教育部國家講座專屬網站--張正尚教授

第26屆教育部國家講座-張正尚教授(YouTube影片)

張正尚教授獲頒INFOCOM終身成就獎

第20屆教育部國家講座-張正尚教授(YouTube影片)

張正尚教授獲第55屆學術獎

教學課程規劃重點

（一）、網路鏈結資料之建模與分析（Modeling and Analysis of Networked Data）

課程說明（Course Description）

Networked data that characterize the interactions between different entities are widely available nowadays. These include user-item networks (that log the purchases of items from users), social networks (that describe the relationship between two persons), and knowledge graphs (that use graph-structured data models to integrate data). Using networked data to gain insights on entities or predict future events becomes increasingly popular. In this course, we will introduce the mathematical tools needed to study dynamic networks. Students taking this course will have a comprehensive understanding of the mathematical theories for structural analysis in networks and the algorithms (and the associated programming tools) needed for analyzing networked data.

Reference books

• M. E. J. Newman, Networks: An Introduction, Oxford, 2010 (the key reference book of this course and buying this book is highly recommended even though the PDF file of the book can be downloaded from the book publisher)
• TED G. LEWIS, Network Science: Theory and Applications, WILEY, 2009

Acknowledgements

• Prof. Duan-Shin Lee for sharing the burden of teaching this course and preparing part of the slides
• Prof. Shuo-Yen Li for providing us his slides for his talk on "Network Coding" at NTHU
• Many TA's of this course in the last five years, including Dr. Li-Heng Liou, Dr. Ping-En Lu, Dr. Yi-Jheng Lin, Mr. Po-Lin Chen, and Mr. Kuo-Yu Liao

本課程綱要如下：

• Mathematics of Networks
  • Network representations(slides)
  • Classes of networks(slides)
  • Independent paths and cut sets(slides)
  • Network coding(Edited from the slides by Prof. Shuo-Yen Li)
  • Diffusion and random walks(slides)
  • Distributed averaging consensus(slides)
  • Majorization theory(slides)
  • Random gossiping(slides)
• Measures and Metrics
  • Centralities and ranking in social networks(slides)
  • Similarities(slides)
  • Clustering coefficients(slides)
  • Homophily(slides)
  • The general probabilistic framework for structural analysis(slides)
  • Structure of networks(slides)
• Network Algorithms
  • Basic concepts of algorithms(slides)
  • Graph partitioning(slides)
  • Community detection and clustering in large social networks(slides)
  • Link prediction and recommendation systems(to be added)
• Network Models
  • Random graphs(slides)
  • The configuration model(slides)
  • Preferential attachment and the duplication model(slides)
  • The small-world model(slides)
  • Exponential random graphs(slides)
• Network Dynamics
  • Percolation and emergence(slides)
  • Epidemic models(slides)
  • A time-dependent SIR model for COVID-19 with undetectable infected persons(paper)(slides)
  • Positively correlated samples save pooled testing costs(paper)(slides)
• Network Embedding
  • Laplacian eigenmaps(slides)
  • Graph neural networks(to be added)
  • Graph transformers(to be added)
  • Tracking and prediction(see the paper "A time-dependent SIR model for COVID-19 with undetectable infected persons")

Lectures in videos for the course "Network Science" (NTHU COM 530500)

