Seminar - Future Internet Protocols: Design Principles & Deployment (2017)

Dr. Vaibhav Bajpai, Ljubica Kärkkäinen

There is a constant demand for the Internet to function better, not only in terms of performance, but also in terms of being more secure, scalable and highly reliable. Towards this end, new protocol design proposals are proposed across multiple layers of networking abstraction to allow the Internet to evolve to sustain these growing needs. In this seminar, we will explore such recent efforts in protocol designs (network layer and above) and the qualitative benefits these proposals bring on top of the existing Internet architecture. We will also study the state of their deployment, the roadblocks that may hinder their adoption on the Internet and real-world evaluations that quantify the improvements with existing protocols.

(Recommended) requirements:

The participants should be already prepared by an undergraduate-level course on computer networks. Familiarity with networking tools used for performance evaluation may be beneficial.

Relevant Conferences: 

• IEEE/ACM Transactions on Networking
• ACM SIGCOMM Computer Communication Review
• ACM SIGCOMM
• IEEE INFOCOM
• USENIX Networked System Design and Implementation
• Internet Engineering Task Force
• Internet Protocol Journal

Time and location:

Thursdays (14:00 - 16:00) in Room 01.07.023.

Review Form:

The review form is available here.

Schedule:  

02/Nov, Session Chair: Humaira	The QUIC Transport Protocol: Design and Internet-Scale Deployment Presenter: Prateek Bagrecha Opponent: Ursula Rieger Taking a Long Look at QUIC: An Approach for Rigorous Evaluation of Rapidly Evolving Transport Protocols Presenter: Thomas Paul Opponent: Richard Otto Reducing Web Latency: the Virtue of Gentle Aggression Presenter: Tizian Sarre Opponent: Humaira
09/Nov,  Session Chair: Thomas	Measuring DANE TLSA Deployment Presenter: Atakan Yenel Opponent: Thomas Paul A Longitudinal, End-to-End View of the DNSSEC Ecosystem Presenter: Ursula Rieger Opponent: Andrei Vlad Are We There Yet? On RPKI’s Deployment and Security Presenter: Richard Otto Opponent: Fabian Helfert
16/Nov,  Session Chair: Atakan	Is the Web HTTP/2 Yet? Presenter: Andrei Vlad  Opponent: Atakan Yenel The Web, the Users, and the MOS: Influence of HTTP/2 on User Experience Presenter: Krishna Opponent: - Measuring HTTPS Adoption on the Web Presenter: Fabian Helfert Opponent: Prateek Bagrecha State of IPv6 Deployment 2017 Presenter: Tobias Zappe Opponent: Petr Romanov
23/Nov,  Session Chair: Ursula	BBR: congestion-based congestion control Presenter: Humaira Opponent: Tobias Zappe TCP ex Machina: Computer-Generated Congestion Control Presenter: Robert Brünig Opponent: Tizian Sarre How Hard Can It Be? Designing and Implementing a Deployable Multipath TCP Presenter: Petr Romanov Opponent: Akash Manjunath Design, Implementation and Evaluation of Congestion Control for Multipath TCP Presenter: Akash Manjunath Opponent: Robert Brünig

Pre-course Meeting:

Wednesday: 05.07.2017 (16:00 - 17:00) in Room 01.07.023. Slides

Topics of Interest:

The topics of interest include (but are not limited) to the following:

Applications

WebRTC	Real-time communications for the web
HTTP/2	Hypertext Transfer Protocol Version 2 (HTTP/2) HPACK: Header Compression for HTTP/2 HTTP2 explained, blog by Daniel Stenberg Is the Web HTTP/2 Yet? The Web, the Users, and the MOS: Influence of HTTP/2 on User Experience

 

Security

 

TLS 1.3	The Transport Layer Security (TLS) Protocol Version 1.3 Transport Layer Security Introducing TLS 1.3, by Nick Sullivan
Email Security (DKIM, SPF, DMARC)	Neither Snow Nor Rain Nor MITM...: An Empirical Analysis of Email Delivery Security DomainKeys Identified Mail (DKIM) Signatures Sender Policy Framework (SPF) for Authorizing Use of Domains in Email, Version 1 Domain-based Message Authentication, Reporting, and Conformance (DMARC)
DNSSEC and DANE	A Longitudinal, End-to-End View of the DNSSEC Ecosystem The Performance Impact of Elliptic Curve Cryptography on DNSSEC Validation DNSSEC meets real world: dealing with unreachability caused by fragmentation Measuring DANE TLSA Deployment DNSSEC and its potential for DDoS attacks: a comprehensive measurement study Making the Case for Elliptic Curves in DNSSEC Domain Name System Security Extensions

