In this work, we focus on framed slotted Aloha (FSA) and passive ultra high-frequency radio frequency identification\r\nmulti-antenna systems with physical layer collision recovery. We modify the tags slightly by adding a so-called\r\nââ?¬Ë?postpreambleââ?¬â?¢ that facilitates channel estimation. Furthermore, we investigate the throughput performance of\r\nadvanced receiver structures in collision scenarios. More specifically, we analyse the throughput of FSA systems with\r\nup to four receive antennas that can recover from a collision of up to eight tags on the physical layer and\r\nacknowledge all tags involved in that collision. Due to the higher collision recovery capabilities, the frame sizes can be\r\nsignificantly reduced, and thus, the throughput can be increased. We also derive analytically optimal frame sizes,\r\ngiven that a certain number of collisions can be resolved. We further study the constraints to the throughput due to\r\nthe structure of our receiver and channel estimation for different collision scenarios. Furthermore, we propose a novel\r\ncollision recovery method with two phases: first, a successive interference cancellation and, second, a projection of\r\nthe constellation into the orthogonal subspace of the interference. Additionally, the inventory time, i.e. the number of\r\nslots necessary to successfully decode all tags in the reader range, is calculated and compared for different receiver\r\ntypes. A validation of our theoretical predictions is achieved by means of simulations. We show that by our proposed\r\nmethods, we can realistically achieve more than ten times higher throughput or, equivalently, a reduction of the\r\ninventory time by more than 90%.
Loading....