1. 913_1_Overview(video)
2. 913_2_Introduction(video)
3. 916_Adjacency matrix(video)
4. 923_Citation network_Bipartite network(video)
5. 927_1_Undirected_directed network_Geodesic path(video)
6. 927_2_Network coding(video)
7. 930_Network code_Diffusion(video)
8. 1004_1_Laplacian(video)
9. 1004_2_Random walk(video)
10. 1007_Resistance network(video)
11. 1014_Majorization(video)
12. 1018_1_Birkhoff decomposition(video)
13. 1018_2_RandomGossip(video)
14. 1021_Centrality(video)
15. 1025_1_Between Centrality(video)
16. 1025_2_Clustering Coefficient(video)
17. 1028_Pearson Coefficient_Assortative Mixing(video)
18. 1101_1_Components_Small-world Effect(video)
19. 1101_2_Degree Distribution_Power Laws(video)
20. 1104_Top-Heavy_Assortative Mixing(video)
21. 1108_1_Runnup time and computational complexity(video)
22. 1108_2_Heaps(video)
23. 1108_3_Graph partitioning_Kernighan-Lin(video)
24. 1115_1_Spectral Modularity Maximization(video)
25. 1115_2_Random graph(video)
26. 1118_Small components(video)
27. 1125_Component size(video)
28. 1129_1_Conjugate distribution(video)
29. 1129_2_Configuration model(video)
30. 1202_Edge probability(video)
31. 1206_1_Generating function for degree.dist(video)
32. 1206_2_generating function for small component(video)
33. 1209_Average size of small component(video)
34. 1216_Percolation(video)
35. 1220_1_Non-uniform removal of vertices(video)
36. 1220_2_SI Model(video)
37. 1223_Final project(video)
38. 1223_SIR model(video)
39. 1213_1_Preferential attachment(video)
40. 1213_2_BA model(video)
41. 1227_1_Addition of extra edges(video)
42. 1227_2_Nonlinear preferential attachment(video)
43. 1230_Vertex copying models(video)
44. 0103_1_Clustering coefficient of ring(video)
45. 0103_2_Exponential random graphs(video)
46. 0117_Reciprocity model(video)

（二）、計算機網路概論（Introduction to Computer Networks）

課程說明（Course Description）

This course primarily introduces current computer networks, including both wired and wireless networks. It covers related network architecture, communication protocols, network interfaces, network traffic control, and related applications. Students can gain a deeper understanding of current computer networks through this course.

Reference books

• Larry L. Peterson and Bruce S. Davie, Computer Networks: A System Approach, 5th Ed. by Morgan Kaufmann Publishers, 2011 (the key reference book of this course and buying this book is highly recommended)
• Jim Kurose and Keith Ross, Computer Networking: A Top-Down Approach

Acknowledgements

• Prof. Yuh-Ren Tsai for sharing the burden of teaching this course and preparing part of the slides
• Prof. Jim Kurose for providing us his slides for his book
• Many TA's of this course in the last ten years

本課程綱要如下：

• Foundation
  • Foundation(slides)
• Getting Connected
  • Getting Connected(slides)
• Internetworking
  • Independent paths and cut sets(slides)
  • Basic internetworking(slides)
  • Evolution of Packet Switching Architectures(slides)
• Advanced Internetworking
  • BGP and IPv6(slides)
• End-to-End Protocols
  • UDP TCP and RTP(slides)
• Congestion Control and Resource Allocation
  • TCP congestion control and Quality of Service(slides)
• End-to-End Data
  • Presentation formatting and multimedia data(slides)
• Network Security
  • Security attacks and principles of ciphers(slides)
• Applications
  • WWW and DNS(slides)
  • Peer-to-Peer Networks(slides)
  • Data Center Network (DCN)(slides)
  • Introduction to Software Defined Networking (SDN)(slides)

Lectures in videos for the course "Introduction to Computer Networks" (NTHU EECS 302001)

