Answers are grounded in this paper and cite their sources.
Indoor pedestrian tracking systems utilizing Bluetooth Low Energy (BLE) technology typically rely on the Weighted K-Nearest Neighbor (WKNN) algorithm for position estimation. However, this method suffers from high computational load in fingerprint matching and significant spatial ambiguity in positioning results. To solve these problems, this paper introduces an improved version of the WKNN algorithm, termed the Improved Self-Adaptive WKNN (ISAWKNN), by incorporating a Dual-Factor Similarity-based K-means (DFS-K-means) clustering approach and a dynamic adaptive K-value selection strategy. Additionally, an improved particle filter (IPF) is developed that uses the Subtraction-Average-Based Optimizer (SABO) to refine the particle distribution. Consequently, the ISAWKNN and IPF algorithms are integrated to form a two-module algorithm named ISAWKNN-IPF, where the initial position estimate from the ISAWKNN algorithm serves as a measurement input for the IPF algorithm to achieve a precise position estimate. Experimental results demonstrate that the ISAWKNN-IPF algorithm not only reduces positioning error by approximately 50% compared to the traditional WKNN approach but also achieves superior performance relative to several state-of-the-art methods. The proposed algorithm has the capability to enhance the quality of existing indoor location-based services.
No takes yet. Share an insight, caveat, or question.
Zhou et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: