Designing Fast Identification Algorithm in the RFID System for Known Tag Sets

快速辨識無線射頻電子標籤演算法之設計

無線射頻辨識是一種自動辨識技術，這種技術利用無線電頻率去獲得電子標籤內的資訊。在多個電子標籤的環境中，數電子標籤同時回應讀取器將會成碰撞的問題而危及識別的速度。故此，某些為調解衝突與改善讀取時間的機制是必要的。本文提出一個新的機制以改善上述問題。另外，此機制可以被訓練來加快辨識已知的電子標籤。僅管有一些相似的機制能夠達到同樣的效果，但此機制可以識別未知的電子標籤等罕見狀況。在模擬情況中，本文針對數個電子標籤防碰撞機制進行績效評估。更精確而言，本文著重在完成讀取所須的時間和訊息數。我們觀察得知本機制能有效減少讀取時間和讀取訊息。在最佳情況下，相較於詢問樹演算法本機制能節省50％的時間；相較於二元樹演算法本機制能節省70％的時間。這些改善不只加快了讀取的速度，也節省了讀取器中之電力。

RFID is an automatic identification technology which exercises the radio frequency to retrieve the information saved in electronic tags. In a multi-tag environment, the collision problem caused by a number of tags simultaneously responding to a reading will jeopardize the speed of identification. Some scheme is necessary to reconcile the conflict and improve the reading time. A novel scheme is proposed in this paper to address this collision problem. In addition, our scheme can be trained to accelerate the identification of a subset of the known tags. Although there are similar schemes capable of achieving this function, our proposal accommodates rare presence of those tags outside the known set. In the simulation, we evaluate the performance of several tag anti-collision protocols. Specifically, we focus on the total reading time and the amount of messages required accomplishing the reading process. We observe that our scheme can effectively reduce both the reading time and the number of reading messages. In the best case, the proposed scheme saves 50% and 70% time when compared with the query tree algorithm (QTA) and binary tree algorithm (BTA) methods respectively. These improvements not only accelerate the reading speed but also save the on-board power in the reader.