(Under Construction)

CSCI 7143-001: Secure Sensor Networks

Fall 2004

Department of Computer Science, University of Colorado at Boulder

Abstract:  Sensor networking envisions a ubiquitous computing future in which scattered smart sensors combine to form dynamic, ad hoc, low power, and self-configurable wireless networks capable of monitoring their external environments and relaying their data to a smart spaces infrastructure.  These sensor networks will operate in a variety of environments, including home, office, and military, and will assist in such activities as locating and identifying people indoors as well as tracking the motion of vehicles outdoors.

In this graduate seminar, we will explore the state-of-the-art in secure sensor networking research by reading, presenting and discussing selected papers, and  design innovative research projects in sensor networks, covering such topics as: encryption, authentication in sensor networks secure routing secure aggregation secure operating system design for smart sensor nodes secure protocols <>secure time synchronization and sensor calibration for sensor networks if time permits, build sample applications using MANTIS OS on sensor nodes, (see right, and http://mantis.cs.colorado.edu)

MANTIS sensor nodes (nymphs) with GPS

We will study how the unique constraints of sensor networks - wireless, low power, small form factor, limited memory/CPU - affect the design of networking protocols as well as the design of the sensor nodes themselves.   We will construct real sensor networks using actual leading edge sensor nodes and develop hands-on familiarity with a variety of physical sensors, including thermal, magnetic, acoustic, and motion sensors.  A thorough understanding of TCP/IP networking, wireless networking, and operating system concepts is a prerequisite.

Grading

Students will be asked to:

• design and build an innovative research project for presentation at the end of the semester.  Students should organize themselves into research project teams.  A final written report will be submitted.

• for each research paper in the assigned reading list:

  • participate in the class discussion of each paper

  • provide written summaries of each paper before class

  • volunteer to present in class certain of the papers on the reading list, on a rotating basis

• develop software programs on sensor nodes

25% Final project presentation to class
25% Final project report
20% Class presentation of papers
15% Paper reviews and participation in discussion
15% Programming projects on sensors

Plagiarism policy.
Religious observances policy.
Disability policy.

Research Projects

Students will be asked to build/create an innovative research project for presentation at the end of the semester.  Students will form teams of 2-3 members and work on projects as a team.  Teams and projects will be decided according to the timeline below.  Read ahead to topics that you'd be interested to do a project in.  A list of suggested project ideas will be available later.  Students are welcome to formulate their own project ideas.

Each team will be required to present their project to the class at the end of the course,
A final project report written in the style of a conference paper will be handed in following the presentation.  If the final project is sufficiently innovative, and is accepted as a paper at a conference, then I'll pay for your trip to that conference (Past projects have led to publications at conferences in Sweden and Monterrey, California).

Paper Reviews and Presentations

Students are required to read, present, and discuss graduate-level research papers throughout the semester.  An average of 2-3 papers per week will be read.  Written reviews of each paper to be discussed in class are due prior to the start of that class, and should be emailed to the instructor rhan@cs.colorado.edu.  Late reviews will not be accepted.   For each paper, students should write a review answering each of the following questions:

1. What problems (with prior work or the lack thereof) were addressed or surveyed by the authors?

2. What solutions were proposed or surveyed by the authors?

3. What are the technical strengths and main contributions of the paper's proposed solutions?

4. What are the technical weaknesses of the paper's proposed solutions?  What suggestions do you have to improve upon the paper's ideas?

Each paper to be discussed in class will be assigned to a student to present in class.  Assignments will rotate thoughout the class.  Papers will be assigned approximately one week in advance of the presentation date.  The presenter of a given paper must email their Powerpoint slides to the instructor rhan@cs.colorado.edu by midnight of the night before the presentation.  The in-class presenter of a particular paper does not have to submit written reviews for any of the papers reviewed that same day in class.  The paper schedule may vary over the course of the semester, e.g. as new papers become published at the most recent conferences.

Programming Projects

There will be one or two software programming projects in which students will build applications on actual physical sensor nodes, deployed outdoors and/or indoors.  Some examples of such projects could be:

• Build a multicast routing protocol for indoor sensor networks.

• Build a sense-and-forward application that takes GPS location data and forwards it back to a base station using the previously built multicast routing protocol.

We will likely be programming in C on MANTIS OS on MICA2 motes.  The late policy on programs is 10% off the grade if late within one day, 20% off the grade for two days late, 30% off the grade for three days late.  Programs that are submitted more than three days late will no longer be accepted.  Programs should be emailed to the instructor.  More information will be available on the Announcements web page.  

Announcements

As the class progresses, announcements will be posted on the Announcements Web page of the class Web site.