Indoor Positioning for the Elderly Based on Fuzzy C-Means Clustering and KNN Algorithm

Location-based services (LBS) for elderly care is a trending topic in smart homes. The main concern is the precise positioning for elderly individuals. To address the issue of poor positioning accuracy and lengthy processing time associated with traditional UHF radio frequency identification (RFID) indoor positioning algorithms that use geometric methods, a new positioning method has been proposed. This method combines fuzzy clustering and k-nearest neighbor (KNN) ratio algorithms and has been applied to indoor positioning of the elderly. RFID is used to detect the location of elderly individuals and collect their location information. Fuzzy C-means clustering replaces traditional hard clustering algorithms. The algorithm estimates feature points in the cluster center reasonably based on the membership degree of reference points. This increases the difference between reference points and reduces the complexity of feature matching. Additionally, the KNN algorithm is combined with the clustering algorithm to perform fine matching. The experimental results show that the proposed algorithm can enhance positioning efficiency by 12.3% compared to the traditional algorithm.