RFID is one of the most pervasive technologies nowadays. Its main attractiveness lies on the ability of identifying objects with no contact or a direct line of sight between the participant devices. An RFID system comprises two main components: an RFID network and a storage system.There are important research directions on each of these subjects, which together allow to improve the performance of RFID technology. The RFID network is the core of an RFID system since in this component resides the whole information of interest. Each object has attached an RFID tag with a unique ID. The RFID reader is the central part of the RFID network, its main function is to manipulate the information from the tags. Therefore, several processes are executed in an RFID network in order to manage such information. Normally, each process is related to an application and its requirements, causing an on-demand execution of a particular process. Moreover, the Identification Process is always executed in an RFID network, which consists of collecting the tag IDs by means of the interaction between an RFID reader and the tags in the interest area. In this thesis, we call such a scenario the centralized approach. Improving the performance of the identification process is vital for the whole RFID system. Most of the research work has been made around the centralized approach for RFID networks; however, we identify in this work that this not only limits the execution of the identification process, but also limits the impact of the future mechanisms.A new architecture for RFID tag identification has arised in recent years. Such approach proposes to evolve from the original centralized approach to one that in some way groups tags in different coverage zones and executes the identification process in a distributed way. The main goal is to improve the identification delay of the system, and eventually the associated implementation costs. In this work, we call such architecture as the distributed approach. We also prove in this work that the contributions made so far in this direction still need lots of improvements. Particularly, we highlight the need of new communication protocols for the distributed approach, as well as the need of analyzing their performance.In this PhD thesis, we review a wide set of RFID identification mechanisms and we evaluate their performance in terms of identification delay. Such parameter is the time needed to identify the whole set of tags present in an interest zone. Then, we describe in detail how to improve such performance parameter for each of the analyzed mechanisms. Finally, we propose an algorithm to help designers choose an identification mechanism according to application requirements. i 6. Mecanismos de identificación para sistemas RFID pasivos: análisis, comparación y clasificación 6.1.