Academic Journal of Computing & Information Science, 2026, 9(5); doi: 10.25236/AJCIS.2026.090508.
Cui Lizhi1,2, Jiang Zhen1,2
1Country School of Electrical Engineering and Automation, Henan Polytechnic University, Jiaozuo, China
2Henan Key Laboratory of Intelligent Detection and Control for Coal Mine Equipment, Jiaozuo, China
Aiming at the issues of inadequate positioning accuracy in RFID indoor positioning technology, a localization method based on Convolutional Neural Network (CNN) feature extraction and Weighted K-Nearest Neighbors (WKNN) is proposed. This method integrates the fingerprint features of Received Signal Strength (RSSI) and Phase. Firstly, the CNN is trained for feature extraction and weight learning using the offline RSSI/Phase dataset. Then, during online matching, the RSSI/Phase data collected at the test points are processed by the trained CNN for feature extraction. Finally, the WKNN algorithm is employed to predict the labeled location of the test points. In the WKNN prediction stage, a Genetic Algorithm (GA) is adopted to optimize the weights of RSSI and Phase, respectively. The experimental results of the proposed method are compared with those of CNN, WKNN, and LANDMARC localization methods, and the impact of different CNN architectures and fingerprint datasets on positioning accuracy is analyzed. The experimental results show that the proposed method achieves significant improvement in both positioning accuracy and adaptability.
Radio Frequency Identification; Convolutional Neural Network; Weighted k-Nearest Neighbors; Genetic Algorithm
Cui Lizhi, Jiang Zhen. Deep Feature Extraction and Optimal Neighbor Selection Method for RFID-Based Localization. Academic Journal of Computing & Information Science (2026), Vol. 9, Issue 5: 67-76. https://doi.org/10.25236/AJCIS.2026.090508.
[1] Wu, L., Ren, J., Li, Y., et al. RFID data-driven product processing cycle prediction for flexible manufacturing workshops[J]. Mechanical Design and Manufacturing, 2025, (01): 265-271.
[2] Wang, S., Shi, C., Zeng, D. Y., & Zou, Y. S. Research on real-time positioning method of AGV based on low-cost IMU and RFID technology. Mechanical Design and Manufacture, 2021,(05): 269-272.
[3] Y. Shao and Q. Xiang, "Intelligent Vehicle Positioning Method Based on UWB and GPS Fusion," 2024 4th International Conference on Electronic Information Engineering and Computer Communication (EIECC), Wuhan, China, 2024: 447-450.
[4] Z. Turgut and A. G. Kakisim, "An explainable hybrid deep learning architecture for WIFI-based indoor localization in Internet of Things environment", Future Gener Comput Syst,2024,7:50-63.
[5] M. D. Jovanovic and S. M. Djosic, "Analysis of Indoor Localization Techniques", Proc. ICEST, 2023:219-222.
[6] S. Böller, A. Bödder, J. Riegler, "SHF RFID based Localization: Combining Time-of-Flight and Angle-of-Arrival," 2024 IEEE International Conference on RFID Technology and Applications (RFID-TA), Daytona Beach, FL, USA, 2024:70-73.
[7] R. Shu, J. Feng, Y. Xu, "Indoor Following Model Based on TDOA Positioning and Q-Learning Algorithm," 2023 2nd International Conference on Automation, Robotics and Computer Engineering (ICARCE), Wuhan, China, 2023: 1-5.
[8] Shang, L., Guan, W., & Gong, R. An improved localization algorithm based on clustering optimization and adaptive KNN[J]. Transducer and Microsystem Technologies, 2023, 42(3): 136-139.
[9] Y. Xie, Z. Wang, and S. Zheng, ‘‘On RFID positioning base on LANDMARC and improved algorithm,’’ in Proc. Chin. Control Conf., Beijing, China, 2010:4831–4836.
[10] C. Peng, H. Jiang and L. Qu, "Deep Convolutional Neural Network for Passive RFID Tag Localization Via Joint RSSI and PDOA Fingerprint Features," in IEEE Access, 2021,9: 15441-15451.