| Nowadays,a new round of industrial revolution takes place in the world at the same time.A series of national strategies,such as “Industries 4.0” in Germany,“Society 5.0” in the Japan,“Industrial Internet” in the USA,“Made in China 2025”,are proposed and implemented.The development direction of the manufacturing industry in the next generation are discussed,closely around the Internet of Things technology,discussing how to connect the devices,the factories and the infrastructure through the Internet.The Internet of Things is known as the third wave of information industry after computer and Internet.According to the agreement,any item is connected to the Internet for communication and information exchange,and functions such as intelligent identification,location,monitoring and management are realized.The information and communication technologies represented by the Internet of Things are rapidly developing at an unprecedented speed and have greatly influenced on the world economic situation.Radio frequency identification is one of the key technologies of the perception layer in Internet of Things.The RFID tags embedded in the object(or attached to the object)transmit their information to the reader.Then RFID technology achieves information interaction and sharing through the open Internet.It realizes the connection between physical objects and the Internet and completes automatic identification,tracking,monitoring,and "transparency" management for any physical objects attached with RFID tags.At present,RFID technology has been widely applied in such fields as logistics,anti-counterfeiting,retail,supply chain management and has made great contributions to the development of the Internet of Things.In an RFID system,the reader can identify RFID tags through a shared communication channel,while the tags cannot exchange information between each other,and cannot perceive the existence of other tags.Collisions occur when multiple tags respond to the reader's query at the same time,resulting in loss of information,increasing identification delay,and reducing RFID system efficiency and reliability.How to effectively identify tags to avoid collision is the focus and realization foundation for the research of RFID system.In order to solve the problem of RFID tag identification for Internet of Things,this paper focuses on improving the system identification efficiency,and makes the following contributions around multi-tag identification in general application environment and specific application environment:(1)According to the global uniqueness of tag ID,binary coding feature and XOR operation,a bit arbitration strategy is designed.In this strategy,the grouping rule based on characteristic value and the rule of collision bits are defined,using three-bit arbitration to replace one-bit of the traditional query tree-based protocol.The bit arbitration strategy divides the eight data combinations evenly into two groups with different characteristics.Aiming at the static RFID system application environment,a bit arbitration query tree protocol based on grouping mechanism is proposed according to the bit arbitration strategy,to solve the bottleneck problem of low efficiency of classical tree-based protocols.If the responded tags belong to two different groups,the reader queries each group of tags at different time,and uses the rule of collision bits to determine the transmission data combinations,which reduces the number of idle slots generated during the tag recognition process.The theoretical analysis and experimental simulation of the proposed protocol are carried out in four aspects: the total number of inqueries,the identification efficiency,the communication complexity and the average transmission delay.The results show that the performance of proposed protocol is better than that of contrast protocols,and the system identification efficiency is increased to 0.54.(2)In order to reduce collision time slots generated by the bit arbitration query tree protocol based on grouping mechanism,a bit arbitration tree anti-collision protocol is proposed.According to the number of collision bits of the transmission data,the cases of the tag collision are analyzed,and then the transmission data combinations are reasonably inferred.If there is one collision bit,the reader can infer two kinds of transmission data directly through one query;if there are two collision bits,two or four kinds of transmission data are inferred directly,which reduces the number of times to search tags.The experimental results show that the proposed protocol reduces the total number of inquiries required by the reader to identify tags,degrades the system energy consumption and improves the identification efficiency.In terms of total number of inqueries and identification efficiency,the theoretical analysis results are basically consistent with experimental simulations.The performance of the proposed protocol is stable,and the identification efficiency is not affected by the tag ID length,but only related to the number of tags.(3)To solve the problem of identification for new-coming tags entering the RFID system,a novel unknown tag identification protocol is designed,which saves the time spent by traditional tag identification protocols for recollecting known tags.The protocol is mainly divided into two phases: the preprocessing phase and the identification phase.In the preprocessing phase,the constructed indication vector is used to represent the expected collision slot and the expected non-collision slot.The tags matching the expected non-collision slots are active,wherein the unknown tags selecting the expected idle slots are identified or labeled,and the known tags selecting the expected readable slots are inactive.In the identification phase,the classical Enhanced Dynamic Framed Slotted Aloha protocol is used to identify the labelled unknown tag.To reduce the time to identify unknown tags,theoretical analysis sets the optimal frame length and the minimum number of polling times.The proposed protocol is simulated from three aspects of the frame length,the total execution time and the identification accuracy.The experimental results show that the proposed protocol is significantly superior to the contrast protocols in terms of time efficiency.When the identification accuracy is 99%,the number of known tags and the number of unknown tags are 10000 respectively,the proposed protocol takes an average of 6.5ms to identify an unknown tag.(4)Aiming at a large-scale RFID system application environment with low density of unknown tags,a low-density unknown tag identification protocol is proposed to identify all the unknown tags in the system effectively.The proposed protocol constructs two indicator vectors to accurately represent that the expected state of the time slot is idle,readable or collision,and effectively distinguishes the known tags and unknown tags by the change of slot state.By inactivating known tags and labeling unknown tags in the labeling phase,the probability of tag collision is decreased,and the interference to unknown tag identification is reduced.Theoretical analysis considers three system parameters: the optimal frame length,the minimum number of polling times,and the best execution time.When the system has 10000 known tags and 100 unknown tags,the execution time of the low-density unknown tag identification protocol is increased by 40.7% over the novel unknown tag identification protocol.The simulation results show that the proposed protocol can quickly identify unknown tags in the sparse unknown tag environment of a large-scale RFID system.In this paper,four kinds of tag identification protocols for the Internet of Things are proposed,which improve the system efficiency,realize fast reading of tag information,and provide meaningful research results and valuable reference for the subsequent research and application of tag identification protocols. |
PDF Full Text Request