Overview

Project Topic

Project topic: Wireless Sensor Networks for Internet of Things Applications

Specifics of the project:

Simplest version: go-box with Raspberry Pi inside, listening to and profiling wireless networks/traffic.
Complex version: (same as above) plus Pi will "phone home" to C2 server
Ultimate version: (same as above) plus external sensors: GPS? camera? elevation? etc.

Covers multiple skill sets:

Wireless networks and how they work
Sensor data acquisition - wireless data and other location data
Hardware side: circuitry, raspberry pi, hardware hacking
Software side: Linux, networking, wireless, cybersecurity

Abstract

Thanks to Moore's Law, computers are getting more powerful, and at the same time, getting small enough to fit in our pockets. But the real power comes from wireless internet (wifi), which can connect all of those devices together. The Internet of Things (IoT) is the idea that we can use tiny computers, cheap sensors, and wifi connections to gather data about anything - a vehicle, a kitchen, a battery, or the weather - and analyze the data to gain a better understanding of how the system works.

In this project, we will deploy several Raspberry Pi microcomputers with wifi antennas, and gather data about the wireless radio signals that are received. This data will be analyzed to create a model of wireless traffic in a given area over time. The project has many aspects to focus on, including databases, statistical analysis, Python programming, Linux networking, and wireless radio antenna design.

Ahead of Time/Ready to Go

Want to have a working demonstration of what kind of data they can obtain about wireless networks, what kind of mathematical analysis you can do, what we will actually do with the data.

Course Materials

Handouts and written materials

Like programming class - provide formal information, but show examples in meetings

Milestones and Goals

Schedule for planning goals and milestones for students
Each student keeps track of their goals and milestones
Progress report - short report on the "what have we done so far"
(don't put off documentation until the end)

Weekly group meetings

Show them examples of things, explain with a white board
Computers are going to be an issue... still.

Online format, how to communicate together?

Slack?
Possibly a re-usable wiki. This would make it possible to transfer information across quarters. (Need?)

To have students prepared for materials:

Have a standard plan of what each student will be doing
Set a weekly date and time for group meetings

Task lists

Students work through a task list
What 6-8 tasks do they need to have accomplished so they'll be where they need to be at end of quarter?

Balance

What is already done, versus what do they actually do?
Again - use the Raspberry Pi balcony sensor project as a prototype
They will need to replicate the setup
What is the setup? The setup is a Raspberry Pi generating sensor data - they need to capture and analyze that data

Resources

External Links

Roaming wifi analysis with wireshark: http://www.revolutionwifi.net/revolutionwifi/2013/01/wi-fi-roaming-analysis-with-wireshark.html

Excellent use case: wifi data for business intelligence

http://www.ymc.ch/en/case-study-retail-wifi-log-file-analysis-with-hadoop-and-impala-part-1
http://www.ymc.ch/en/case-study-retail-wifi-log-file-analysis-with-hadoop-and-impala-part-2
http://www.ymc.ch/en/case-study-retail-wifi-log-file-analysis-with-hadoop-and-impala-part-3
Focus on machine learning, business intelligence, data science workflow
No fumbling with tools: install OpenWRT, increase log verbosity for all wifi events, and dump logs into database

More business applications of wifi data analysis: hotels

Customer data rates, device types/mfg, length of sessions, etc.
http://www.hospitalitynet.org/news/4068025.html

PlaceLab: mashup of GPS and wifi to triangulate positions

low-cost, low-accuracy, high-availability solution to location-finding
client-side, open source, democratic, privacy-conscious
"GPS, the most famous location system, has a major limitation in that it is virtually useless indoors due to its dependency on receiving satellite transmissions. PlaceLab's dependence on 802.11 beacons, GSM masts and Bluetooth devices make it very useful indoors, but almost useless in rural areas. As mentioned before, PlaceLab aims to provide a high coverage in terms of the time in which the service is available to users and since most users are assumed to be in urban areas this is less of an issue. "
http://ntrg.cs.tcd.ie/undergrad/4ba2.05/group1/index.html