Dead Zone Mitigation for Vehicle Passive Entry Systems

The patent introduces predictor modules as intermediary components that generate predicted signal strength values when direct measurement is unavailable. These predictors act as mediators between the antenna modules and the passive entry system, providing estimated signal strength data that allows the system to function in dead zones while maintaining measurement reliability through multiple prediction methods (e.g., based on historical data, spatial interpolation, or machine learning models).

The system performs preliminary actions by pre-calculating and storing signal strength patterns, propagation models, and environmental characteristics in advance. When a dead zone is detected, these pre-computed data and models are quickly applied to generate accurate predictions without requiring real-time complex calculations, enabling seamless passive entry functionality even in areas with no direct signal measurement capability.

The patent implements a hierarchical prediction strategy where multiple predictors are generated but not all are equally weighted or processed. The system uses a combination of simple predictors (e.g., based on recent history) and complex predictors (e.g., machine learning models), activating only the necessary level of complexity based on the situation. This partial action approach achieves sufficient location accuracy without the excessive computational burden of running all possible prediction methods at full capacity simultaneously.

The system incorporates feedback mechanisms where the results from multiple predictors are continuously evaluated and compared. When predictors converge on a consistent location estimate, the system accepts this result. When they diverge, the system adjusts its prediction strategy based on feedback from recent accuracy measurements, environmental conditions, and resource availability, dynamically balancing accuracy requirements with computational constraints.