A Non-Contact Indoor Positioning Sensing Strategy with Commercial WiFi Devices

With the proposal of the world's first integrated sensing and communication (ISAC) standard, IEEE 802.11bf, the wireless fidelity (WiFi) indoor positioning topic has attracted more attention. To address the problem of multi-person online positioning, in this paper, an indoor positioning sensing strategy is proposed, including an optimized preprocessing process and a novel machine learning (ML) method called NCA-Based-ML. Experimental results show that compared with support vector machine (SVM), multilayer perceptron (MLP), naive bayes (NB) and random forest (RF), the proposed method has higher localization accuracy and is robust in multi-scene circumstances.