1. 0214_1 Foundation(video)
2. 0214_2 Foundation(video)
3. 0217 Foundation(video)
4. 0221_1 Foundation(video)
5. 0221_2 Foundation(video)
6. 0221_3 Foundation(video)
7. 0224 Foundation(video)
8. 0303 Getting Connected(video)
9. 0307_1 Getting Connected(video)
10. 0307_2 Getting Connected(video)
11. 0310 Getting Connected(video)
12. 0314_1 Getting Connected(video)
13. 0314_2 Getting Connected(video)
14. 0317 Getting Connected(video)
15. 0321_1 Getting Connected(video)
16. 0321_2 Getting Connected(video)
17. 0324 Getting Connected + Switching and Forwarding(video)
18. 0328_1 Switching and Forwarding(video)
19. 0328_2 Switching and Forwarding(video)
20. 0407 Switching and Forwarding + Basic internetworking(video)
21. 0411_1 Basic internetworking(video)
22. 0411_2 Basic internetworking(video)
23. 0414 Basic internetworking(video)
24. 0418_1 Basic internetworking + BGP and IPv6(video)
25. 0418_2 BGP and IPv6(video)
26. 0421 BGP and IPv6 + UDP TCP and RTP(video)
27. 0428 UDP TCP and RTP(video)
28. 0502_1 UDP TCP and RTP(video)
29. 0502_2 UDP TCP and RTP + TCP congestion control and Quality of Service(video)
30. 0505 TCP congestion control and Quality of Service(video)
31. 0509_1 TCP congestion control and Quality of Service(video)
32. 0509_2 TCP congestion control and Quality of Service(video)
33. 0512 TCP congestion control and Quality of Service(video)
34. 0516_1 Presentation formatting and multimedia data(video)
35. 0516_2 Presentation formatting and multimedia data(video)
36. 0519 Presentation formatting and multimedia data(video)
37. 0523_1 Security attacks and principles of ciphers(video)
38. 0523_2 Security attacks and principles of ciphers(video)
39. 0526 Security attacks and principles of ciphers(video)
40. 0530_1 WWW and DNS(video)
41. 0530_2 WWW and DNS(video)
42. 0602 WWW and DNS(video)
43. 0606_1 WWW and DNS + Peer-to-Peer Networks(video)
44. 0606_2 Peer-to-Peer Networks(video)
45. Data Center Network (DCN) + Introduction to Software Defined Networking (SDN)(video)

研究重點及方向(目前研究簡介)(YouTube影片)

一、網路科學、大數據分析及其應用

2005年美國國家學院科學研究委員會依據科學方法將「網路科學」定義為「有組織的網路知識」。這個需要整合各種不同領域的知識：包含網際網路、電力網路、社群網路、實體網路以及生物網路。對於網路科學，主要的數學工具是動態圖形研究，用來處理巨量資料（Big Data）。這些年來已可常見在Nature和Science的文章討論此類問題，諸如生物心臟的同步跳動、基因調控網路、合作與競爭的演化、動態社群網路的社群偵測、及兩黨政治形成的數學模型等。然將網路科學和大數據分析的應用發揮得淋漓盡致的莫非是Google、Meta（facebook）和Apple這些大公司，他們如白雪公主童話故事中的魔鏡般的不時監控我們的一舉一動，回答我們的問題，並進一步預測我們的下一個動作。顯而易見的是誰掌握這些魔鏡，誰將在未來的發展佔盡優勢。 　　網路科學和大數據分析因包含相當廣闊之領域，基本的研究問題包含中心度分析（centrality analysis），社群偵測（community detection）和資料聚類（data clustering），及網路嵌入（network embedding）與應用。張教授近十年來的研究，提出共同及廣義之機率架構，稱為取樣網路，可用於有向或無向網路中的中心度分析，社群偵測及網路嵌入，已有一系列的論文發表。

