rate assigned to each link in order to obtain multiple paths for each traffic request value and a value of traffic split rate(S206). The maximum utilization(a) of a link and a resource utilization are minimized in setting up multiple paths, and a multiple path satisfying a limit condition of used hops is set(S207).
Abstract translation: 目的：提供满足限制条件的多路径分配方法，通过在MPLS（多协议标签交换）网络中划分多路径的最优流量来优化互联网骨干网络的性能。 规定：记录在每个节点处记录给定网络链路连接信息，链路容量信息，节点信息和输入/输出节点之间所需的流量的业务请求矩阵（S201）。 设置网络的一般情况的NTI（网络拓扑信息）（S202）。 通过使用流量请求矩阵和NTI来最小化最大链路利用率的流量工程问题通过使用MIP（混合整数编程）（S203）进行数学建模。 通过使用MIP进行数学运算，获得满足该最大链路利用率（a）的最大链路利用率（a）的最小值和对应的业务速率（Xijkl）（S204）。 最大链路利用率（A）的最小值被固定为常数“C”，并且使用于网络的每个资源的总和（新的网络优化参考）被最小化（S205）。 通过使用分配给每个链路的流量速率来执行使用MIP的数学建模，以便为每个流量请求值和流量分流速率的值获得多个路径（S206）。 链路的最大利用率（a）和资源利用率在设置多个路径时被最小化，并且设置满足所使用跳数的限制条件的多路径（S207）。
-
公开(公告)号：KR100744559B1
公开(公告)日：2007-08-01
申请号：KR1020060064012
申请日：2006-07-07
Applicant: 한국전자통신연구원
Abstract: 본 발명은 동일한 데이터는 동일한 해싱 값을 생성하고, 유사한 데이터의 경우 해싱 값의 차가 데이터의 유사도와 반비례 관계를 가지도록 해싱 값을 생성하는 유사성 기반 해싱 알고리즘을 이용한 데이터 해싱 방법, 데이터 처리 방법, 및 데이터 처리 시스템을 제공한다. 그 데이터 해싱 방법은 전산 데이터를 입력하는 단계; 및 동일 데이터에 대해서는 동일 해싱 값을 생성하고, 유사한 데이터의 경우 해싱 값의 차가 데이터의 유사도와 반비례 관계를 가지도록 해싱 값을 생성하는 유사성 기반 해싱(Similarity-based Hashing:SBH) 알고리즘을 이용하여 전산 데이터에 대한 해싱 값을 생성하는 단계;를 포함한다. 본 발명의 데이터 해싱 방법은 생성된 해싱 값들의 차가 데이터의 유사도와 반비례 관계를 가지므로, 데이터들의 검색, 비교, 분류 등을 O(1) 또는 O(n) 정도의 시간 복잡도 내에 신속하게 처리할 수 있는 장점을 가진다.
데이터 비교, 데이터 분류, 문서 비교, 문서 분류, 파일 비교, 파일 분류, 유사성, 동일성, 해싱, 핑거프린트-
Patent Agency Ranking
公开(公告)号：KR100461727B1
公开(公告)日：2004-12-17
申请号：KR1020020078488
申请日：2002-12-10
Applicant: 한국전자통신연구원
Abstract: PURPOSE: A device for collecting IP packets including header information of a link layer is provided to define a common part for analysis, and to realize the defined common part in hardware system, then to make analysis software for configuring a monitoring system lighter, thereby constructing a network monitoring system which can measure in real time. CONSTITUTION: A network interface connection port(200) receives bit stream-type traffic. A PPS(Precise Positioning Service) port(206) generates a precise time stamp. A framer(201) makes the traffic inputted through the network interface connection port(200) into a PDU packet of a link layer. An FPGA(Field Programmable Gate Array)(202) records IP length and IP packets collected according to the designated length in an FIFO buffer(207) by detecting the generated time stamp information and the header of an IP frame and the length of the IP frame, and performs a designated filtering function. A booter(210) performs an initialization process. FIFO buffers(203,204) supply buffer functions for transmitting data between the FPGA(202) and a MPC(Micro Processor Controller)(205).
