Mobile Device Travel Jump Detection for Fraud Identification

The patent applies local quality by making different parts of the monitoring system perform different functions: basic location tracking for all devices, and enhanced analysis with signal quality measurements only for suspicious events. This allows the system to maintain high detection accuracy for fraudulent activities while avoiding the computational overhead of analyzing every device event in detail, thereby reducing false positives.

The system performs preliminary filtering of handover events before detailed analysis. By pre-identifying and excluding conventional handover patterns using historical data and mobility profiles, the system prepares the data in advance to focus computational resources only on potentially fraudulent events, improving both accuracy and reducing false positives.

The patent segments the detection system into multiple independent components: event collection module, handover filtering module, signal quality analysis module, and fraud detection module. Each component performs a specific function with manageable complexity. The segmentation allows the system to achieve high detection accuracy through coordinated modules without requiring any single component to be overly complex.

The patent introduces intermediate processing layers between raw network events and fraud detection. Handover filtering and signal quality measurement act as intermediaries that preprocess data, removing obvious legitimate events and enhancing suspicious patterns before final analysis. This intermediary processing reduces the complexity of the final detection algorithm while maintaining high accuracy.