 Transports 

QUIC	The QUIC Transport Protocol: Design and Internet-Scale Deployment   Innovating Transport with QUIC: Design Approaches and Research Challenges QUIC: A UDP-Based Multiplexed and Secure Transport Using Transport Layer Security (TLS) to Secure QUIC  QUIC Loss Detection and Congestion Control Hypertext Transfer Protocol (HTTP) over QUIC
MPTCP	TCP Extensions for Multipath Operation with Multiple Addresses How Hard Can It Be? Designing and Implementing a Deployable Multipath TCP Multipath TCP Multipath TCP: An annotated bibliography
SCTP	Stream Control Transmission Protocol SCTP: state of the art in research, products, and technical challenges SCTP: new transport protocol for TCP/IP SCTP: a proposed standard for robust Internet data transport

 

Congestion Control and Loss Recovery 

BBR	BBR: congestion-based congestion control BBR, the new kid on the TCP block, by Geoff Huston
TCP RACK, TLP and SACK	RACK: a time-based fast loss detection algorithm for TCP Tail Loss Probe (TLP): An Algorithm for Fast Recovery of Tail Losses A Conservative Loss Recovery Algorithm Based on Selective Acknowledgment (SACK) for TCP

 

Networks 

IPv6	Internet Protocol, Version 6 (IPv6) Specification Measuring IPv6 adoption State of IPv6 Deployment 2017
LISP	The Locator/ID Separation Protocol (LISP) Designing a Deployable Internet: The Locator/Identifier Separation Protocol On the cost of caching locator/ID mappings The Internet Protocol Journal, Volume 11, No. 1, March 2008.
TRILL	Routing Bridges (RBridges): Base Protocol Specification The Internet Protocol Journal, Volume 14, No. 3, September 2011. Rbridges: transparent routing

 

Learning outcomes (study goals):

The topics covered in this seminar revolve around novel network protocols and architectures. The papers will give students the technical knowledge and understanding on the latest advancements in the field of emerging networking protocol solutions. The participants will also learn how to critically read and discuss research papers. This will be achieved by reviewing papers individually, and actively participating in group discussions during the seminar presentations. Students will also have the opportunity to advance their soft skills through presentation and session moderation. Participants will learn how to act as a chair of a session. Presentations will involve learning to not only stay within time limits but also to appreciate the Q/A session.

 Detailed goals of the seminar

Specifically, after the seminar, the student should be able to: -

• Understand the need for new Internet architectures, protocols considered.
• Explain the technical detail[s] of the discussed protocols and frameworks.
• Discuss design principles and the performance of the presented solutions.
• Understand the importance of peer review and be able to conduct such process independently.
• Present research in a concise way and within the allotted time (conference-style settings)

 Teaching and learning methods: 

1. A written paper review before the presentation (20% grade)
2. Weekly presentations during the semester (70% grade)
3. Group discussions (10% grade)

Each participant presents 1 paper during the seminar. See for example student presentations. To ensure everybody has read the paper, participants that witness the presentation are required to hand in a review of the paper (review form will be provided) being presented. The answers to the review forms should be brief and concise. An email with filled out review form to us (see below for contact details) would do. Internet Measurement Conference made reviews for accepted papers public for the 2012 and 2013 programmes which can be used to get an impression of a paper review. Paper allocations will be done on a best-effort basis, based on preferences (favorite 2-3 topics) solicited over email during the semester. A topic will be randomly assigned if no preference is sent. The first seminar course slot (19/Oct) will be used to set the agenda for the seminar. 

Further Reading

• Srinivasan Keshav. How to read a paper. SIGCOMM Computer Communication Review, July 2007
• William G. Griswold. How to Read an Engineering Research Paper (Last accessed: June 2017)
• Graham Cormode. How NOT to review a paper. ACM SIGMOD, March 2009
• Alan J. Smith. The Task of the Referee. IEEE Computer, April 1990 

Contact:

• Dr. Vaibhav Bajpai <bajpaiv@in.tum.de>
• Ljubica Kärkkäinen <kaerkkal@in.tum.de>