Abstract translation: 目的：提供一种用于收集包括链路层的头部信息的IP包的设备，以定义用于分析的公共部分，并且实现硬件系统中定义的公共部分，然后制作用于配置监测系统更轻的分析软件，从而构建 一个可以实时测量的网络监控系统。 构成：网络接口连接端口（200）接收比特流型业务。 PPS（精确定位服务）端口（206）产生精确的时间戳。 成帧器（201）使通过网络接口连接端口（200）输入的业务量成为链路层的PDU分组。 FPGA（现场可编程门阵列）（202）通过检测生成的时间戳信息和IP帧的报头以及IP的长度，将根据指定长度收集的IP长度和IP分组记录在FIFO缓冲器（207）中 帧，并执行指定的过滤功能。 引导程序（210）执行初始化过程。 FIFO缓冲器（203,204）提供用于在FPGA（202）和MPC（微处理器控制器）（205）之间传输数据的缓冲器功能。
-
公开(公告)号：KR1020040052015A
公开(公告)日：2004-06-19
申请号：KR1020020079733
申请日：2002-12-13
Applicant: 한국전자통신연구원
IPC: H04L12/26
CPC classification number: H04L43/026 , H04L43/045 , H04L43/062
Abstract: PURPOSE: A traffic measurement system and a traffic analyzing method are provided to offer a detailed traffic analysis result by measuring traffics from multiple points and correlatively analyzing the measured traffics. CONSTITUTION: A measuring unit(10) collects packets flowing on the Internet line, extracts and processes information required for traffic analysis, and provides it to an analyzing server(20). A packet collector(100) collects information on packets flowing on the Internet line, records accurate time received from a visual receiver(40) by each packet, and provides it to a flow generator(110). The flow generator(110) generates packets having the same address, the same protocol and the same port number as one flow. A transmitter(130) transmits information stored in a storing unit(120) to the analyzing server(20). The analyzing server(20) classifies traffics in detail, obtains traffic statistics, and generates a traffic report on a line. A data receiver(210) receives collected information from multiple measuring units(10) and provides it to a traffic analyzer(200). The traffic analyzer(200) analyzes the information and stores a result in a data storing unit(230) or provides it to a report output unit(220). The report output unit(220) processes the data to a report form and stores it in the data storing unit(230). A user interface(240) displays the report and various analysis data for a user.
Abstract translation: 目的：提供流量测量系统和流量分析方法，通过测量多个点的流量和相关分析测量流量来提供详细的流量分析结果。 规定：测量单元（10）收集在互联网上流动的数据包，提取和处理流量分析所需的信息，并将其提供给分析服务器（20）。 分组收集器（100）收集关于在因特网线路上流动的分组的信息，记录由每个分组从视觉接收机（40）接收到的精确时间，并将其提供给流发生器（110）。 流发生器（110）产生具有与一个流相同的地址，相同协议和相同端口号的分组。 发送器（130）将存储在存储单元（120）中的信息发送到分析服务器（20）。 分析服务器（20）详细分类流量，获取流量统计信息，并生成一条流量报告。 数据接收器（210）从多个测量单元（10）接收收集的信息，并将其提供给流量分析器（200）。 流量分析器（200）分析信息并将结果存储在数据存储单元（230）中或将其提供给报告输出单元（220）。 报告输出单元（220）将数据处理为报告表单并将其存储在数据存储单元（230）中。 用户界面（240）显示用户的报告和各种分析数据。
-
公开(公告)号：KR1020070059888A
公开(公告)日：2007-06-12
申请号：KR1020060064012
申请日：2006-07-07
Applicant: 한국전자통신연구원
CPC classification number: H04L9/0643
Abstract: A data hashing method using a similarity-based hashing algorithm, a data processing method and a system using the same are provided to perform retrieving, comparing, and sorting of data within time complexity with O(1) or O(n) by making the difference of the hashing values in inverse proportion to the similarity of the data. A data hashing method using a similarity-based hashing algorithm includes the steps of: inputting computation data(S110); and generating the same hashing value for the same data, and generating a hashing value for the computation data using the similarity-based hashing algorithm to generate the hashing value to make difference of the hashing values in inverse proportion to the similarity of data(S170). The hashing value generation step includes the steps of: calculating a fingerprint value from contents of the computation data(S130); changing a corresponding component value of an N-dimension hash vector in response to the fingerprint value according to a predetermined regulation(S150); determining whether the contents of the computation data are an end part(S160); and if the contents of the computation data are the end part, numerating the hash vector into the hashing value.
