A bit-error based capture model for EPCglobal RFID systems

The tag inventory process in systems according to the EPCglobal UHF RFID Air Interface Protocol is based on framed slotted ALOHA (FSA). For FSA, the well-known maximum throughput of 0.37 is reached when the number of allocated slots is equal to the number of contending tags. These values do not hold when captures happen in the system, i.e. when information can be recovered from collision slots. In this work, we propose a new capture model that allows us to calculate the capture probability in a noisy channel, depending on the occupation number of a slot. We will apply the model to AWGN, as well as fading channels, and show the expected capture probabilities for these cases. We will further show how the proposed model can be used to estimate the throughput improvement in a system when forward error correction is employed during the tag inventory.