[Dataset] rice/midas (v. 2010-01-21)

top

the initial version

@MISC{rice-midas-2010-01-21,
  author = {Ardalan Amiri Sani and Lin Zhong and Ashutosh Sabharwal},
  title = {{CRAWDAD} data set rice/midas (v. 2010-01-21)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/rice/midas},
  month = jan,  
  year = 2010
}
					

Accelerometer and compass readings along with network usage and application 
data usage information from 11 smartphone users, each for one week in the field.

This data comes with the following license from Rice Efficient Computing Group.

1. We grant you a nonexclusive, nontransferable license to use the data and/or 
code for commercial, educational, and/or research purposes only. You agree to 
not redistributed the data/code without written permission from us.

2. If the download includes traces collected from human subjects, they are 
anonymized. To respect the privacy of those human subjects whose activity is 
captured by the data, you will not attempt to reverse the anonymization 
process. This may include but is not limited to identifying specific MAC 
addresses, Cell IDs, the actual users, or their location.

3. You agree to acknowledge the source of the data/code, i.e., the MobiCom'10 
paper "Directional Antenna Diversity for Mobile Devices: Characterizations and 
Solutions", in your publications and/or reports.

4. We provide no warranty whatsoever on any aspect of the data and/or code.
Use at your own risk.

We employ a commercial smartphone with a built-in tri-axis accelerometer and 
compass, HTC G1, to collect data regarding device orientation from real mobile 
device usage. We have developed a software logger to collect accelerometer and 
compass readings along with information regarding wireless network status, and 
applications being used for HTC G1. The logger runs in the background and 
collects data about every 10ms, including both voice and Internet usage. We 
employ a larger battery and an extended battery compartment cover for the HTC 
G1 so that the smartphone has 10 to 15 hours of battery lifetime with normal 
usage, long enough not to significantly affect participants’ usage.

We have collected accelerometer and compass readings from 11 smartphone users, 
each for one week along with both voice and data usage. The average usage time 
for a participant is 25.9 hours.

In practice, there are a few challenges to the estimation algorithm from the 
hardware limitations of accelerometer and compass in smartphones. First, both 
the accelerometer and the compass are prone to hardware noise. In particular, 
the compass can be easily interfered by electromagnetic activities nearby, 
including those by the host device. From controlled measurement conducted in 
the lab, we observe that such interferences and hardware noise tend to be much 
higher frequency than what could be possibly introduced by device rotation or 
mobility. Therefore, when analysing the logs, we suppress the noise and 
interference with a moving average window of 25 samples. Moreover, the 
numerical readings of a sensor can be different from the physical value. For 
example, when there is no acceleration, the accelerometer reading can be 
non-zero. When processing the logs, we remove this offset by data demeaning. 
Finally, while we leverage the impact of gravity on the accelerometer reading 
to estimate the orientation, the accelerometer reading also reflects the 
mobility of a device through acceleration. We note that the gravity is almost 
constant while the external acceleration due to mobility changes much faster. 
Therefore, in our analysis, the impact from mobility is suppressed with the 
mentioned low pass filter.

[Traceset] rice/midas/android (v. 2010-01-21)

top

the initial version.

@MISC{rice-midas-android-2010-01-21,
  author = {Ardalan Amiri Sani and Lin Zhong and Ashutosh Sabharwal},
  title = {{CRAWDAD} trace set rice/midas/android (v. 2010-01-21)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/rice/midas/android},
  month = jan,  
  year = 2010
}
					

Accelerometer and compass readings along with network usage and application 
data usage information from 11 smartphone users, each for one week in the field.

[Trace] rice/midas/android/txt (v. 2010-01-21)

top

the initial version

@MISC{rice-midas-android-txt-2010-01-21,
  author = {Ardalan Amiri Sani and Lin Zhong and Ashutosh Sabharwal},
  title = {{CRAWDAD} trace rice/midas/android/txt (v. 2010-01-21)}, 
  howpublished = {Downloaded from http://crawdad.cs.dartmouth.edu/rice/midas/android/txt},
  month = jan,  
  year = 2010
}
					

Accelerometer and compass readings along with network usage and application 
data usage information from 11 smartphone users, each for one week in the field.

The traces are combined into 4 gzipped tar archives, with several users in each
archive. Data of each user is in its own subdirectory and split into
subdirectories by weeks.

Each line of the log files is one sample of sensors readings,which have the 
following format:

        Timestamp, Accelerometer x-axis, Accelerometer y-axis, Accelerometer 
z-axis, Compass x-axis, Compass y-axis, Compass z-axis, Call state, Data state,
Wifi state, Application package name

Below you can find the explanations of each of these parameters:

1. Timestamp: Timestamp of the sample reported in system time which is in 
milliseconds since January 1, 1970 00:00:00 UTC.

        see http://developer.android.com/reference/java/lang/System.html#currentTimeMillis()

2. Accelerometer x-axis: Accelerometer reading on device x axis
        
        see http://developer.android.com/reference/android/hardware/SensorEvent.html 
for definition of axes of the device

3. Accelerometer y-axis: Accelerometer reading on device y axis

4. Accelerometer z-axis: Accelerometer reading on device z axis

5. Compass x-axis: Compass reading on device x axis

6. Compass y-axis: Compass reading on device y axis

7. Compass z-axis: Compass reading on device z axis

8. Call state: Shows device call state and takes one of the following values:

        0 = CALL_STATE_IDLE
        1 = CALL_STATE_RINGING
        2 = CALL_STATE_OFFHOOK

        for details, see http://developer.android.com/reference/android/telephony/PhoneStateListener.html#onCallStateChanged(int, java.lang.String)

9. Data state: Shows device data connection state and takes one of the following values:

        0 = DATA_DISCONNECTED
        1 = DATA_CONNECTING
        2 = DATA_CONNECTED
        3 = DATA_SUSPENDED

        for details, see http://developer.android.com/reference/android/telephony/PhoneStateListener.html#onDataConnectionStateChanged(int) 

10. Wifi state: Shows device wifi state and takes one of the following values:

        0 = WIFI_STATE_DISABLING
        1 = WIFI_STATE_DISABLED
        2 = WIFI_STATE_ENABLING
        3 = WIFI_STATE_ENABLED
        4 = WIFI_STATE_UNKNOWN

        for details, see http://developer.android.com/reference/android/net/wifi/WifiManager.html#getWifiState()

11. Application package name: Package name of the foreground application. It 
can be used to identify the foreground application.

[Author] Ardalan Amiri Sani

top

[Author] Lin Zhong

top

[Author] Ashutosh Sabharwal

top

[Paper] sani-directional

top