| Summary: | 博士 === 國立臺灣科技大學 === 電子工程系 === 98 === Radio Frequency Identification (RFID) technology has generated huge interest in recent years. It is believed that traditional barcodes will be widely replaced by RFID because it can provide a more powerful and convenient electronic identification service. In 2004, Ari Juels introduced a proof for the simultaneous reading of two RFID tags and named it as the “Yoking Proof”. Afterward, the yoking proof had been extended to prove the simultaneous presence of an arbitrarily large number of tags within the broadcast range of a reader, which is generally called the “Grouping Proofs”. This dissertation extends the RFID grouping proofs from multiple tags in a single group to multiple tags in multiple groups.
In this dissertation, a comprehensive survey of the existing RFID grouping proof protocols is provided. In addition, classifications of grouping proof protocols according to yoking or non-yoking, tag reading order correlation or uncorrelation, and random or select response properties are provided.
Some of existing protocols adopted the concept of yoking proof to construct the grouping proofs. That is, a tag is required to wait for a specified tag response before any further actions. Such tag reading order correlation in a grouping proof is inefficient and can raise the failure rate of verification. In this dissertation, the Saito and Sakurai’s protocol is modified in which a grouping proof protocol with tag reading order uncorrelation is proposed. The proposed protocol maintains a fixed length of the grouping proof and improves the overall efficiency during the grouping proof verification process.
Moreover, most of the previous grouping proof protocols designate that a reader is responsible for both the queries and computations of a proof for a verifier. There is no information for the reader to judge the completeness of the proof in advance. In addition, DoS is also possible since a malicious tag can obstruct the generation of a legitimate proof or cause a useless proof to be created. To overcome these problems, an online “Select-Response” grouping protocol is proposed in which the verifier is actively involved instead of just waiting a proof from the reader. With this fundamental change, the verification process becomes simple and flexible. In addition, the proposed protocol can perform multi-group verification. It also possesses collision-free, missing tag identification, and tag reading order uncorrelation properties.
In case of a group with large number of tags, the unreliability of radio wave communication could potentially cause high failure rates in the currently existing grouping proofs. Therefore, a novel way of performing the grouping proof is proposed in which the “Dynamic Binary Tree Anti-collision Algorithm” is employed to subgroup the tags. This solution combines the “Tag Reading Order Uncorrelation” idea and the “Select-Response” scheme from my proposed protocols. Each subgroup will generate its own “Subgrouping Proof”. The yoking proof is then generated between subgroups and finally the grouping proof of the whole group is assembled. Thus, the efficiency of the whole group verification will be improved.
Finally, an application of the Select-Response grouping proof concept is realized in a supply chain system. Two protocols are also presented which includes the tags reading protocol with mutual authentication and the updating protocol. The handover process of group tags is also modified that made it more reasonable. The proposed protocol satisfies most of the requirements of supply chain systems and can be applied on the multi-batch supply chain systems.
|
|---|
