CSC/ECE 772: Survivable Networks Spring, 2009 Rudra Dutta
CSC/ECE 772 About the course Subject of interest Background Copyright Rudra Dutta, NCSU, Spring, 2009
About the Course
About CSC/ECE 772 Understand network/service continuity principles, issues, constraints (descriptive: what's out there, work in progress) Evaluate alternatives, strengths, weaknesses (critical: what's wrong with... ?, how else can we do... ?) Ability to design and implement relevant protocols (skill-oriented: projects) Ability to perform original research (research-oriented: research presentations, literature report) Copyright Rudra Dutta, NCSU, Spring, 2009
Grading (tentative) 3 Programming Projects: 45% (grade will be based on code, report, performance results) Research Presentation: 15% (in-class presentation on 1-2 research papers) Literature Report: 20% (report on a research area within the field of network survivability) Final Exam: 20% (take-home exam; based on lectures, reading assignments) Class Participation (attend class and actively participate in discussion) Copyright Rudra Dutta, NCSU, Spring, 2009
Subject of Interest
Overview Continuity is an important goal in networks It has to be designed toward Involving better components, redundancy Design may be informed by optimization goals Not all things may be equally worth protecting There are various options in design approach Different viewpoints, some correlated Metrics needed to evaluate how good design is In evaluating, original design goals must be considered Restoring connectivity only Restoring capacity also - Bandwidth, switching, buffers Copyright Rudra Dutta, NCSU, Spring, 2009
Historical Development Equipment makers/carriers historically concerned with reliability In the past, attention to ensure certain levels of availability Transition equipment redundantly designed Network as a whole not considered No expectation of split-second recovery Failures above these levels -> service outage until: Repair was completed, or Manual effort at partial rerouting was completed Human involvement - additional errors Current goal of fault tolerance: Instantaneous recovery from most significant/frequent failures Eliminate human involvement Growth in use of communication services Fiber optic technology Competition - survivability a differentiating factor Copyright Rudra Dutta, NCSU, Spring, 2009
Background Topics Network provisioning and resource design Optimization, complexity System Reliability Copyright Rudra Dutta, NCSU, Spring, 2009
Course Modality Initially - lectures To cover background topics Fundamentals of survivability (Catch up on: “Research how-to” “Reading how-to” On instructor’s website) Then - Group readings - informal presentations Close - Individual topics to present formally Written report accompanying individual topics End - Take-home final Copyright Rudra Dutta, NCSU, Spring, 2009
Research Report Focus on a specific topic Research, standardization, implementation Must represent state-of-art Expected to read several papers (15-20) Deliver an authoritative, informative, synthesis Organize and classify prior work Describe important problems Summarize and critique solution approaches and results List open problems Conference-quality report Competent writing Suitable technical style 8 - 10 page limit Copyright Rudra Dutta, NCSU, Spring, 2009
Research Presentation One formal presentation by each student Interactive counterpart to research report Graded Preceded by short informal presentations throughout semester, not planned to be graded Must identify 2-3 key papers at least one week in advance for all to read Others allowed to submit questions beforehand Or bring to presentation Must educate and inform Enable audience to answer their own questions by the end Plan for about 45 minutes, plus discussion Copyright Rudra Dutta, NCSU, Spring, 2009
Informal Presentations Instructor will announce papers “Primary” readers announced Representative questions may be announced Papers will be jointly read in class Primary readers must be prepared to answer questions, make short explanation / presentation Bring slides or handouts only if you have to Preferably plan to use whiteboard only Several related papers will be read in a single lecture period Expect to discuss to find the similar and dissimilar features of the papers w.r.t. area, problem, methodology, solution Copyright Rudra Dutta, NCSU, Spring, 2009
Representative Sources R. K. Ahuja, T. L. Magnanti, J. B. Orlin, Network Flows. Prentice-Hall, 1993 R. Bhandari, Survivable Networks: Algorithms for Diverse Routing. Kluwer Academic Publishers, 1999 W. D. Grover, Mesh-Based Survivable Transport Networks: Options and Strategies for Optical, MPLS, SONET and ATM networking. Prentice Hall, 2003 K. Trivedi, Probability and Statistics with Reliability, Queueing and Computer Science Applications. Wiley Interscience, 2002 M. Pioro, D. Medhi, Routing, Flow, and Capacity Design in Communication and Computer Networks. Morgan Kauffmann, 2004 Copyright Rudra Dutta, NCSU, Spring, 2009