Advanced Computer Networks (CSL724)

• Instructor:  Vinay Ribeiro


• Location: LH 416


• Time: Tue, Wed, Fri (5-5:50pm) - Slot K


• Course Email List: CSL724@courses.iitd.ac.in (will be active a week or so into the semester)


• Evaluation (tentative)

  Item	Marks
  Project	40
  Paper Reviews	20
  Class Paper Presentation	10
  Exams	30
  Attendance	5
  Total	105

  
  
  
• Attendance Policy:
• You get 5 bonus marks if you don't bunk any class. You lose 1 bonus mark for each class bunked. The definition of "bunk" is that a student missed class without a good reason (illness is a good reason).
• In addition, students getting less than 75% attendance will get one grade lower.



• Honour Code: Copying of any material verbatim from an Internet source (except for direct quotations or definitions) or from each other is not allowed for paper reviews and project reports. I plan to use software called Turnitin to check for such plagiarism. To have an idea about how seriously plagiarism is treated in the professional community see IEEE's Plagiarism FAQ. You will get negative marks for any paper review in which you have plagiarised. Ask for a deadline extension if you need more time instead of copying from various sources.
  
  
• Topics and Reading List (search Google Scholar to download papers)
  • Flow and Congestion Control: high-speed TCP, data center TCP
    1. Allman et al.: TCP Congestion Control, RFC 2581, 2001.
    2. Brakmo, Lawrence S., and Larry L. Peterson. "TCP Vegas: End to end congestion avoidance on a global Internet." Selected Areas in Communications, IEEE Journal on 13.8 (1995): 1465-1480.
    3. Floyd, Sally, and Van Jacobson. "Random early detection gateways for congestion avoidance." Networking, IEEE/ACM Transactions on 1.4 (1993): 397-413.
    4. Kumar, Anurag, D. Manjunath, and Joy Kuri. Communication networking: an analytical approach. Elsevier, 2004. (pg. 373-380: A Stochastic Model for a wide-area TCP connection)
    5. Kelly, Tom. "Scalable TCP: Improving performance in highspeed wide area networks." ACM SIGCOMM computer communication Review 33.2 (2003): 83-91.
    6. Floyd, Sally. "HighSpeed TCP for large congestion windows." (2003).
    7. Wei, David X., et al. "FAST TCP: motivation, architecture, algorithms, performance." IEEE/ACM Transactions on Networking (ToN) 14.6 (2006): 1246-1259.
    8. Ha, Sangtae, Injong Rhee, and Lisong Xu. "CUBIC: a new TCP-friendly high-speed TCP variant." ACM SIGOPS Operating Systems Review 42.5 (2008): 64-74.
    9. King, Ryan, Richard Baraniuk, and Rudolf Riedi. "TCP-Africa: an adaptive and fair rapid increase rule for scalable TCP." INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies. Proceedings IEEE. Vol. 3. IEEE, 2005.
    10. Vasudevan et al: Safe and effective fine-grained TCP retransmissions for datacenter communication, SIGCOMM 2009
    11. Alizadeh, Mohammad, Albert Greenberg, David A. Maltz, Jitendra Padhye, Parveen Patel, Balaji Prabhakar, Sudipta Sengupta, and Murari Sridharan. "Data center tcp (dctcp)." ACM SIGCOMM computer communication review 41, no. 4 (2011): 63-74.
  • Internet Architecture: original design, short-comings, suggested solutions
    1. Clark: The Design Philosophy of the DARPA Internet Protocols - SIGCOMM 1988
    2. Stoica et al.: Internet Indirection Infrastructure - SIGCOMM 2002
    3. Balakrishnan et al.: A Layered Naming Architecture for the Internet - SIGCOMM 2004
    4. Jacobson et al: Networking Named Content - CONEXT 2009
  • Modern Routing Technology: high-speed router design
    1. Mckeown: Fast Switched Backplane for a Gigabit Switched Router
    2. Mckkeown: The iSLIP Scheduling Algorithm for Input-Queued Switches - IEEE Trans. on Networking, 1999
    3. Chuang, Shag-Tse, Sundar Iyer, and Nick McKeown. "Practical algorithms for performance guarantees in buffered crossbars." In INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies. Proceedings IEEE, vol. 2, pp. 981-991. IEEE, 2005.
    
    
• Instructions on how to review a paper: Reading a computer science research paper.

• Class Project
• Form teams of 3 students each.
• The goal of the project will be to implement a TCP protocol that improves throughput over a WiFi connection. The first step of the project will be to implement and test TCP NewReno protocol to run over UDP. Your TCP protocol will be at the application layer but must take care of packet retransmissions, congestion control, flow control etc. After testing and debugging you will study its performance when running between two nodes by WiFi (either a direct WiFi link or through a WiFi Access Point). Then you will have to suggest small improvements in the protocol to improve throughput. You will need to study how the WiFi protocol works to be able to do this.


