Browsing by Author "Lin Ying-Dar"

Now showing 1 - 20 of 46

ATM網路上的虛擬教室系統應用、管理、評估---總計畫
(1996)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學研究所
ATM網路提供動態頻寬分配及效益保證式的服務,這使得多媒體網路應用時機成熟,本群體計畫配合交大 ATM 網路實驗平臺,將開發 ATM 虛擬教室應用及ATM 網路系統,並進行系統效益評估,作為應用控制模組及網管系統管理程序調整的根據.虛擬教室應用的核心技術問題在於發展動態一對多的連線傳送協定,使其能根據資料源位置的改變做動態頻寬分配調整,發揮頻寬最佳使用效率,應用子計畫將使用盡力式及保證式 ATM 服務,分別建構這個多點連線協定 ,管理子計畫的網管系統將用以監督 ATM 網路及其上的虛擬教室應用及其他應用的進行狀況, 並在適當時機調整網路,評估子計畫將分析多媒體應用在盡力式及保證式ATM 服務下的表現, 並對目前在此兩種服務下具階層式及主從式相關性之資料源的控制及管理方法做分析、模擬及實測以求得最佳的控制參數及管理程序,這些結果可用在應用及管理系統的設計.本群體計畫最後階段將發展延伸 ATM 應用至國家資訊基礎網路及有線電視網路的軟硬體架構.
DECISION METHOD CONSIDERING TIME AND POWER CONSUMPTION FOR OFFLOADING COMPUTATION AND COMPUTING SYSTEM
(2013-08-08)
Lin Ying-Dar
;
Huang Ting-Jun
;
Lai Yuan-Cheng
;
Chu Tsung-Hsien
A decision method considering time and power consumption for offloading computations is provided. The method includes: obtaining a computing mission; obtaining a plurality of static effect factors from an effect factor table; collecting a plurality of dynamic effect factors. The method also includes: generating power consumptions and computing time that each of the computing unit and each of the server consumes according to the dynamic effect factors and the static effect factors; calculating cost values of executing the computing mission according to the power consumptions and the computing time; determining a target unit to execute the computing mission according to the cost values, in which the target unit is a computing unit or a server; executing the computing mission by the target unit. Therefore, the power consumptions and the computing time are simultaneously considered, and computations are adequately allocated to one of the computing unit and the servers.
Machine-implemented method and system for determining whether a to-be-analyzed software is a known malware or a variant of the known malware
(2013-08-06)
Chiang Yi-Ta
;
Lin Ying-Dar
;
Wu Yu-Sung
;
Lai Yuan-Cheng
A machine-implemented method for determining whether a to-be-analyzed software is a known malware or a variant of the known malware includes the steps of: (A) configuring a processor to execute the to-be-analyzed software, and obtain a to-be-analyzed system call sequence that corresponds to the to-be-analyzed software with reference to a plurality of system calls made in sequence as a result of executing the to-be-analyzed software; (B) configuring the processor to determine a degree of similarity between the to-be-analyzed system call sequence and a reference system call sequence that corresponds to the known malware; and (C) configuring the processor to determine that the to-be-analyzed software is neither the known malware nor a variant of the known malware when the degree of similarity determined in step (B) is not greater than a predefined similarity threshold value.
MACHINE-IMPLEMENTED METHOD AND SYSTEM FOR DETERMINING WHETHER A TO-BE-ANALYZED SOFTWARE IS A KNOWN MALWARE OR A VARIANT OF THE KNOWN MALWARE
(2012-05-17)
Chiang Yi-Ta
;
Lin Ying-Dar
;
Wu Yu-Sung
;
Lai Yuan-Cheng
A machine-implemented method for determining whether a to-be-analyzed software is a known malware or a variant of the known malware includes the steps of: (A) configuring a processor to execute the to-be-analyzed software, and obtain a to-be-analyzed system call sequence that corresponds to the to-be-analyzed software with reference to a plurality of system calls made in sequence as a result of executing the to-be-analyzed software; (B) configuring the processor to determine a degree of similarity between the to-be-analyzed system call sequence and a reference system call sequence that corresponds to the known malware; and (C) configuring the processor to determine that the to-be-analyzed software is neither the known malware nor a variant of the known malware when the degree of similarity determined in step (B) is not greater than a predefined similarity threshold value.
