IoT Communication Pattern Convergence Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for collecting communication information from IoT devices face challenges in efficiently gathering data, as the duration of data collection can be either too long or too short, and setting a target number of patterns for learning can be difficult due to varying requirements for each device.
Innovation Solution
A determination apparatus that includes a collection unit, an extraction unit, a calculation unit, and a determination unit to collect, extract, and analyze communication patterns, approximating the change in the cumulative number of patterns to a function that expresses a predetermined curve, determining the degree of convergence, and stopping data collection when sufficient information is gathered.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If information collection continues for a long period, then sufficient communication patterns can be collected, but collection efficiency decreases and unnecessary data is gathered
Solution Approach 1:
The system continuously monitors the cumulative number of extracted patterns and compares it against the approximated saturation curve. When the actual cumulative value approaches the predicted saturation point (within a predetermined margin), the system automatically stops collection. This feedback mechanism prevents both under-collection and over-collection, optimizing both reliability and productivity.
Solution Approach 2:
The system performs preliminary approximation of the saturation curve using initial collected data before the collection process is fully complete. This allows the system to predict the saturation point in advance and adjust the collection duration accordingly, rather than using a fixed predetermined period.
2Productivity
If a fixed target number of patterns is set for collection, then collection can be stopped at a predetermined point, but it is difficult to set an appropriate target number that is not too large or too small
Solution Approach 1:
The system dynamically determines the collection target based on the actual convergence characteristics of each device. Instead of using a static predetermined target number, the system approximates the saturation curve from collected data and calculates the dynamic target as the cumulative value at the saturation point minus a predetermined margin. This adapts to different devices' communication patterns automatically.
Solution Approach 2:
The system changes the collection target parameter based on observed data characteristics. The target number of patterns is not fixed but is recalculated based on the approximated saturation curve parameters (asymptotic value and convergence rate), allowing automatic adaptation to different device types and communication behaviors.
3Productivity
If information collection is stopped too early, then collection efficiency is improved, but insufficient information is gathered for reliable pattern recognition
Solution Approach 1:
The system uses continuous feedback by comparing the actual cumulative pattern count against the approximated saturation curve. The collection process is monitored in real-time, and the stop condition is triggered only when the actual value approaches the predicted saturation point within a predetermined margin, ensuring sufficient data is collected before stopping.
Solution Approach 2:
The system collects slightly more data than the minimum required by continuing until the cumulative value approaches (but does not exceed) the saturation point minus a predetermined margin. This partial excessive action ensures sufficient data for reliable pattern recognition while avoiding unnecessary extended collection.
Data Source
AI summary
A determination apparatus according to a first embodiment collects information regarding communication performed by an IoT device. The determination apparatus extracts patterns used for detecting unauthorized communication performed by the IoT device from information that has been collected. Also, the determination apparatus approximates a change in the cumulative value of the number of patterns to a function that expresses a predetermined curve, thereby calculating the degree of convergence of the change. Also, the determination apparatus determines whether or not the degree of convergence is no less than a predetermined value.