Article Preview
Top1. Introduction
The “Internet of Things” (IoT) consists in a vision where objects become part of the Internet: every physical object has its unique identification, and is accessible from the network, providing an expanded Future Internet (Coetzee & Eksteen, 2011). In this scenario it is expected, for example, that the users use the Internet to check the location of people and their belongings within a pre-defined area. Thus, it is needed that readers periodically send requests to store the data about people and objects.
RFID (Sheng, Li, & Zeadally, 2008) is a key technology of the IoT, since small passive RFID tags allow to link millions and billions of physical products with the virtual world (Wu, Zeng, Feng, & Gu, 2013). When a large number of tags are used, there is a high probability that there will be more than one tag within a reader zone at some time. When the tags transmit their responses simultaneously to the reader, collisions will happen because the communication is done over a shared wireless channel. Therefore, RFID tag anti-collision mechanisms will play an important role in the IoT (Wu et al., 2013; Chunli & Donghui, 2012; Jia, Feng, Fan, & Lei, 2012).
Many efforts have been made in the literature to improve the performance of anti-collision protocols (Wu et al., 2013; Leonardo & Victor, 2012; Felemban, 2012; Jia, Feng, & Yu, 2012; Guilan & Guochao, 2010; Zhong, Chen, Wu, & Pan, 2012; Jian Su and Guang-Jun Wen, 2012; Chunli & Donghui, 2012; Han, Park, & Lee, 2012). However, little research has been conducted for IoT scenarios (Guilan & Guochao, 2010). According to (Namboodiri, DeSilva, Deegala, & Ramamoorthy, 2012) there are several disadvantages of using the Q algorithm, the standard Algorithm for Class 1 Generation 2 RFID systems, because too many packets need to be transmitted, in a single identification process, between the reader and tags. This process generates considerable overhead and increases the power consumption, since the energy consumption is proportional to the number of actions of the readers (Klair, Chin, & Raad, 2009). In an IoT scenario, where readers regularly consult the tags and make them available on the Internet, the problem is compounded, generating even more overhead.
The aim of this paper is to propose a mechanism to increase the chances to meet QoS re-quirements for IoT tracking scenarios, whose nodes are RFID tags. The mechanism reduces the number of delay slots (idle and collision), and consequently the amount of messages exchanged in the network, when compared to the Pure Q Algorithm and to the Binary Tree Slotted Aloha - BTSA algorithm. The proposed mechanism is based on the principle that the tags do not need to reply to all reader queries if they don’t change their locations.
The proposed mechanism had its performance evaluated through simulated experiments in the simulator ns-2, and the results confirm its effectiveness. For instance, in a scenario with 500 tags, and using the proposed mechanism, there was a reduction in the number of delay slots of about 24%–43% when compared to the classical mechanisms.
The contributions of this paper are:
- •
A mechanism to decrease the delay slots in RFID systems used to deploy IoT applications;
- •
As a consequence of the first contribution, a decrease of energy consumed by readers and an increase to the chances to meet QoS requirements.
Besides the mechanism that reduces the number of delay slots, this paper is different from those found in the literature and advances the state of the art because it performs experiments simulating real IoT scenarios (Welbourne et al., 2009) with an RFID ns-2 module, varying the number of tags, and because the proposed mechanism is compatible with the global standard communication protocol for passive RFID tags.