Sessionizing Network Flows for Accurate App Usage Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for analyzing app usage in wireless communication networks overestimate the time spent using applications due to the summation of multiple network traffic flows, making it difficult to accurately reflect user engagement metrics.
Innovation Solution
Combining multiple network flows into sessions based on start and end times, using a model or algorithm to determine session start and end times, and analyzing these sessions to estimate app usage, which allows for accurate engagement metrics and targeted promotions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple network traffic flows are summed with respect to duration, then complete network behavior is captured, but app usage time is greatly overestimated
Solution Approach 1:
The patent segments network traffic flows by creating separate session records for different applications. Each session record contains duration information only for its specific application, preventing the aggregation of unrelated traffic flows. This segmentation allows accurate measurement of individual app usage time while still capturing complete network behavior through multiple separate records.
2Loss of information
If event data records capture all network flows, then comprehensive network behavior is recorded, but accurate engagement metrics cannot be obtained
Solution Approach 1:
The patent applies local quality by creating application-specific session records where each record is optimized for its particular application. Each session record contains quality attributes (duration, data volume) specific to that application's usage pattern. This allows comprehensive network behavior to be captured while maintaining precise engagement metrics for each application through differentiated record structures.
3Quantity of substance
If summed network traffic flows are used for analysis, then complete usage data is available, but user behavior analysis becomes inaccurate
Solution Approach 1:
The patent segments usage data into application-specific session records, each containing only the traffic flows relevant to that particular application. This segmentation preserves the complete quantity of usage data while ensuring accurate user behavior measurement by preventing contamination from unrelated application traffic. Each session record can be independently analyzed for its specific application's user behavior patterns.
Data Source
AI summary
Systems and methods discussed herein are directed to combining multiple network flows with respect to usage of applications into sessions. A record of a plurality of network flows may be obtained where the record relates to usage of the application by a user device in a wireless communication network. The plurality of network flows may be combined into a session, where an earliest start time of a network flow represents a session start time for the session and a latest end time of a network flow represents a session end time for the session. Based at least in part on the session start time and the session end time, a usage of the application during the session by the user device may be determined. Based at least in part on a plurality of sessions, usage of the application in the wireless communication network may be estimated.