標題: | 用於Auto-ID環境下減少碰撞的機制 Mechanisms for Reducing Collision in an Auto-ID-based Environment |
作者: | 何丁武 Ding-Wu Ho 羅濟群 Chi-Chun Lo 資訊管理研究所 |
關鍵字: | Auto-ID;RFID;碰撞解決協定;Tag與Reader溝通協定;Auto-ID;RFID;Anti-collision protocol;Tag protocol |
公開日期: | 2003 |
摘要: | Auto-ID是一個新的電子化物品辨識系統,其可以用來取代傳統的Bar Code來辨別貨品。他架構的底層採用RFID(Radio Frequency Identify)的技術,使得物品上的標籤(底下均稱為Tag),利用無線電波的方式,被讀取器(底下均稱為Reader)讀取到並送至後端系統加以辨識與應用。
於Auto-ID 規格中,Reader跟Tag之間的碰撞機制主要是採用Binary-Tree Protocol(底下均稱BT)。BT是個類似喊號的機制,由Reader在BT上來回搜尋喊號。然而這個機制在當Tag數量很多時,其辨識效率不彰,若要將Auto-ID應用在一個較為動態的環境,則需要去降低碰撞發生的次數,進而改善其效率問題。
在此,我們嘗試採用以CSMA/CA為基準的機制與混合式的BT機制,除了改善原先BT機制的碰撞次數外,辨識效率亦可所提升。基於CSMA/CA的機制為因應Auto-ID的環境,對CSMA/CA做個修改,讓其得以適用;混合式BT機制,則是將BT與QT兩者混合,在BT連續碰撞的地方,改採QT的方法來向Tag查詢。
模擬結果,基於CSMA/CA的機制不會因為延遲了Time Slot的時間,而使得媒體總存取數與原先BT機制相比之下有所增加,由此亦可證實辨識效率並不因此而有所降低;另外,在同時間於Reader讀取範圍裡的Tag數若不大於512個的情況下,基於CSMA/CA的機制比混合式BT機制更為適用;反之,若有所超過,因為基於CSMA/CA的機制會產生辨識漏失的現象,故混合式BT機制會比較合適。本文所提機制最大貢獻在於使Reader可更有效率地讀取到Tag的ID資訊,且在模擬實驗中亦得到一個驗證。這樣的貢獻可使Auto-ID更適用於動態的環境。 Auto-ID architecture proposed a new technology to identify products and goods embedded special designed tag, instead of conventional identification triggered by Bar Code. This new technology of identification practices RFID proposed standard coordinated with the distribution information structure. The anti-collision mechanism addressed in Auto-ID specification adopts a similar procedure of calling the roll. The inefficiency and higher collision probability will occur in the procedure of roll call when many products and goods must be identified. According above, the mechanisms proposed here must be developed to foster the efficiency of identification. The proposed anti-collision mechanisms, adapted CSMA/CA and combined BT scheme, try to boost the original anti-collision mechanism and reduce the inefficiency of identification. According to the simulation, adapted CSMA/CA mechanism is suitable when the amount of tag is not larger than 512; otherwise, combined BT scheme is more suitable than adapted CSMA/CA mechanism since adapted CSMA/CA would cause the identifying loss. Verified with the conclusion of simulation, the mechanisms proposed here can foster the efficiency of identification as anticipated. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT009134524 http://hdl.handle.net/11536/58212 |
顯示於類別: | 畢業論文 |