| The RFID technology has reached a peak since the adoption in 2005 of the EPCglobal Class 1 Gen 2 standard (EPC-C1G2). RFID has replaced other automatic identification systems like barcodes in some sectors. Besides, it has attracted the attention of scholars in RFID technology. The future of RFID looks promising, with a wide range of application fields, such as home automation, checkout systems, healthcare, public transport, and so forth. Currently, developments are underway in various areas of RFID: physical design, development of middleware, anti-collision algorithms, optimal network planning, etc. These investigations seek to improve RFID performance and reduce costs of deployment and operation.;The contribution of this thesis is twofold. First, the anti-collision mechanisms are studied in depth, and new solutions are suggested both for active and passive RFID systems. Second, the co-existence of several readers in dense reader environments is also investigated in this work. The major outcomes are described next.;One of the main goals of this thesis has been to analyze the performance of the anti-collision mechanisms in passive RFID. Standards as well as relevant proposals based on Frame Slotted Aloha (FSA) have been modeled using Discrete Time Markov Chains (DTMC). From these models the most relevant performance metrics in the identification process are computed. Namely, the Mean Identification Time (and throughput) in static scenarios, and the Tag Loss Ratio for semi-static and dynamic cases. Based on these results, optimal configuration criteria are provided.;Moreover, since FSA algorithms are inefficient under variable populations of tags, Dynamic FSA (DFSA) are regarded as an efficient solution to cope with this issue. A comprehensive survey of DFSA mechanism is developed in this work, including a DTMC characterization of the optimal DFSA mechanism, and a comparative analysis of current proposals performed via simulation. In addition, we propose a new anti-collision algorithm, the Multi-Frame Maximum-Likelihood DFSA algorithm (MFML-DFSA), which outperforms the previous ones, and has a lower computational cost which enables a feasible implementation in current readers.;On the other hand, for active tags, the goal of minimizing power consumption is added to the design constraints of anti-collision protocols. Despite the increased capabilities of these equipments, conventional FSA approaches have been the most extended so far. Since active devices already integrate carrier sensing, two new anti-collision algorithms are suggested based on this observation. A non-persistent Carrier Sense Multiple Access (CSMA) mechanism based upon the quasi-optimal Sift distribution, which is indented to minimize the mean identification time and to reduce the activity periods of the tags. And a p-persistent CSMA, which is able to minimize energy waste switching off a majority of tags during contention slots.;Finally, dense RFID reader environments are studied in this doctoral thesis. In some installations, a single reader is not enough to cover a large identification area, or several identification gates may be present. Hence, several readers are required, and the overall performance is negatively affected by two types of collisions: Reader to Tag Collisions (RTC) and Reader to Reader Collisions (RRC). Our focus was on the review of the scheduling solutions aimed at minimizing these types of collisions. The performance of the different mechanisms is studied in order to provide comparative results of efficiency and network usability. The drawbacks extracted are used to suggest the key design properties of an efficient scheduler. |
