Efficient Access Algorithms for Dynamic Many-tag Passive RFID Storage Systems

In this work, we investigate efficient algorithms for accessing information stored in a large number of RFID tags. Interrogators, equipped with RFID scanners, regularly arrive at the tag system, read from and write messages to the tags, and then leave. We develop techniques for the interrogators to read the newest messages from the tags as quickly as possible, since newer information is often the most relevant and interesting. We borrow ideas from aloha and query tree, two traditional singulation schemes, to form our many-tag access algorithms. Our query tree-based algorithms access tags faster, but are not as robust as our aloha-based algorithms. We study two scenarios. In the static scenario, the tag population is fixed. Here, it is naive to initially singulate all the tags, and then query them afterward to read the newest messages. Our results indicate that better-performing algorithms instead progressively segment the tag population at each round to find the newest messages. In the dynamic scenario, tags are continually arriving and departing, causing information to quickly disappear. We combat this by encoding messages, dividing the coded bits into multiple chunks, and then spreading them across multiple tags. This requires us to access a large number of tags when reading. Therefore, our results show that better-performing algorithms initially singulate all the tags (as opposed to the static scenario), and then continually query them afterward individually for data chunks in order to recover the newest messages.