2019年年底新型冠狀病毒（COVID-19）爆發，陸續在各個國家大肆傳染，造成世界的大災難。張教授及其研究團隊運用網路科學方面的知識，著手研究COVID-19傳染的狀況，並預測未來COVID-19的發展。在（代表著作1 (paper)）提出了一個time-dependent susceptible-infected-recovered (SIR) 數學的模型來追蹤與預測COVID-19的感染人數。並探討防疫措施，包含保持社交距離、限制與會人數，以及宣導基本的衛教知識等，以遏止COVID-19的傳播。在（代表著作2(paper)）進一步研究COVID-19群體測試（pooled testing）的方法。此論文比較文獻中各種池化測試矩陣表現，並且運用多通道會合問題中通道跳頻序列的組合性質，設計出一種池化測試矩陣，可以大幅降低檢測成本。在（2016年以後發表的參考著作1(paper)）提出來一個在發生社區感染的情況下，如何運用社群偵測演算法來節省池化測試成本的方法。目前研究團隊進行中的研究，包括（一）探討台灣抑制 COVID-19 Delta變異株盛行原因，（二）由已知的案例中追蹤第一個病例的來源。未來研究方向將繼續做關於COVID-19的研究，希望這些研究成果能對對控制疫情做出一點貢獻。

二、6G通訊系統及其相關問題之研究

由於現今的行動服務逐漸深入生活中的每個面向，從工業、教育、醫療、交通與娛樂等，均因新科技帶來完全不同的互動，如智慧醫療、智慧城市、人工智慧物聯網（AIoT）、自駕車、無人機、穿戴裝置、工廠自動化，以及邊緣運算。第四代行動通訊技術（4G）已無法完全提供完善的服務因應各種新興應用與產業，因此第五代行動通訊技術（5G），甚至是第六代行動通訊技術（6G）的推出勢在必行。國際電信聯盟（ITU）公布5G的正式名稱為IMT-2020，同時確立三大應用場景，分別為增強型行動寬頻（enhanced Mobile BroadBand, eMBB）、超可靠低延遲通訊（Ultra-Reliable Low-Latency Communications, URLLC），以及大量連結物聯網（massive Machine-Type Communications, mMTC）。然而目前建置的5G通訊系統中，並不包含URLLC和mMTC，因此B5G/ 6G已成為近期世界各國的重點發展項目，也是我國的重點投入項目之一。

URLLC設計主要是為了因應關鍵任務應用，例如智慧型運輸與交通系統、智能電網、VR、工業自動化與遠程醫療。以上應用案例皆需要超高可靠度與超低延遲，才能有效避免意外事故的發生。因此嚴苛的通訊需求如錯誤率至少低於10^(-5)、網路延遲低於1毫秒以下，以及通訊時的保密性和安全性是絕對必須的。在（2016年以後發表的參考著作6(paper)）張教授的研究團隊提出兩個適用於多通道無線網路多重接取之通訊協定，此協定滿足以下性質：（1）異步時脈：所連接設備的時脈不需要同步。（2）無允諾傳輸：為了避免往返延遲，設備無需接收方的授權即可進行傳輸。（3）支持大量已連接設備：已連接設備的數量可能遠遠大於通道的數量。（4）能夠保證流通量和延遲：當活躍設備的總數不超過某個閾值參數時，成功傳輸的活躍設備的最大延遲為一個常數上界。在（代表著作4(paper)）中，針對編碼多重存取，提出波以松接收器的抽象接收器模型，建立完整機率分析的理論架構，對於B5G/6G網路中不同使用者的應用場景，提供差異化服務。進一步在（代表著作5(paper)）中，基於網路計算工具，提出抽象的阿羅哈接收器，結合波以松接收器的理論架構，在媒體存取控制層建立一個通用的理論架構，不僅統一過去十年來在編碼多重存取分析的論文成果，更進一步對發展中的power-domain non-orthogonal multiple access（NOMA）提供分析的結果。

研究方向將繼續做關於URLLC及mMTC的編碼多重接取技術的研究。希望能建構在我們提出的架構上，進行下列數個後續研究：（一）擴展到多用戶以及多接收器、（二）穩定性分析、（三）數值研究B5G/6G系統可能的應用案例、（四）透過波以松及阿羅哈接收器對物理層的通道進行建模及分析、（五）最佳化系統參數以更好地設計未來的B5G/6G系統、（六）探索物理層中不平等錯誤保護（UEP）問題的可能應用，以及（七）與編碼壓縮感測的相關研究。