User Behavior Modeling in IP Networks for QoE Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for assessing Quality of Experience (QoE) in IP networks are limited in accurately measuring subjective user experiences and emulating realistic user behavior, particularly for edge devices, leading to inadequate testing and service quality assurance for Internet Service Providers (ISPs) and enterprises.
Innovation Solution
A method for establishing QoE scores for various user activities over IP networks using a combination of objective and subjective metrics, with embedded meta-commands in normal network packets to synchronize test control information, allowing for realistic emulation of user behavior and improved network testing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional QoE assessment methods are used, then implementation is simpler, but measurement precision of subjective user experiences deteriorates
Solution Approach 1:
The patent segments user behavior into distinct behavior profiles (e.g., browsing profile, downloading profile, streaming profile) with specific activity patterns. Each profile contains segmented parameters such as activity duration, packet intervals, and bandwidth usage patterns. This segmentation enables precise measurement of subjective QoE experiences by matching actual user behavior to appropriate profiles, thereby improving measurement precision without requiring a monolithic complex system.
Solution Approach 2:
The patent applies preliminary action by pre-defining behavior profiles with characteristic activity patterns before actual network testing occurs. These profiles include pre-configured parameters such as typical user activity sequences, packet transmission intervals, and bandwidth consumption patterns. By establishing these profiles in advance, the system can accurately assess QoE for subjective experiences without requiring complex real-time analysis of every user action.
2Reliability
If realistic user behavior emulation is implemented, then service quality assurance improves, but device complexity increases
Solution Approach 1:
The patent creates simplified copies of real user behavior through behavior profiles that replicate characteristic activity patterns without requiring actual users. Each profile copies essential aspects of user behavior such as activity duration, packet transmission timing, and bandwidth usage patterns. This copying approach enables reliable service quality assurance by testing against realistic scenarios while keeping edge devices relatively simple, as they only need to recognize and respond to these predefined behavioral patterns rather than analyze complex real user actions.
Solution Approach 2:
The patent uses parameter changes to represent different user behavior types by modifying key parameters within behavior profiles. Instead of implementing complex emulation of each specific user action, the system changes parameters such as activity duration, packet intervals, and bandwidth consumption to reflect different user behaviors (e.g., light browsing vs. heavy downloading). This approach improves service quality assurance across different user scenarios while maintaining edge device simplicity through parameter-based differentiation rather than structural complexity.
3Measurement precision
If meta-commands are embedded in network packets, then test control synchronization improves, but loss of information increases
Solution Approach 1:
The patent merges test control information with normal network packets by embedding meta-commands within the packet structure. Rather than using separate communication channels for test control, the system combines control commands with data packets, utilizing existing packet fields to carry additional control information. This merging approach improves test control synchronization by ensuring that control information travels with the data it governs, while minimizing information loss by using established packet formats that preserve original packet integrity.
Data Source
AI summary
A method of modeling user behavior in an IP network, comprising the steps of allowing a testing user to create a behavior profile specifying one or more user activities from a plurality of user activities performed by a user over the IP network, one or more bandwidth usage levels corresponding to the one or more user activities, and an importance level associated with each user activity; associating one or more applications from a plurality of applications with the one or more user activities; emulating the behavior profile over the IP network by performing the one or more user activities using the one or more applications; and measuring a quality of experience (QoE) score for each user activity from the one or more user activities.