Surveillance Device Dynamic Allowable Range Anomaly Detection
Solution Overview
Problem
Existing surveillance techniques for detecting anomaly frames in on-board vehicle networks struggle to accurately identify anomalies, especially when data frame changes are within predetermined threshold limits or have varying states, leading to false negatives or false positives.
Innovation Solution
A surveillance device and method that calculates changes in data frame parameters, generates distributions of these changes based on previous states, and sets allowable ranges for determining anomaly detection, allowing for more precise identification of anomalies by considering the specific state of the vehicle before the change.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a fixed threshold value is used to determine consistency between data types, then the detection method is simple to implement, but anomaly frames with values that cannot be actually detected cannot be identified
Solution Approach 1:
The patent applies dynamics by transitioning from fixed threshold values to dynamically determined allowable ranges. The surveillance device calculates allowable ranges based on the specific state of the apparatus before the change, making the detection criteria adaptive rather than static. This enables the system to accurately detect anomaly frames by comparing actual changes against state-dependent allowable ranges, resolving the contradiction between implementation simplicity and detection accuracy.
2Device complexity
If an absolute threshold value is used for amount of change detection, then the detection rule is simple, but frames with abnormal changes depending on vehicle state cannot be detected if within the threshold
Solution Approach 1:
The patent applies parameter changes by making the allowable range a variable parameter that depends on the apparatus state before the change. Instead of using a constant absolute threshold, the system calculates different allowable ranges based on the specific state (e.g., vehicle speed, temperature, pressure levels). This enables reliable detection of abnormal changes that are context-dependent, resolving the contradiction between rule simplicity and detection reliability.
3Loss of information
If predetermined threshold limits are used for anomaly detection, then false alarms are reduced through simple rules, but detection precision decreases when data changes vary with vehicle state
Solution Approach 1:
The patent resolves this contradiction by implementing dynamic allowable ranges that adapt to the apparatus state. The system calculates state-dependent allowable ranges and compares actual changes against these dynamic thresholds. This maintains low false alarm rates by using scientifically grounded ranges while achieving high detection precision through state-aware comparison, eliminating the need for simple but inaccurate predetermined thresholds.
Data Source
AI summary
A surveillance device according to the present invention includes: a memory; and at least one processor coupled to the memory. The processor performs operations. The operations includes: calculating, based on a value of a parameter that is contained in a received data frame and represents a state of an apparatus, a change in the state of the apparatus; and determining whether the change is included in an allowable range determined in accordance with the state of the apparatus before the change.