�� FDMA. Transmission channel is split up into different carrier frequencies that are simultaneously available. It requires a complex receiver at the reader.�� CDMA. Tag IDs are multiplied with a pseudo-random sequence before transmission. It demands elevated power consumption.�� TDMA. Transmission channel is divided between the participants chronologically.In RFID systems, TDMA procedures are the most used techniques in RFID and they have the largest group of anti-collision methods. These can be categorized in: Aloha-based protocols which are probabilistic, tree-based protocols which are deterministic, and hybrid protocols which are a mixture of the previous ones [6].2.1.1. Aloha-Based ProtocolsThe aloha protocol is the origin of the Aloha-based protocols.
An improvement of that is the slotted-Aloha, which introduces the slot concept. A slot is a period of time during which the reader sends a command and the tags respond to the reader. Slotted-Aloha divides time into slots thus improving its throughput [3]. Later, framed-slotted-Aloha (FSA) is developed. In FSA all nodes must respond choosing a slot into a fixed length frame (a group of slots). As the throughput of the FSA decreases with the increase of the total amount of nodes, a dynamic-framed-slotted-Aloha (DFSA) is developed [7,8]. This protocol changes the length of the frame dynamically using an estimator to adjust the frame size. Some protocols like I-Code [8] change the frame size at the end of the last frame slot, and other algorithms, as the EPC C1G2 Slot Counter [9], adjust the frame size after a slot transmission.
Early cited, the tag starvation problem affects probabilistic algorithms, this is a tag that may not be correctly read during a reading cycle. Besides, estimation involves some disadvantages [13]: an increase in the computational cost of the reader [8] and the tag [20]; an error that degrades the efficiency; and lastly, an initial frame length cannot be set according to the estimated number of tags.2.1.2. Tree-Based ProtocolsThe main feature of this kind of protocols is that they are deterministic. This is that all tags in the reader’s interrogation zone are going to be identified. These protocols usually have simple design tags and work well with uniform set of tags but are slower than Aloha-based protocols. They can be categorized into [6]: Tree Splitting (TS), Query Tree (QT), Binary Search (BS) and Bitwise Arbitration (BTA).
A virtual tree to organize and identify each tag was firstly proposed by the authors Dacomitinib of the TS in [10]. This algorithm splits the set of tags in B subsets (B > 1) after a collision. These subsets become increasingly smaller until they contain one tag. The TS does not need clocking circuitry but they must maintain a counter, so if a tag get discharged, it loses cycle information. Moreover, the QT is proposed in [11].