When the radio frequency identification (RFID) system inventories multiple tags, the recognition rate will be seriously affected due to collisions. Based on the existing dynamic frame slotted Aloha (DFSA) algorithm, a sub-frame observation and cyclic redundancy check (CRC) grouping combined dynamic framed slotted Aloha (SUBF-CGDFSA) algorithm is proposed. The algorithm combines the precise estimation method of the quantity of large-scale tags, the large-scale tags grouping mechanism based on CRC pseudo-random characteristics, and the Aloha anti-collision optimization mechanism based on sub-frame observation. By grouping tags and sequentially identifying them within subframes, it accurately estimates the number of remaining tags and optimizes frame length accordingly to improve efficiency in large-scale RFID systems. Simulation outcomes demonstrate that this proposed algorithm can effectively break through the system throughput bottleneck of 36.8%, which is up to 30% higher than the existing DFSA standard scheme, and has more significant advantages, which is suitable for application in large-scale RFID tags scenarios.

Addressing the challenges of passive Radio Frequency Identification (RFID) indoor localization technology in Non-Line-of-Sight (NLoS) and multipath environments, this paper presents an innovative approach by introducing a combined technology integrating an improved Kalman Filter with Space Domain Phase Difference of Arrival (SD-PDOA) and Received Signal Strength Indicator (RSSI). This methodology utilizes the distinct channel characteristics in multipath and NLoS contexts to effectively filter out interference and accurately extract localization information, thereby facilitating high precision and stability in passive RFID localization. The efficacy of this approach is demonstrated through detailed simulations and empirical tests conducted on a custom-built experimental platform consisting of passive RFID tags and an R420 reader. The findings are significant: in NLoS conditions, the four-antenna localization system achieved a notable localization accuracy of 0.25 m at a distance of 5 m. In complex multipath environments, this system achieved a localization accuracy of approximately 0.5 m at a distance of 5 m. When compared to conventional passive localization methods, our proposed solution exhibits a substantial improvement in indoor localization accuracy under NLoS and multipath conditions. This research provides a robust and effective technical solution for high-precision passive indoor localization in the Internet of Things (IoT) system, marking a significant advancement in the field.