| The Internet of Things (IoT) is included in the national development strategies in 2010. Furthermore, network-related research works have been strongly supported, and are being rapidly advanced steadily. Some IoT systems continue to achieve a variety of applications. Similarly to other network systems, the IoT system adopts hierarchical architecture including perception layer, network layer, and application layer. The perception layer mainly uses such wireless communication technologies such as radio frequency identification (RFID), wireless sensor networks and infrared to complete the acquisition of object-related data, and the RFID is a main technology of the perception layer. In the IoT system, the network layer adopts computer network layer technology, and through the layer data from the perception layer is transmitted to the data center. The application layer is for analysis or processing of data in the center in order to provide information for users.The RFID system consists of readers and numerous tags, and these tags passively response to commands from the reader. When multiple tags simultaneously transmit identity data to the reader, data collision will occur resulting in no reception of identity data. This data collision reduces tag identification speed, which is the tag collision problem. Furthermore, the data collision also leads to the missing tag identification phenomenon in the dynamic environment. As one of the three key technologies of RFID, tag anti-collision protocols can reduce the collision rate. To solve the tag collision problem, there have been many tag anti-collision protocols, however, their identification efficiency is still low. With fast development of IoT, these protocols are unable to meet requirement of massive, fast and reliable tag identification under the dynamic environment. Therefore, development of high-performance tag anti-collision protocol will become an urgent task.Existing tag anti-collision protocols assume that no tag arrives in identification area of the reader in the identification process, so the reader can effectively estimate the number of tags in current stage using identification cases of the previous stage and obtain optimal operation parameters of the anti-collision protocol. However, new tags may arrive in each stage of the identification process in the dynamic environment, so it is difficult for the existing protocols to effectively estimate the number of tags in the current stage. Therefore, these protocols are not able to optimally operate in the dynamic environment, correspondingly, there exists decline of performances and tag missing identification phenomenon. Summarization and further analysis of tag missing identification show that there are two main reasons for inapplicability of these protocols to the dynamic environment. One reason is that these protocols determine their parameters without considering new arrival tags, and other reason is their low identification efficiency. Thereforce, this dissertation proposes two dynamic framed slotted aloha protocols (algorithms) from the perspective of the tag estimation in the dynamic environment. This dissertation also proposes two high-efficiency anti-collision protocols identifying the vast majority of tags at each stage basing on the reservation mechanism. Furthermore, a RFID system consisting of multiple readers expands the tag identification coverage with higher throughput rate than one with a reader, thus, the missing rate can be reduced to a minimum. From the above three aspects, a series of research works have been carried out, and the corresponding achievements are as follows:(1) Aiming at solving the problem that the identification efficiency of classical Dynamic Framed Slotted Aloha protocols decrease under the dynamic environment because these protocols do not calculate the new arrival tags, a new anti-collision protocol named CDFSA (Continuous Dynamic Framed Slotted Aloha) supporting continuous arrival tags is proposed. For the environment, the tag identification process model of the continuous tag arrival is established. Furthermore, on the basis of the process model, the Poisson process superposition principle and the frame efficiency optimization method are used to determine frame length resulting in efficient tag identification. Additionally, for the steady work conditions of the proposed protocol, the upper limit of tag arrival rate is obtained based on the stochastic service system theory.(2) An arrival rate prediction algorithm with low computational complexity and determination methods of protocol parameters are proposed by researching the performance decrease problem of the widely used EPCglobal Class 1 Generation 2 (EPC-C1G2) system under the dynamic environment. On the basis of the prediction algorithm and the determination methods, a new tag anti-collision algorithm named G2-CDFSA (Continuous arrival Dynamic Framed Slotted Algorithm for the EPC-C1G2) meeting the EPC-C1G2 system under the dynamic environment is proposed.(3) An anti-collision protocol adopting reservation mechanism (Reservation Slot with Multi-Bits Aloha, RSMBA) is proposed. It can allocate time slots among tags according to their needs. The RSMBA divides each inventory round into the reservation procedure and the identification procedure by adopting a new reservation mechanism during the process of tag identification. Collision frame slots are significant reduced and empty frame slots are eliminated at very low cost of reservation time overhead. A mathematical model for optimization of the protocol efficiency is established. Furthermore, the optimal protocol parameters are obtained by numerical computation method based on the model. The identification efficiency of the RSMBA is close to 91.28%, so the proposed protocol is suitable to tag identification requirement under the dynamic environment. Furthermore, an estimation method of the tags’number is proposed for the protocol, and its identification performances are evaluated based on the estimation method. In addition, analysis of the tag circuit complexity show its lower complexity than one of the EPC-C1G2 system.(4) A reservation mechanism anti-collision protocol named SARCE (Slot Assignment based on Reservation with Capture Effect) is proposed considering the capture effect. Optimal operation parameters of the proposed protocol are obtained by establishing its mathematical analysis model. Performance evaluation shows that the identification performance of the SARCE is significantly higher than the existing protocols, and its efficiency reaches 92.6%. Furthermore, a prototype system based on the reservation mechanism is developed to verify the validity of the RSMBA and the SARCE.(5) Considering multiple readers layout requirements of the dynamic environment, this dissertation puts forward a RFID data integration middle-ware as the foundation for further research of tag identification under the dynamic environment in order to rapidly and reliably identify tag in the large scale scene. The middle-ware is also as application scheme of tag anti-collision protocols. In addition, considering the classic RFID application scenarios, the embedded monitoring system is developed on the basis of hierarchical architecture. |
PDF Full Text Request