Projects

CpE 664 Sensor Actuator Networks

(Previously CS / EE 591 V)

Course Format and Credit hours

3 hr Lecture / Presentations

3 hr Credit

Prerequisites

For CS students: CS 450, CS 350 (Or equivalent) – Knowledge of operating systems concepts

For EE or other disciplines: EE 329 (Or equivalent) – Knowledge of signal processing

Schedule

MWF 12:00 to 12:50

Location

ESB 201

Office Hours

Tuesday and Wednesday 1:00 to 2:00 p.m. or by appointment

Course Objectives

Fueled in part by recent advances in MEMS and wireless communication technologies, sensor actuator networks and networks of embedded devices in general have become extremely popular and have been used in several applications such as environmental monitoring, perimeter security, structural control, asset tracking and personal healthcare systems. The objectives of this course are to introduce students to the state of the art in sensor actuator networks and to provide hands on training in programming these networks.

We will have a significant reading list from recent literature to accompany the lectures. Lectures will emphasize aspects of distributed systems such as fault-tolerance, reliability, and security. Case studies from existing applications will be used. Each student will have to complete a project.

The course is aimed both at students who wish to do research in the sensor actuator networks area, as well as at students from related disciplines, such as signal processing, wireless communications, databases, algorithms, etc., who wish to understand what new challenges sensor actuator networks pose for their own discipline

The course will serve as an elective option in the Computer Systems (Area 3) emphasis area. The course is appropriate for students with a major or minor in this area because it teaches the requisite system engineering principles for assembling large scale networked systems. The students will first learn elements of distributed computing and then apply them in the context of networked applications. Students will also gain hands-on experience in programming such systems using a network of wireless sensor motes. The course has been designed to accommodate students from CS (with background in operating systems, computer networks, algorithms etc.) and EE disciplines (with background in communication, signal processing, analog hardware etc.) who can collaborate on team projects and design systems that often require interdisciplinary expertise. The course is also expected to provide the basic scientific knowledge required to carry out research in the emerging area of cyber-physical systems, a largely interdisciplinary research area with networking and systems engineering at its core.

Note: This course will complement the EE 562 course on wireless communication systems. While EE 562 focuses on physical layer issues of wireless networks, in this course we will focus on protocol level issues in the higher layers of the sensor network stack.

Expected Learning Outcomes

Upon successful completion of this course:

1. Students will be introduced to some existing applications of wireless sensor actuator networks

2. Students will be introduced to elements of distributed computing and network protocol design and will learn to apply these principles in the context of wireless sensor networks

3. Students will learn the various hardware, software platforms that exist for sensor networks

4. Students will get an overview of the various network level protocols for MAC, routing, time synchronization, aggregation, consensus and distributed tracking

5. Students will learn to program sensor network platforms using TinyOS, C and Java and will get an opportunity to have hands on training in developing applications on wireless motes, smart phones and other embedded platforms

6. Students will read and present seminal papers on various issues in sensor networks, opening a path to research in this area

7. Students will understand what research problems sensor networks pose in disciplines such as signal processing, wireless communications and even control systems

Course Syllabus (and % of time per topic)

1. Introduction and examples (5%)

2. Elements of distributed computing and network protocol design (10%)

3. Hardware and software platforms for sensor networks, (10%)

4. Protocols for MAC layer, routing layer and localization (10%)

5. Protocols for network reprogramming (10%)

6. Coordination and consensus in sensor actuator networks (10%)

7. Protocols for information aggregation and querying (10%)

8. Embedded camera networks (10%)

9. Recent advances, Cyber-physical systems, Large scale complex systems (10%)

10. TinyOS (15%)

Project

For this course, students will be required to work on a project, individually or in a team of at most two persons. The project will usually involve programming using TinyOS or C / Java or Matlab. Students will either use a remote sensor network testbed or in-house sensors for their project. (In case of students from non-CS background who are not comfortable with programming, alternate projects can be designed.) An oral presentation and a written report will be due in the finals week. These projects may evolve into long term ones, which can be used for a Masters thesis or may lead to a PhD dissertation project.

Syllabus (pdf)

Some initial papers to read

Connecting the physical world with pervasive networks

Energy and performance considerations for smart dust

The computer for the 21^st century

Proactive computing

Next century challenges: Scalable coordination in sensor networks

Interesting Links (sensor actuator networks):

WSN Blog

Smart parking spaces

Some good advice for graduate students:

Edsger Dijkstra, The Three Golden Rules for Successful Scientific Research