Dynamic Timeout Threshold for Communication Flow Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting the end of data communication in networks face challenges in accurately setting a timeout period, leading to either false detection of communication stop or delayed detection, especially in best effort type networks with potential disconnection or power failures.
Innovation Solution
A communication apparatus and method that sets a threshold based on packet intervals to detect the end of data communication by comparing the packet stop period to the threshold, incorporating a threshold setting unit and a timeout detecting unit to dynamically determine the timeout period.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a relatively short predetermined time (e.g., about one second) is set for timeout determination, then communication stop detection speed is improved, but false detection of communication end increases due to normal packet communication pauses
Solution Approach 1:
The patent applies dynamics by making the timeout threshold adaptable rather than fixed. The threshold is dynamically adjusted based on the measured packet intervals from normal communication operations. By learning the typical packet transmission patterns and setting the timeout threshold based on these learned intervals (e.g., several times the average interval), the system adapts to different communication scenarios, allowing fast detection when appropriate while avoiding false positives during normal pauses.
2Reliability
If a considerably long predetermined time (e.g., about 600 seconds) is set for timeout determination, then detection accuracy for forcible communication termination is improved, but detection timing is significantly delayed
Solution Approach 1:
The system dynamically adjusts the timeout threshold based on learned packet intervals from actual communication patterns. Instead of using a universally long timeout period, the threshold is set adaptively (e.g., 3-5 times the average packet interval) based on the specific communication flow characteristics. This allows the system to detect forcible terminations accurately while minimizing detection delay, as the threshold reflects actual communication timing rather than a conservative fixed value.
Solution Approach 2:
The system performs self-learning by monitoring packet intervals during normal communication operations. It automatically characterizes the communication pattern and uses this learned information to set appropriate timeout thresholds without external configuration. This self-service approach enables the system to optimize detection timing based on actual usage patterns, reducing unnecessary delays while maintaining accuracy.
3Device complexity
If a fixed predetermined time is used for timeout determination, then device complexity is reduced, but adaptability to different communication patterns is worsened
Solution Approach 1:
The system implements self-service by automatically learning communication patterns and configuring its own timeout threshold based on observed packet intervals. The communication apparatus monitors its own operation, characterizes the timing patterns, and autonomously sets the timeout threshold (e.g., as a multiple of the learned average interval). This eliminates the need for manual configuration or complex external control while maintaining high adaptability to different communication scenarios.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting the timeout threshold parameter based on observed communication patterns. Instead of using a fixed parameter value, the system modifies the timeout parameter adaptively according to the learned packet intervals. This allows the same basic mechanism to work effectively across different communication types and patterns without requiring multiple fixed configurations or complex decision logic.
Data Source
AI summary
A packet interval (a time difference between sequentially received packets) in a certain previous time is calculated and a cumulative distribution function of the calculated packet interval is obtained. Meanwhile, a function value-corresponding to-flow end that is a boundary value of the cumulative distribution function to be considered as a flow end (timeout) in the range of a cumulative distribution function value is stored. From the cumulative distribution function obtained as described above, a value of a packet interval mapped to a cumulative distribution function value that matches the function value-corresponding to-flow end is set as a threshold. That is, the threshold is dynamically set based on a previous packet interval. Therefore, a packet stop period from a last packet reception time to a current time is measured, and it is detected as a flow timeout at a timing when the measured value is equal to or greater than the threshold.