CRAWDAD intel/placelab

Location-aware dataset collected using Place Lab software.

These traces contain 802.11, GSM and GPS trace data collected using Place Lab software, for 3 different neighborhoods in the Seattle metro area. Total trace duration is approximately 2 hours, with around 55,000 total readings.

date/time of measurement start: 2004-09-26

date/time of measurement end: 2004-09-29

collection environment: The accuracy of Place Lab depend on the number and mix of beacons in the environment, making it difficult to make absolute statements about the system's performance. To quantify the accuracy of Place Lab and how they vary by area, we measured both 802.11 beacon density and corresponding Place Lab accuracy in an urban, a residential and a suburban area.

network configuration: For each area (see the traceset included), we drove around the areas with a laptop with an Orinoco 802.11 interface, a GPS unit  (Wired Garmin Rhino GPS unit), and a Nokia 6600 cell phone.

data collection methodology: We collected 802.11 and GSM beacons periodically using Place Lab software. We also took GPS readings for measuring "ground truth" location to be used for accuracy estimation. Total trace duration is approximately 2 hours, with around 55,000 total readings.

Traceset

intel/placelab/placelab

Place Lab traceset for location accuracy analysis.

• file: pervasive05_traces.tar.gz
• measurement purpose: Location-aware Computing
• methodology: For each locale (see the traces included - downtown, ravenna, and kirkland), we drove around the areas for sixty minutes with a laptop, a GPS unit, and a Nokia 6600 cell phone. 802.11 scans were performed at 4Hz using an Orinoco 802.11 interface in the laptop. GPS readings were taken at approximately 1Hz using an external serial GPS unit. Finally, the GSM measurements were taken at 1Hz by the Nokia 6600 and relayed to the laptop via Bluetooth4. At all times we tried to navigate within areas in which GPS lock would not be lost as GPS forms the round truth location to be used to estimate beacon positions and Place Lab's accuracy.
• limitation: Unfortunately, our Nokia cell phones only allow us to know the ID of the current cell tower with which the phone is associated, making it impossible to learn the full set of towers in range. While this allows us to know if coverage is available, it does not let us learn about density or Place Lab's accuracy if all towers in range were known. Thus all GSM-based Place Lab results are calculated using the single available cell ID.

intel/placelab/placelab Traces

• downtown: Place Lab log collected from Downtown, Seattle.
• configuration: Collected from Downtown Seattle - a mix of commercial and residential urban high-rises.
• format: File names are as follows: downtown{no}.{month}.{day}.{year}.txt

- no: serial number 

- month, day, year: measurement start date in MM.DD.YY format 

All files are in the Place Lab log format. 

(For documentation on the log format and tools 

that can parse them, visit http://www.placelab.org )

• ravenna: Place Lab log collected from Seattle's Ravenna neighborhood.
• configuration: Collected from Seattle's Ravenna neighborhood - a medium-density residential neighborhood
• format: File names are as follows: ravenna{no}.{month}.{day}.{year}.txt

- no: serial number 

- month, day, year: measurement start date in MM.DD.YY format 

All files are in the Place Lab log format. 

(For documentation on the log format and tools 

that can parse them, visit http://www.placelab.org )

• kirkland: Place Lab log collected from Kirkland, Washington.
• configuration: Collected from Kirkland, Washington - a sparse suburb of single-family homes
• format: File names are as follows: kirkland{no}.{month}.{day}.{year}.txt

- no: serial number 

- month, day, year: measurement start date in MM.DD.YY format 

All files are in the Place Lab log format. 

(For documentation on the log format and tools 

that can parse them, visit http://www.placelab.org )