Spring 17: ECE 598RR: Mobile Sensing and Sensor Fusion

Course Description: This course exposes students to ongoing research in mobile sensing, i.e., techniques, algorithms, and systems that leverage the sensors in smartphones, smartwatches, drones, and IoT devices, to deliver real-world applications. Topics include:
  1. GPS and related outdoor applications
  2. Indoor positioning systems
  3. Activity and gesture recognition
  4. Sensor data analytics in sports, vehicular, and smart environments
  5. Vibration sensing and acoustics
  6. Drone motion tracking
  7. Sensor data leakage and inference: security and privacy
  8. Augmented Reality
The course will teach students some popular/useful analytical techniques and then show the application of these techniques to real systems. Examples techniques are Trilateration, 3D Dead Reckoning, Orientation tracking, HMM, Viterbi Decoding, Doppler, Maximum Likelihood. Example applications are listed above. Overall, the course will take a systems/practical approach, helping students to:
  • Understand the state of the art in mobile sensing and computing
  • Understand how to initiate/engage in mobile computing research.
  • Develop good research taste in general, and approach design-questions systematically (via a semester-long research project)

Prerequisites: Probability, linear algebra, and programming maturity (any one of MATLAB, R, Python, or Java/C). The course will start with some mathematical refreshers and basic introduction to signal processing and data analytics. An Android tutorial may also be presented. Experience on mobile programming (Android/iOS) is a plus, but not necessary.

Course project: Students will propose their own project ideas (as a function of their interest & background); instructor will offer guidance.

Course Load: Read and submit reviews for ~25 research papers (~1 per class), present once in the semester, 1 final project.

Grading: Reviews (30%), Class presentation (15%), Final Project (55%)

Time and Location:   M/W 3 to 4:20pm @ 2013 ECE Building
Instructor:                  Romit Roy Choudhury (croy@illinois.edu)
Office hours:              M/W 4:20 to 5:00pm

Course TA:                Sheng Shen <sshen19@illinois.edu>               

Reference books:       (1) Introduction to Linear Algebra, Gilbert Strang (MIT)
                                    (2) Understanding Digital Signal Processing, Richard Lyons (UCSC)
                                    (3) Mathematical Foundations of Computer Networking, Srinivasan Keshav (University of Waterloo)
                                    (4) A Top Down Approach to Computer Networking, James Kurose (UMass)
                                    (5) Data Analytics, David Forsyth (UIUC)

Tentative Course Calendar (subject to change)
1.1 Course overview, logistics, and expectations [ppt]

1.2 Linear algebra refresher

2.1 Signal processing basics and refresher

2.2 Probability refresher

[notes1] [notes2] [notes3] [notes4]



Location Sensing via GPS
3.1 Basics: GPS trilateration, Carrier phase, Diff. GPS: SafetyNet [ppt]   [due Wed Feb 1]

3.2 Application: Coin GPS [ppt] + Relative GPS    [both due Mon Feb 6]

       Optional readings (no need to submit review): Relative GPS 2


Coin: Romit
Relative: Doyoun, James

Location Sensing via RF

4.1 Basics: Intro to comm. (AM, FM, QAM), multipath, CSI, OFDM [notes1, notes2]

4.2 Basics: WiFi and beamforming

5.1 Application: RADAR, Place Labs    [both due Web Feb 15]

5.2 Application: EZ, ArrayTrack    [both due Mon Feb 20]

       Optional readings (no need to submit review): DecaLoc and use of chinese remainder theorem




[Mon Feb 20]
EZ: Varun, Kartik
Array: Xianze, Xuewei

Location Sensing via IMU 6.1 IMU basics: discuss MEMS, Google video on IMU, analyzing time series data from sensors, bias in compass, sensor noise properties, averaging, dead reckoning. [notes]

6.2 Application: SurroundSense, UnLoc, WalkCompass    [all due Mon Feb 27]


Motion Tracking via IMU
Vibration Sensing
7.1 Basics: 3D orientation, rotation, sensor fusion.

7.2 Basics: Probabilistic graphical models, HMMs, Viterbi, Kalman filters, Particle filters [notes]

8.1 Applications: A3, ArmTrak    [both due Wed Mar 8]

8.2 Applications: MoLe    [due Mon Mar 13]

9.1 Applications: Zee, MagLoc    [due Wed Mar 15]

10.1 Basics: Vibra motors, ringing, resonance

10.2 Applications: Ripple 2    [due Wed Mar 29]

11.1 Application: Sports Analytics    [due Mon Apr 3]
        Project Discussion



[Wed Mar 8]
A3: Alex, Greg
Arm: Romit

[Mon Mar 13]
MoLe: Cathy, Shrey

[Wed Mar 15]
Zee: Jesse, Liyi

[Wed Mar 29]
MagLoc: Xuanyao, Jinwen
Vibration: Nirupam

[Mon Apr 3]
Sports: Romit
Motion Tracking via RF 12.1 Basics: Doppler, FFT points, classification, DTW [notes]

12.2 Applications: WiSee, WiTrack    [both due Mon Apr 10]


[Mon Apr 10]
12.2.WiSee: Yishuo, You
12.2.WiTrack: Anadi, Zicheng

Sensor Security, Privacy 13.1 VibraPhone [due Wed Apr 12],   BackDoor

13.2 Project Discussion

13.3 GyroPhone, Inferring Secrets from Android Public Sensor Data    [both due Wed Apr 19]



[Wed Apr 19]
13.3.Gyro:Dengfeng, Wing
13.3.Secrets:Ben, Michael

Time Permitting:

Acoustic Sensing

Touch Sensing

14.1 Basics:

14.2 Applications: 3D Mouse, Snooping keyboard

15.1 Capacitive touch, Gesture authentication

15.2 Assorted Sensing Applications (BackDoor, Tagoram, DecaWave, Heartrate, Emotion recognition)

Time permitting

Final Presentation Date: 8am, May 8 (Monday)
Final Report Due: noon, May 12 (Friday)