Answers are grounded in this paper and cite their sources.
Fingerprint-based localization methods are regarded as a promising solution in sixth-generation (6G) wireless communication because of their ubiquitous infrastructure and high precision in lab-level experiments. However, received signal strength (RSS) instability and fingerprint spatial ambiguity (FSA) significantly undermine the precision of localization methods under real-world applications. Signal spatial difference (SSD) is one of the approaches which can effectively address RSS instability and FSA in fingerprint localization. However, few of these approaches are effectively incorporated into localization methods. In this paper, we propose an SSD-based advantage actor-critic (A2C) method, SSD-A2C, for 3D indoor localization. This method is the first to combine SSD and reinforcement learning (RL), achieving a highly efficient method that can address FSA and RSS instability. A 3D indoor localization simulation environment is developed based on ray-tracing (RT) simulation results of two real-world indoor scenarios, and the proposed method is trained and evaluated by it. Experiment results proved the excellent performances of SSD-A2C in minor localization errors and a high success rate of predicting the desired location. SSD is found to be more appropriate than RSS for RL. The strengths, shortcomings, and future research directions of the proposed method are also discussed in this paper.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: