A Multi-Agent-based Heterogeneous RFID Reader Management

• Main Authors: Cheng-Yi Peng, 彭正儀
• Other Authors: An-Chi Liu
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/87478549317744276640

碩士 === 逢甲大學 === 資訊工程所 === 97 === Automatic identification systems such as RFID (radio frequency identification) has been developed and become popular for recent years. It has been used in various application domains include manufacturing, health care, defense, logistics, etc. RFID technology has several advantages over other automatic identification systems. It does not require line-of-sight alignment, multiple RFID tags can be read simultaneously, RFID reader offers a large reading range for information collection. For the enterprises that perform their transactions depending on RFID technology, the management of RFID devices is of increasing important. It is since a reader’s fault could be often costly damages to the entire enterprises. In this thesis, a RFID reader management mechanism for effective and automated health diagnosis is provided. We use traditional network management concepts and add proxy agent technology to establish the management of heterogeneous RFID readers. We choose KPIs (Key Performance Index) as the health degree of RFID network such as CPU usage and Tag throughput via network management tools. To diagnose the anomalies of the RFID network, a fuzzy-based mechanism was applied. In addition, we use multi-agent technology to resolve the performance load for RFID middleware in the RFID network. To verify the feasibility of our mechanism, we design an experimental environment for collecting KPIs from each network segment. In our experimental results, the management system could detect anomalies and achieve to load balancing in the RFID network effectively.