• Late submission policy for project reports -- take advance permission (with a good reason) for late submission, otherwise marks will be reduced by 50%


• Report style file: Please use the following Latex style file for project reports. This is a BTech project report style file. Modify the chapter/section headings appropriately for your project reports. Your project progress reports should be cumulative, containing material from the previous progress reports.



Item	Project Marks
Proposal (project team, objectives, tentative timeline to achieve various objectives (according to due dates for progress reports) -- 2 pages (due by 31/7/14)	1
Progress Report 1 [use Latex template given above] (due by 21/8/15)	4
Progress Report 2 (due by 24/9/15)	4
Progress Report 3 (due by 6/11/15)	6
Final Report and Demo (due by 24/11/15)	25
Total	40




• Class Paper Presentation
• Each project group will present one paper on Congestion Control from the past few ACM SIGCOMM conferences.
• List of papers for in-class presentations. Each project group should present one paper. Choose one paper and enter your choice (paper number: 1 - 11) in the online Doodle Poll . The selection is first-come-first-served.
  1. Improving Datacenter Performance and Robustness with Multipath TCP, Costin Raiciu (University College London & University Politehnica Bucharest); Sebastien Barre (Université catholique de Louvain); Christopher Pluntke, Adam Greenhalgh, Damon Wischik, Mark Handley (University College London), SIGCOMM 2011
  2. Deadline aware datacenter TCP (D2TCP), Balajee Vamanan et al., SIGCOMM 2012
  3. On-Chip Networks from a Networking Perspective: Congestion and Scalability in Many-Core Interconnects, George Nychis et al., SIGCOMM 2012.
  4. pFabric: Minimal Near-Optimal Datacenter Transport, Mohammad Alizadeh (Insieme Networks / Stanford University), Shuang Yang (Google / Stanford University), Milad Sharif, Sachin Katti, Nick McKeown, Balaji Prabhakar (Stanford University), and Scott Shenker (University of California, Berkeley / ICSI), SIGCOMM 2013.
  5. FCP: A Flexible Transport Framework for Accommodating Diversity, Dongsu Han (KAIST), Robert Grandl, Aditya Akella (University of Wisconsin-Madison), and Srinivasan Seshan (Carnegie Mellon University), SIGCOMM 2013
  6. TCP ex Machina: Computer-Generated Congestion Control, Keith Winstein and Hari Balakrishnan (MIT CSAIL), SIGCOMM 2013.
  7. CONGA: Distributed Congestion-Aware Load Balancing for Datacenters, Mohammad Alizadeh (Cisco); Tom Edsall (Cisco); Sarang Dharmapurikar (Cisco); Ramanan Vaidyanathan (Cisco); Kevin Chu (Cisco); Andy Fingerhut (Cisco); Vinh The Lam (Google); Francis Matus (Cisco); Rong Pan (Cisco); Navindra Yadav (Cisco); George Varghese (Microsoft), SIGCOMM 2014
  8. Friends, not Foes - Synthesizing Existing Transport Strategies for Data Center Networks, Ali Munir (Michigan State University); Ghufran Baig (LUMS); Syed Muhammad Irteza (LUMS); Ihsan Ayyub Qazi (LUMS); Alex Liu (Michigan State University); Fahad Dogar (Microsoft Research), SIGCOMM 2014
  9. Adaptive Congestion Control for Unpredictable Cellular Networks, Yasir Zaki (New York University Abu Dhabi), Thomas Potsch (University of Bremen), Jay Chen (New York University Abu Dhabi), Lakshminarayanan Subramanian (New York University), Carmelita Gorg (University of Bremen), SIGCOMM 2015
  10. Congestion Control for Large-Scale RDMA deployments, Yibo Zhu (University of California, Santa Barbara), Daniel Firestone (Microsoft), Chuanxiong Guo (Microsoft), Jitendra Padhye (Microsoft), Shachar Raindel (Mellanox), Ming Zhang (Microsoft), Yehonatan Liron (Mellanox), Haggai Eran (Mellanox), Mohamad Haj Yahia (Mellanox), Marina Lipshteyn (Microsoft), SIGCOMM 2015.
  11. TIMELY: RTT-based Congestion Control for the Datacenter, Radhika Mittal (University of California at Berkeley), Vinh The Lam (Google, Inc.), Nandita Dukkipati (Google, Inc.), Emily Blem (Google, Inc.), Hassan Wassel (Google, Inc.), Monia Ghobadi (Microsoft), Amin Vahdat (Google, Inc.), Yaogong Wang (Google, Inc.), David Wetherall (Google, Inc.), David Zats (Google, Inc.), SIGCOMM 2015
  12. Li, Li, Ke Xu, Dan Wang, Chunyi Peng, Qingyang Xiao, and Rashid Mijumbi. "A measurement study on TCP behaviors in HSPA+ networks on high-speed rails." In Computer Communications (INFOCOM), 2015 IEEE Conference on, pp. 2731-2739. IEEE, 2015.