Method and system for offloading a computation to other computing units in the same device or to a server on a network based upon a computing time and a power consumption of each of the computing units and the server
(2015-01-27)
Lin Ying-Dar
;
Huang Ting-Jun
;
Lai Yuan-Cheng
;
Chu Tsung-Hsien
A decision method considering time and power consumption for offloading computations is provided. The method includes: obtaining a computing mission; obtaining a plurality of static effect factors from an effect factor table; collecting a plurality of dynamic effect factors. The method also includes: generating power consumptions and computing time that each of the computing unit and each of the server consumes according to the dynamic effect factors and the static effect factors; calculating cost values of executing the computing mission according to the power consumptions and the computing time; determining a target unit to execute the computing mission according to the cost values, in which the target unit is a computing unit or a server; executing the computing mission by the target unit. Therefore, the power consumptions and the computing time are simultaneously considered, and computations are adequately allocated to one of the computing unit and the servers.
MIB驅動的ATM網路管理
(1995)
羅濟群
;
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊管理研究所
ATM網路的UNI MIB正在進行標準化的工作,並已有初步架構,本計畫將根據該MIB的架構研究ATM 網路的效益管理及錯誤管理方法.在效益管理方面,我們將著重於VP的Reconfiguration來避免或補救網路擁塞的現象,本計畫書將VP的調整分成四類:Rerouting, Resizing, Establishment及Release,並說明各種類的驅動條件,本研究將設計各種類的詳細協定即以模擬程式評估其成效.在錯誤管理方面,我們將研究如何經由Alarm correlation的方法由許多Alarm中診斷網路的問題,本計畫書將ATM網路問題分成四類:Congestion, CAC Failure, UPC failure 及Link/node failure,並提出了Heuristic alarm correlation 的規則,本研究將根據這些規則設計詳細的演算法及以模擬程式評估其可靠度.本計畫的成果將包括:(1)VP reconfiguration protocol for ATM performance management及(2)Alarm correlation algorithm for ATM fault management.
POWER-SAVING DATA SCHEDULING SYSTEM IN LTE AND METHOD THEREOF
(2015-05-14)
Lin Ying-Dar
;
Kuo Yu-Hsien
;
Tung Li-Ping
;
Lai Yuan-Cheng
A power-saving data scheduling system for LTE and a method thereof are provided. The power-saving data scheduling system includes a period decision module for selecting a cycle having a shortest delay time as a discontinuous reception (DRX) cycle for each of user from quality of service (QoS) requirements related to network services and channel conditions, and a start offset decision module for calculating the number of users in each of periods of the DRX cycle. In addition, a DRX-aware scheduling module is provided for extending the on period by increasing priority of the user and resetting an inactivity timer if a period required by the user's load is shorter than an off period. Optimal DRX parameters and DRX-aware scheduling are used to resolve the existing issues that the channel conditions, system load and QoS are not taken into consideration when the DRX parameters are determined. In addition, the DRX parameters of a user may be changed dynamically without updating all other users' parameters so as to avoid consuming unnecessary power. Moreover, a power-saving data scheduling method for LTE is also provided.
SYSTEM FOR REAL TRAFFIC REPLAY OVER WIRELESS NETWORKS
(2013-12-05)
Lin Ying-Dar
;
Ku Chia-Yu
;
Li Pei-Hsuan
;
Lai Yuan-Cheng
A real traffic replay over wireless networks with environment emulation and an event-driven automata-synchronized replay method are disclosed. The system includes a first traffic capture module configured to capture packet traces and environment effects in a real environment, a first event transform module configured to transform the packet traces into a sequence of events, and a reproduce module configured to receive the environment effects from the first traffic capture module and the sequence of events from the first event transform module such as to control a replay of the environment effects in the real environment and the sequence of events coordinately by means of automata.
