Anomalous Device Operation Detection via Behavior Pattern Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Detecting anomalous operation of electronic devices is challenging, especially when they cannot be physically inspected or monitored directly, as it involves identifying unapproved hardware or software modifications that can lead to untested or unwanted device states, potentially causing damage or unfair advantages.
Innovation Solution
A computer-implemented method using machine learning models, such as behavior artificial neural networks and convolutional neural networks, processes user input and visual scene data to determine behavior pattern labels and anomaly likelihood, enabling efficient detection of anomalous device operation through data fusion and federated learning, which preserves privacy and reduces bandwidth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct hardware inspection is used to detect anomalous operation, then detection accuracy is improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The patent introduces behavior data as an intermediary representation of device operation. Instead of directly inspecting hardware states, the system collects behavior data from software interfaces and processes this data through machine learning models to infer anomalous operations. This intermediary approach enables accurate detection without requiring direct hardware inspection capabilities.
Solution Approach 2:
The patent replaces physical hardware inspection mechanisms with computational analysis. Machine learning models process behavior data to detect patterns indicative of anomalous operations, substituting the need for physical inspection hardware with software-based detection algorithms that operate on digital representations of device behavior.
2Reliability
If comprehensive monitoring of all device operations is implemented, then anomaly detection capability is improved, but data processing requirements and computational cost increase
Solution Approach 1:
The patent extracts only the necessary behavior data relevant to detecting anomalous operations from the full range of possible device operations. By focusing on specific behavior patterns and using machine learning models trained to identify anomalies, the system processes only essential information rather than all available data, reducing computational requirements while maintaining detection reliability.
Solution Approach 2:
The patent applies partial monitoring by selectively observing specific behavior patterns rather than continuously monitoring all device operations. The machine learning models are designed to detect specific types of anomalous behavior, allowing the system to process data selectively based on what is most relevant for anomaly detection, thereby reducing overall computational cost.
3Measurement precision
If behavior data processing using machine learning models is used, then anomaly detection accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary actions by pre-training machine learning models on behavior data before actual anomaly detection is needed. The models are trained in advance to recognize patterns associated with anomalous operations, so that during actual use, the system can quickly make predictions without performing complex analysis in real-time, thus reducing processing time while maintaining high accuracy.
Data Source
AI summary
A computer implemented method for autonomous anomalous device behaviour detection. Including receiving behaviour data, wherein the behaviour data is indicative of a user's inputs to an electronic device. A behaviour pattern label and an indication to the likelihood of an anomaly occurring are determined based on the received behaviour data, wherein the behaviour pattern label belongs to behaviour pattern label hierarchy.