Abstract translation: 提供了使用基于相似度的散列算法的数据散列方法，数据处理方法和使用该散列算法的系统，以通过使O（1）或O（n）执行在时间复杂度内的数据的检索，比较和排序 哈希值的差异与数据的相似性成反比。 使用基于相似度的散列算法的数据散列方法包括以下步骤：输入计算数据（S110）; 并为相同的数据生成相同的散列值，并且使用基于相似度的散列算法来产生用于计算数据的散列值，以产生散列值以使哈希值的差与数据的相似性成反比（S170） 。 哈希值生成步骤包括以下步骤：从计算数据的内容计算指纹值（S130）; 响应于根据预定规定的指纹值来改变N维哈希向量的对应分量值（S150）; 确定所述计算数据的内容是否是结束部分（S160）; 并且如果计算数据的内容是结束部分，则将散列向量计算为哈希值。
-
9.发明授权국가 간 인터넷 트래픽 분석 결과를 행렬 형태로시각화하는 인터넷 트래픽 분석 시스템, 방법, 및 기록 매체 失效互联网流量分析系统，方法和存储媒体，按矩阵类型提供国家到国家互联网流量分析结果
公开(公告)号：KR100617308B1
公开(公告)日：2006-08-30
申请号：KR1020040017254
申请日：2004-03-15
Applicant: 한국전자통신연구원
IPC: H04L12/28
Abstract: 국가 간 인터넷 트래픽 분석 결과를 행렬 형태로 시각화하는 인터넷 트래픽 분석 시스템, 방법, 및 기록 매체가 개시된다. 상기 인터넷 트래픽 분석 시스템에서는, 트래픽 분석부가 국가 이름에 대응하는 일련번호를 추출하여 트래픽 분석 결과로 레코드를 생성하면, 행렬 리포트 생성부는 레코드 파일에서 레코드를 순차적으로 입력받아, 트래픽 통계치를 나타내는 행렬 형태의 자료 구조로서 리포트 출력 데이터를 생성한다.
-
公开(公告)号：KR100567320B1
公开(公告)日：2006-04-04
申请号：KR1020030085806
申请日：2003-11-28
Applicant: 한국전자통신연구원
IPC: H04L12/28
CPC classification number: H04L43/026
Abstract: 본 발명은 인터넷 트래픽 측정을 위한 플로우 생성 방법에 관한 것으로, 상기 방법은 가상 타이머를 사용하여 패킷의 수집시각과 플로우의 생성시간과의 종속성을 없애주고, 플로우의 생성 완료 처리 간격을 사용자의 요구에 따라 단위 시간만큼 건너뛰도록 제어한다. 그리고, 필요한 경우 단편화된 패킷 처리 방법을 선택적으로 사용하고, 플로우 생성 완료에 사용되는 타임아웃 메카니즘에 FIN 타임아웃을 추가하여 적용함으로써 인터넷 트래픽 측정의 정확도와 효율성을 높여준다.
-
-
-
-
-
-
-
-
-
Search Results
16
records
(took 0.017 seconds)