System for real traffic replay over wireless networks
(2015-01-20)
Lin Ying-Dar
;
Ku Chia-Yu
;
Li Pei-Hsuan
;
Lai Yuan-Cheng
A real traffic replay over wireless networks with environment emulation and an event-driven automata-synchronized replay method are disclosed. The system includes a first traffic capture module configured to capture packet traces and environment effects in a real environment, a first event transform module configured to transform the packet traces into a sequence of events, and a reproduce module configured to receive the environment effects from the first traffic capture module and the sequence of events from the first event transform module such as to control a replay of the environment effects in the real environment and the sequence of events coordinately by means of automata.
以Linux為基礎的網路安全與頻寬管理閘道器之實作與研究(I)
(2002)
林盈達
;
Lin Ying-Dar
;
交通大學資訊科學系
以Linux為基礎的網路安全與頻寬管理閘道器之實作與研究(II)
(2003)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學學系
以內容為基礎之網路安全---子計畫三：網路內容分類的系統架構---設計、實作與評估(I)
(2004)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學學系
如何在大量自由格式的文字中，快速地做特徵比對，以過濾不必要資訊，是提升內容網路安全服務(如IDS, Antivirus, Content Filtering with URLs and keywords, Antispam) 之運作速度的主要瓶頸點。本計劃以系統架構的角度分四個步驟，解決這個問題。四個步驟包括： 1. 整合與白黑箱測試 – 安裝這些proxies 於同一平台，並進行測試指出耗時程式段落。這裏的黑箱測試指的是從系統的外部量測效能及比對準度。白箱測試則是量測程式中各段落的執行時間。 2. 4-in-1 proxy 的整合 – 減少封包重複處理或多次傳遞於kernel/user space 間的封包流程。一方面每個內容處理步驟只做一次，另一方面儘可能減少kernel/user space 間的封包傳遞。 3. one-time 特徵比對 – 集四個proxies 之特徵比對工作於同一段程式，也就是用單一類的演算法來進行各種類型的比對工作。一方面減化比對的演算法，另一方面是後續的硬體實作更容易。 4. ASIC 硬體加速 – 以硬體加速特徵比對，以提高整體的系統效能。除了設計硬體的部分以外，我們也將與系統其餘的軟體部分做協同運作及測試。第一年進行步驟1 與2，第二、三年分別進行步驟3 與4。每年將產生一個可運作的 prototype gateway，顯現加速的效果。此外也將產出相關論文及專利。
以內容為基礎之網路安全---子計畫三：網路內容分類的系統架構---設計、實作與評估(III)
(2006)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊工程學系(所)
以內容為基礎之網路安全---子計畫三：網路內容分類的系統架構、設計、實作與評估(II)
(2005)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學學系(所)
以多層到達程序為基礎的寬頻網路控制
(1995)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學研究所
基於使用者行為的智慧型手機硬體元件耗電模型化、預測及調整
(2016)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊工程學系（所）
基於使用者行為的智慧型手機硬體元件耗電模型化、預測及調整
(2015)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊工程學系（所）
子計劃二：網路化創造性學習環境之可行性研究 (III)
(2000)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學學系
寬頻網際網路之服務品質確保---子計畫五：整合服務與差別服務網路之相容運作技術
(2000)
林盈達
;
Lin Ying-Dar
;
國立交通大學資訊科學學系
寬頻網際網路服務品質保證---子計畫VI：整合服務與差別服務網路之相容運作技術(II)
(2000)
林盈達
;
Lin Ying-Dar
;
交通大學資訊科學系