PLACE: Physical layer cardinality estimation for large-scale RFID systems

Yuxiao Hou, Jiajue Ou, Yuanqing Zheng, Mo Li
2015 2015 IEEE Conference on Computer Communications (INFOCOM)  
Estimating the number of RFID tags is a fundamental operation in RFID systems and has recently attracted wide attentions. Despite the subtleties in their designs, previous methods estimate the tag cardinality from the slot measurements, which distinguish idle and busy slots and based on that derive the cardinality following some probability models. In order to fundamentally improve the counting efficiency, in this paper we introduce PLACE, a physical layer based cardinality estimator. We show
more » ... at it is possible to extract more information and infer integer states from the same slots in RFID communications. We propose a joint estimator that optimally combines multiple subestimators, each of which independently counts the number of tags with different inferred PHY states. Extensive experiments based on the GNURadio/USRP platform and the large-scale simulations demonstrate that PLACE achieves approximately 3∼4× performance improvement over state-of-the-art cardinality estimation approaches.
doi:10.1109/infocom.2015.7218579 dblp:conf/infocom/HouOZL15 fatcat:cbcil4u5fvhk5k6ol5btgoqbz4