5G Localization with Blockage Detection Based on Beam RSRP Patterns in Urban Environment

We propose a computationally more efficient localization method compared to artificial intelligence (AI)-based fingerprinting methods utilizing beam reference signal received power (RSRP) patterns for 5G urban environments. Our approach detects blockages using beam RSRP patterns and applies the AI model only for blocked cases. It then integrates the AI model with conventional localization methods, which offer low complexity and reliability. Numerical experiments show that the proposed method demonstrates relatively improved performance and efficiency in terms of computational complexity compared to AI-based fingerprinting.