Robust statistical methods for detection of missing RFID tags

The technology of radio frequency identification is being deployed in many applications, such as logistics and inventorying. However, a generic problem in all RFID systems is to ensure that the RFID readers can reliably read a set of RFID tags, such that the probability of missing tags stays below some acceptable value. This article introduces statistical methods to deal with the problem of missing RFID tags. These methods are applied at the reliability layer, which initiates multiple reader sessions by invoking certain MAC-layer arbitration (anti-collision) protocols to collect tag responses. The reliability layer obtains a running estimate of the probability of having at least one tag missing. This estimate is used to detect if an additional reader session is required. We present several estimators, which can be used at the reliability layer to obtain an estimate of the probability of missed tags after R reader sessions have been carried out. These estimators are derived under idealized assumptions. However, when tested under more realistic conditions, which violate these ideal assumptions, the estimators exhibit high robustness and provide a very close approximation of the true probability of missing tags.