IoT Sensors and Systems

  • Course: CS6501/ECE6501 (008) - Fall 2017
  • Instructor: Brad Campbell
  • Time: M/W 10:30am-11:45am
  • Location: 340 Rice
  • Office Hours: Tu 2pm-3:30pm, 512 Rice

The Internet of Things promises to revolutionize how we interact with computers by making embedded computation ubiquitous. New devices will be added to cities, to homes, to factories, to ourselves (inside and out), to cars, and to many other facets of life. The hope is that this influx of technology will help us solve many pressing societal issues in areas such as energy, personal health, the environment, and safety. The challenges lie in designing and scaling the hardware platforms, networking protocols, and programming paradigms to enable this new class of computing to be used productively.

This course will start by covering several key application areas and various systems that address issues within those applications. With the application drivers in place, we will explore the hardware platforms that support the sensing, energy, and deployment requirements of the various application domains. With methods to build the IoT devices and ensure they are low power, we move to techniques for providing connectivity to individual and networks of devices. We then will investigate systems for programming and providing security primitives for the networks of devices, as well as how the IoT can expose new privacy concerns.

As a graduate seminar, this course will focus on reading, analyzing, and discussing research papers. The course will focus on very recent research (last 2-5 years), but will also include older, more foundational papers for certain topics.

Deliverables

This seminar requires three main deliverables from students.

1. Paper Reviews

Before each class you must read and write a short review of each assigned paper. The review should answer the following questions:

  • What is the problem this paper addresses, and why is it important?
  • What is the hypothesis of this paper?
  • What are two key assumptions that this paper makes?
  • What are the two main strengths of this paper?
  • What are the two main weaknesses of this paper?
  • Which figure or experiment was most compelling in support of the hypothesis, and why?

Additionally, the review should include ratings for how you perceive the paper in the following categories:

  • Presentation (1-5):
  • Interest (1-5):
  • Impact (1-5):
  • Overall (1-5):
  • Confidence (1-5):

The reviews must be entered in the review site before the start of each class and should be used as reference notes for the in-class discussion.

Note: it will always be easier to find weaknesses than it is to find strengths when reviewing a paper. After all, the authors had a limited number of pages! Try to find the merits that lead to the paper getting accepted while you read each paper.

2. In-class Discussion Lead

You must select one class where you will be the discussion lead for the assigned papers. You should come prepared to give an overview of the paper, and guide a discussion about the strengths, weaknesses, and potential for impact of the paper.

3. Semester Project

This class will feature a semester long project focused on idea creation, idea motivation, and the scientific method. You will choose a topic, develop a hypothesis within that topic, develop a research plan to explore and evaluate that hypothesis, provide some motivating preliminary data, and communicate these results in the form of a written paper.

Note: this does not require actually building the system. In fact, it requires that you do not build the system, but rather work on the science aspects before undertaking the engineering.

The project will have four milestones:

  1. Monday, September 25: Two page paper covering your hypothesis, motivation, and expected results.
  2. Wednesday, October 25: Four page paper with planned experiments.
  3. Monday, November 27: Five to six page paper with preliminary results.
  4. Wednesday, November 29 and Monday, December 4: Project presentations.

For the two in-class project workshops, you will be expected to evaluate and provide feedback for your peers’ papers.

Grading

Your grade for the course will be based on: individual or group project (34%), paper reviews and peer-review of final projects (33%), and in-class participation and discussion lead (33%).

Schedule

Class Topic Lead Details
Wed Aug 23 Introduction Campbell  
Mon Aug 28 Ubiquitous computing and the IoT Campbell No reviews: (1) Ubiquitous Computing, (2) IoT Research Challenges
  Applications    
Wed Aug 30 Outdoor and Wildlife Monitoring Campbell (1) Great Duck Island, (2) Bat Tracking
Mon Sep 4 Urban Environments Campbell (1) Array of Things, (2) Air Quality
Wed Sep 6 Buildings and Energy Campbell (1) Sentinel, (2) CapNet
Mon Sep 11 Democratic Monitoring Campbell (1) SeaGlass
Wed Sep 13 Personal and Population Health Campbell (1) LIBS, (2) Opo
Mon Sep 18 Infrastructure and Industrial Campbell (1) Pipeline, (2) Alps
Wed Sep 20 Safety and Security Anderson (1) Gunshot Detection, (2) LOOKUP
Mon Sep 25 Project Workshop 1   Bring first version to class.
  Sensors, Hardware, Energy    
Wed Sep 27 Hardware Platforms and Smart Dust Campbell (1) TelosB, (2) M3, No review: (3) Firestorm
Mon Oct 2     No class (Reading day)
Wed Oct 4 Energy Harvesting 1 Campbell (1) HydroWatch, (2) Monjolo
Mon Oct 9 Energy Harvesting 2 Campbell (1) Tragedy of the Coulombs, No review: (2) DoubleDip
Wed Oct 11 Hardware Generation and Sensing 1 Campbell (1) EDG, (2) Soli, No review: (3) Synthetic Sensors
  Networking and Interfacing    
Mon Oct 16 Low Power Wireless Crump (1) LPL, (2) LWB
Wed Oct 18 Networking   (1) IP is Dead, Long Live IP, (2) Interscatter
Mon Oct 23 Gateways and Whitespaces Campbell (1) Gateway Problem, (2) SNOW
Wed Oct 25 Project Workshop 2   Bring second version of the project paper to class.
Mon Oct 30     No class (SOSP).
  Programming and Security    
Wed Nov 1 Operating Systems TockOS (1) SOS, (2) DINO
Mon Nov 6     No class (SenSys). Work on project paper.
Wed Nov 8     No class (SenSys). Work on project paper.
Mon Nov 13 Blockchain: Hype or Useful? Campbell No Review: (1) Blockchain Overview (2) Sharing
Wed Nov 15 IoT Frameworks Hamid (1) Ravel, (2) HomeOS
Mon Nov 20 IoT Security Ulkuatam (1) SmartAuth, (2) Electromyography
Wed Nov 22     No class (Thanksgiving)
Mon Nov 27 Misusing Sensors Agrawal (1) Gyrophone, (2) PitchIn.
  End    
Wed Nov 29 Wrap-up and Future Directions Campbell Final papers due Friday.
Mon Dec 4 Project Presentations All 10 minute talks, 5 min Q&A

Notes for the future!

  • Maybe replace LWB with Glossy (or do both). Flooding + network protocol is a lot to cover.
  • OS papers were good choices.