Study On Anti-Collision Technology Of RFID System In Internet Of Things

With the rapid development of network technology, communication technology andintelligent embedded technology, the "Internet of Things"(IoT) frequently appears in ourworld. As an important part of the next generation network, the IoT is widely concerned bythe governments, academia and industry. The purpose of IoT is to achieve interconnection andinformation transfer between all things, and to achieve this, the most important is to recognizethe person or thing in IoT, which is the foundation of IoT. Automatic identification technologyprovides the basic means of the connection between objects and network. With the expansionand development of IoT, automatic identification technology is widely used in IoT, as one ofthe core technologies.Because of the adaptability of the network demand, RFID (Radio FrequencyIdentification, RFID) becomes the most important technology in numerous automaticidentification technologies. The biggest advantage of RFID technology is its ability of multitarget recognition, which fully meets the features of IoT like "mass storage" and "full sense".However, the multi-objective anti-collision recognition problem is the key technical problemof the RFID system applications, which is very difficult and complex and needs to beconstantly researched and explored to get more efficient and safer anti-collision technologythat can meet lower energy consumption requirement, so as to further strengthen the technicaladvantages of RFID.With the research of anti-collision recognition of multi RFID tag on theory andtechnology for many years, it has formed many algorithms. The algorithms can be dividedinto several categories. The current researches on RFID anti-collision algorithms mainly focuson time division multiplexing method. Two main parts of the method is probabilistic ALOHAalgorithms and deterministic binary tree search algorithms. In both of them, probabilisticALOHA algorithms are more suitable for the IoT applications which have large number oftags.Based on above, this thesis presents the present common types of ALOHA algorithmsand binary tree search algorithms, analyses and compares the algorithm work flow, impactand efficiency. Through the analysis and study of several improving methods for RFID taganti-collision algorithms, for applications having large number of tags, the thesis proposes adynamic frame slotted ALOHA improved anti-collision algorithm named TG-DFSA(Tag-Grouping Dynamic Framed Slotted ALOHA anti-collision algorithm).Anti-collision algorithm based TG-DFSA algorithm in dynamic frame slotted ALOHAon general improvement, will read and write processing classified stage requires the identification tag region. Tag identification process is divided into two stages: the first stage isbased on a response to each time slots in a frame contains the number of tags, theclassification of randomly selected time slot reader all tags within the region, and to identifyindividual only slot label; the second stage is the specific time slot conflict of the recognitionprocessing. The algorithm can quickly and accurately adjust the frame length, accelerated skipidle slots and processing only time slot, and the additional frame uniform effective treatmentof colliding tags.The validity of the algorithm through simulation experiment, and compared theperformance of TG-DFSA algorithm and general frame slotted ALOHA algorithm anddynamic frame slotted ALOHA algorithm. The experimental results demonstrate that theTG-DFSA algorithm can effectively improve the throughput and channel utilization, reducethe time and energy overhead of the recognition process; in tag intensive, are a number ofTG-DFSA algorithm compared with other algorithms, the superiority is more prominent, hasgood applicability for things networking perception layer of the RFID system, provides aefficient, rational multiple tag identification solutions for the study of RFID technology torealize the overall perception and recognition in IoT.