Enhanced PDR-BLE Compensation Mechanism Based on HMM and AWCLA for Improving Indoor Localization.

Area of Science:

• Robotics
• Sensor Fusion
• Indoor Localization Systems

Background:

• Indoor localization remains challenging due to signal uncertainties and sensor inaccuracies.
• Existing methods often require linearization or struggle with varying environmental conditions.
• Pedestrian Dead Reckoning (PDR) systems are prone to cumulative position drift and heading errors.

Purpose of the Study:

• To develop an enhanced PDR-BLE compensation mechanism (EPBCM) for improved indoor localization accuracy and resilience.
• To address challenges in Received Signal Strength Indicator (RSSI) smoothing and sensor data compensation.
• To reduce position drift, zero-velocity errors, and heading inaccuracies in indoor environments.

Main Methods:

• Utilized an Unscented Kalman Filter (UKF) and Kalman Filter (KF) for Received Signal Strength Indicator (RSSI) smoothing without linearization.
• Implemented a fusion of conflicting information and activity detection using a Hidden Markov Model (HMM) for accelerometer data.
• Compensated for inadvertent body acceleration and magnetic distortion using estimated orientation.

Main Results:

• The EPBCM algorithm demonstrated superior performance in an indoor localization scenario.
• Achieved enhanced orientation estimation and more accurate zero-velocity measurements.
• Significantly reduced position drift, leading to higher overall position accuracy compared to existing methods.

Conclusions:

• The EPBCM provides a robust and accurate solution for indoor localization using PDR and BLE.
• The method effectively handles sensor uncertainties and environmental variations.
• EPBCM offers a promising advancement for reliable indoor positioning systems.