MIT Researches Map Cell Phone Usage

stlhawkeye writes "MIT researchers with the Mobile Landscape Projects have mapped a city based on cell phone usage. The article includes a map of Graz, Austria with a color-coded overlay indicating cell phone usage in various parts of the city. Using call origin and destination data, they are able to not only reverse-engineer a topographic map of the geography and landscape, but one of phone usage as well. The implications of the research have practical applications in law enforcement, emergency management, and traffic management. There are also, of course, privacy implications."

Comments

Link in article broken.

[bluesoul88]

The correct address for MIT's "Mobile Landscape" project can be found here.

Privacy implications?

[sploxx]

What are the privacy implications if the study only uses anonymized location data, i.e. "in this field of 100m x 100m", there is a cell phone which now moves to this field etc.?

I think there are none. At least not any new ones than those implications by using cell phones at all.
The data about who uses which cell when does exist already and it needs to exist, in the current state, at all times in the phone system (how would you route calls without this information?)

Privacy concerns can surely be raised about storing such tracking profiles attached to particular persons. But just anonymized usage patterns?

Re:Invasion of Privacy

[negative3]

It depends on the standard; some are harder to eavesdrop on than others. Any fool with a 30kHz (even 10kHz will work) bandwidth FM demodulator and the proper tuner can get an analog cellphone signal with no problem (AMPS == Ancient Mobile Phone Service, and truly should be abandoned). This is effortless for a graduate student in digital communications. GSM presents a problem for eavesdroppers because all of the time slots are aligned at the base station - if the eavesdropper is not sitting at the base station the time slots could be horribly skewed & overlap or lag. CDMA presents even greater problems because time slots are not the issue but actually figuring out the spreading code, chip rate, etc. is a huge problem. But if this was done cooperatively with a service provider, I'd say that the chances are slight, especially after reading the first line of TFA: Using anonymous cellphone data provided by the leading cellphone operator in Austria, A1/Mobilkom. Do you think the data given to the students is the actual recorded calls? I'd expect it to be time of arrival and any available spatial data. So for now, let's leave the conspiracy theories to

Another aspect: cell phone companies design their systems based on call density & concentration - this could have been real news a decade ago. It's standard practice now. I can draw the cell phone usage in a city if you answer a few questions: where are the rush hour routes? where is the business district? what are the peak rush hour times? You can get a much better picture by actually analyzing a lot of data but the fundamental result will be the same!