An overview of mobile device network traffic and network interface usage patterns

Abstract

In this paper, we present and evaluate the outcomes of a measurement study amongst Android mobile device users, who volunteered their device-level network activity data through a newly developed mobile application in January 2013. We evaluate the submitted data in two hour time intervals with respect to device-level network traffic amounts, application network activity times, and data distribution (as measure of connectivity) between mobile (cellular) and wireless LAN networks.We find fairly homogeneous values with low levels of autocorrelation or long range dependence for the device-level amounts of data, but an indication for self-similarity for the summed application network usage times, which are positively correlated. In addition, we observe that the average distribution for cellular interface usage exhibits clear patterns for the day of the week as well as the time of the day. The combination of these findings can find direct utilization in future mobile device utilization modeling efforts.

Capacity level modeling of mobile device bandwidth requirements employing high utilization mobile applications

“…The proposed model is in turn suitable for broad application scenario ranges, even if only basic or aggregate parameters of main applications on mobile devices are known by, e.g., a cellular network provider wishing to evaluate capacities. A description of the characteristics of the device-level activities for volunteer-submitted data that captured network activities in 2-hour intervals over varying time horizons was presented in [12]. Due to space limitations, we refer to [12] for greater details about the dataset, which we utilize as basis.…”

“…A description of the characteristics of the device-level activities for volunteer-submitted data that captured network activities in 2-hour intervals over varying time horizons was presented in [12]. Due to space limitations, we refer to [12] for greater details about the dataset, which we utilize as basis. 2 The remainder of this article is structured as follows.…”

A User-adaptive Modeling for Eating Action Identification from Wristband Time Series

“…Usage of various other types of wearables that are enabled by smartphone base-stations can be found in the literature [22,23,37]. However, the wearable sensor of choice for eating activity analysis has been the accelerometer, which detects any movement of the arm or fingers.…”

Identifying In-App User Actions from Mobile Web Logs

“…However, the technique is not feasible because of encryption in modern mobile Web traffic. In work by [15] and [13] a device-level point of view is taken to investigate network activity characteristics of mobile devices. Both authors gather and analyse data from users who voluntarily installed an app on their device, which captured their applications' network activity over a period of time.…”