簡易檢索 / 詳目顯示

研究生: 薛旻欣
Xue, Min-Xin
論文名稱: 在階級制信任網路下最佳傳播排程之延伸成果
Extended Results in Optimal File Dissemination Scheduling under Hierarchical Trust Relationship
指導教授: 韓永楷
Hon, Wing-Kai
口試委員: 李哲榮
Lee, Che-Rung
蔡孟宗
Tsai, Meng-Tsung
學位類別: 碩士
Master
系所名稱: 電機資訊學院 - 資訊工程學系
Computer Science
論文出版年: 2020
畢業學年度: 108
語文別: 英文
論文頁數: 54
中文關鍵詞: 檔案傳輸信任關係階級制信任最佳排程NP困難傳輸能力能力較強大之節點
外文關鍵詞: File Dissemination, Trust Relationship, Hierarchical Trust, Optimal Scheduling, NP-Hardness, Dissemination Power, Superiorized Nodes
相關次數: 點閱:4下載:0
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 在通訊網路模型中,點對點 (P2P) 技術為傳統集中式網路的替代方案,可降低
    頻寬、縮短傳輸時間、增加網路規模的彈性、及提供更多自我錯誤修補的機制。
    然而點與點之間的誠實信任成為了新的隱憂,Ku 等作者 (2012) 最先研究在階
    級式信任網路中的檔案傳輸問題,提出了在信任階級模型為有根滿二元樹時,
    線性時間的最佳排程演算法。隨後,Tien 等作者 (2019) 將信任階級模型延伸
    為有根的二元有向無環圖,並提出了線性時間的最佳排程演算法。另一方面,
    Lin(2017) 提出當信任階級模型為最大外分支度 6 的有向無環圖時,最佳排程問
    題將為 NP 困難問題。
    本論文將信任階級模型進一步延伸為有根的二元一般圖,並提出可於線性時
    間內算出最佳排程的演算法。接著我們引入了有更強傳輸能力的超級節點,並
    且發現:在二元有向無環圖上,儘管只有兩種不同傳輸能力的節點,其最佳排
    程問題已為 NP 困難問題;但當信任階級模型僅為路徑 (一元樹) 時,則可在多
    項式時間內求得最佳排程。
    總的來說,階級式信任關係的結構、節點的最大外分支度、與多樣的傳播能
    力皆為影響檔案傳輸問題難度的重要參數。試以不同的參數組合來規範檔案傳
    輸問題,並將之界定為多項式時間內可解或 NP 困難題型,將成為未來一個有
    趣的研究方向。


    Peer-to-Peer (P2P) technology was proposed as an alternative solution to the
    traditional centralized communication network model, which has the advantages
    of bandwidth reduction, shorter dissemination time, better scalability, and faulttolerance. Yet, this brings new concerns, such as trust. Ku et al. (2012) first
    proposed the file dissemination problem under hierarchical trust relationship, and
    showed that when the trust hierarchy is a rooted full binary tree, an optimal file
    dissemination schedule can be computed in linear time. Tien et al. (2019) later
    extended the result so that the trust hierarchy can be any rooted directed acyclic
    graph (DAG) whose vertices have out-degree at most 2; the time to compute an
    optimal file dissemination schedule remains to be linear. In contrast, Lin (2017)
    showed that, if the maximum out-degree of a node is allowed to be up to 6,
    the problem of finding an optimal file dissemination schedule in a rooted DAG
    suddenly becomes NP-hard.
    In this thesis, we extend the file dissemination problem to consider trust hierarchy to be a general directed graph, with out-degree of every node at most 2,
    and show how to find an optimal file dissemination in linear time. Next, we introduce the concept of superior nodes, which have stronger dissemination power
    than others. We show that allowing superior nodes will make the file dissemination problem NP-Hard, even when there are only two types of dissemination
    powers, and the trust hierarchy is a rooted DAG whose nodes have out-degree
    at most 2. However, if we restrict the trust hierarchy to be a path, the problem
    becomes polynomial-time solvable.
    In summary, topology of the trust hierarchy, maximum degree of the nodes,
    and diversity of dissemination powers are some key parameters that control the
    hardness of the file dissemination problem. Finding the boundary classes (i.e., the
    combinations of these parameters that separate between polynomial-time solvable
    instances and NP-hard instances) will be an interesting direction to investigate.

