Optimization of UWB indoor positioning based on hardware accelerated Fuzzy ISODATA

Area of Science:

• Wireless Communication
• Indoor Positioning Systems
• Signal Processing

Background:

• Ultra-Wideband (UWB) technology is crucial for indoor positioning.
• Non-line-of-sight (NLOS) conditions significantly increase positioning errors in complex environments.
• Existing methods struggle to mitigate NLOS effects effectively, limiting accuracy.

Purpose of the Study:

• To enhance the accuracy and efficiency of UWB-based indoor positioning systems.
• To reduce positioning errors caused by non-line-of-sight (NLOS) factors.
• To accelerate the computational performance of the positioning algorithm.

Main Methods:

• Utilized a fuzzy iterative self-organizing data analysis clustering algorithm (ISODATA) to process UWB data and mitigate NLOS influences.
• Implemented FPGA acceleration for the computationally intensive parts of the ISODATA algorithm.
• Validated the system's performance against the MATLAB platform for speed comparison.

Main Results:

• Achieved indoor positioning accuracy within 2 cm by stabilizing positioning errors.
• Reduced the impact of NLOS factors on UWB positioning accuracy.
• Accelerated algorithm execution speed by approximately 100 times compared to MATLAB, significantly improving computational efficiency.

Conclusions:

• The fuzzy ISODATA algorithm effectively reduces NLOS errors in UWB indoor positioning.
• FPGA acceleration dramatically enhances the algorithm's processing speed, making real-time applications feasible.
• The combined approach offers a robust and efficient solution for high-accuracy indoor positioning.