Indoor NLOS Positioning System Based on Enhanced CSI Feature with Intrusion Adaptability

Area of Science:

• Wireless communication
• Signal processing
• Indoor positioning systems

Background:

• Channel State Information (CSI) is crucial for indoor applications like localization and intrusion detection.
• Non-line-of-sight (NLOS) conditions and continuous system operation pose practical challenges for accuracy and efficiency.
• Existing passive positioning systems can lead to unnecessary computation.

Purpose of the Study:

• To propose an enhanced CSI-based indoor positioning system with pre-intrusion detection for NLOS scenarios (C-InP).
• To improve feature discrimination under NLOS conditions.
• To reduce computational load by integrating intrusion detection before positioning.

Main Methods:

• Developed a modified calibration method for phase transformation to enhance feature discrimination in NLOS.
• Implemented outlier filtering for amplitude using variance distribution and median sequence.
• Utilized binary and improved multiple Support Vector Classification (SVC) models for intrusion detection and fingerprint localization.

Main Results:

• The C-InP system demonstrated superior performance in NLOS environments compared to existing systems.
• Achieved a Mean Distance Error (MDE) of 0.49 m in integrated rooms.
• Achieved a Mean Distance Error (MDE) of 0.81 m in complex garage environments.

Conclusions:

• C-InP effectively addresses challenges in CSI-based indoor positioning under NLOS conditions.
• The integrated intrusion detection module enhances system practicality and efficiency.
• The proposed methods significantly improve localization accuracy in complex indoor settings.