    1 Introduction 2 2 Problem Description 4 2.1 Overview of the Variants . . . . . . . . . . . . . . . . . . . . . . . . 5 3 Preliminaries 7 3.1 The SplitTree Operation on a Binary Tree . . . . . . . . . . . . . 7 3.2 Generalization of the SplitTree Operation . . . . . . . . . . . . . 8 4 Standard OFDS on a Degree-At-Most-2 Graph 11 4.1 Standard OFDS on Reformable Graphs . . . . . . . . . . . . . . . . 13 5 OFDS on a Root-Superiorized Tree 16 5.1 The Greedy Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . 17 5.1.1 Correctness of the Algorithm (Proof of Claim 1) . . . . . . . 21 5.2 The Load-Balancing Algorithm . . . . . . . . . . . . . . . . . . . . 26 6 OFDS on a Path with Superiorized Nodes 30 6.1 Our Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 6.2 Proof of Correctness . . . . . . . . . . . . . . . . . . . . . . . . . . 34 6.2.1 Sufficient Condition for Domination . . . . . . . . . . . . . . 35 6.2.2 Optimality of the Greedy Choice . . . . . . . . . . . . . . . 36 7 NP-Hardness Results 38 7.1 OFDS on a General DAG . . . . . . . . . . . . . . . . . . . . . . . 38 7.2 OFDS on Out-Degree-At-Most-2 DAG with Two Types of Capacities 41 7.2.1 The “Only If” Part . . . . . . . . . . . . . . . . . . . . . . . 43 7.2.2 The “If” Part . . . . . . . . . . . . . . . . . . . . . . . . . . 43 7.3 Standard OFDS on Out-Degree-At-Most-3 DAG . . . . . . . . . . . 45 8 Conclusion and Future Work 50

    [1] B. Cohen, “Incentives Build Robustness in BitTorrent,” in Workshop on Economics of Peer-to-Peer Systems (P2PEcon), 2003.
    [2] M. Castro, P. Druschel, A.-M. Kermarrec, A. Nandi, A. Rowstron, and A. Singh, “SplitStream: High-Bandwidth Content Distribution in a Cooperative Environment,” in International Workshop on Peer-to-Peer Systems (IPTPS), pp. 292–303, 2003.
    [3] C. Gkantsidis, T. Karagiannis, P. Rodriguez, and M. Vojnovic, Planet Scale Software Updates,” in ACM SIGCOMM Conference, 2006.
    [4] T. T. Do, K. A. Hua, and M. A. Tantaoui, “P2VoD: Providing Fault Tolerant Video-on-Demand Streaming in Peer-to-Peer Environment,” in IEEE International Conference on Communications (ICC), pp. 1467–1472, 2004.
    [5] Y. Duan and J. Canny, “Scalable Secure Bidirectional Group Communication,” in IEEE International Conference on Computer Communications (INFOCOM), pp. 875–883, 2007.
    [6] A. M. Farley, “Broadcast Time in Communication Networks,” SIAM Journal on Applied Mathematics, vol. 39, no. 2, pp. 385–390, 1980.
    [7] C.-F. Ku, K.-H. Yang, and J.-M. Ho, “An Optimal Scheduling for File Dissemination under a Full Binary Tree of Trust Relationship,” in IEEE Global Telecommunications Conference (GLOBECOM), pp. 751–757, 2012.
    [8] Y.-P. Tien, W.-C. Lin, J.-M. Ho, and W.-K. Hon, “Optimal File Dissemination Scheduling Under a General Binary Tree of Trust Relationship,” in IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom), pp. 42–49, 2019.
    [9] W.-C. Lin, “Optimal File Dissemination on DAG under the Trust Model,” Master’s thesis, National Tsing Hua University, 2017.
    [10] T. H. Cormen, C. E. Leicerson, R. L. Rivest, and C. Stein. MIT Press and McGraw-Hill, 2nd edition ed., 2001.
    [11] A. J. Walker, “New Fast Method for Generating Discrete Random Numbers with Arbitrary Frequency Distributions,” Electronics Letters, vol. 10, no. 8, pp. 127–128, 1974.

    QR CODE