Research On Anti-Collision Algorithm Based On ALOHA For RFID System

As one of automatic identification technologies, RFID (Radio Frequency Identification) communicates bi-directionally and wirelessly by means of radio signals in the space. RFID was used for military affairs originally in World Warâ…¡and has run to a wide range of applications in industry and daily life. The core technology of the Internet of Things is RFID which collects terminal data at the bottom in real time, and the internet achieves automatic identification and information sharing.Multi-target recognition is the main advantage of RFID, however causes data collision between the feedback signals in the uplink from these tags at the same time. The collision threatens data integrity of RFID system. Reliable anti-collision mechanism is always needed to effectively reduce the damage of data integrity caused by collision. Taking system cost and technical level into account, the most popular method to solve collision is TDMA (Time Division Multiple Access), which consists of the deterministic algorithms based on binary search and the probabilistic algorithms based on ALOHA. ALOHA algorithm is irregular and unrestrained from the tag ID length, so that it is the main research contents in this paper.A random framed-slotted ALOHA based Q-algorithm is proposed to mitigate the tag collision problem in EPC C1G2 protocol. When tags receive a command with parameter Q from an interrogator, each tag randomly generates an integer between 0 and 2Q-1 to load the slot counter. Tags with 0 in their slot counters could respond, while others decrease their slot counters by one or load a new Q value to their slot counters according to the specific command. FSA-DS algorithm (Framed-Slotted ALOHA Algorithm with D-value Threshold Judgment and Slot-Predicting Binary Selection) was proposed in this paper against inflexibility of adjustment for Q and other defects in original protocol during interaction. Innovations are as follows:1. Adjustment for frame sizeFrame size should be as close with the number of tags to be read as possible in order to maintain a high efficiency system. FSA-DS algorithm provided the value of accumulation factor C of EPC protocol so that the decision threshold for adjusting Q, namely Th(Q) could been derived. Then the proposed algorithm modified the prior frame size by comparison with the preset decision threshold Th(Q) according to DTJ (D-value Threshold Judgment) scheme.2. Hash in each frameThe system throughput would be improved obviously in theory if tags decentralize in the frame and choose different time slots automatically. The SPBS (Slot-Predicting Binary Selection) scheme was used in each frame to accelerate idle slots and solved collisions instantly with binary selection, so that to reduce the invalid slots essentially.3. The condition to end the identificationThere is none of any provision in this respect in EPC C1G2 protocol. For this reason, the end condition was led into FSA-DS according to the value of Th(Q) in DTJ. Q is initialized to 4 in the original protocol. On the basis of DTJ, the disparity between the number of idle slots Nidle and collision slots Ncoll must reach Th(1) (that is 2) which is the decision threshold of Q for transformation from 1 to 0. When Q equals to 1, if and only if all time slots are idle or it can be said none of tags response, the relationship between Nidle and Ncoll could meet the above situation. Since Q=1 could be the conditions to end the identification so that the interrogator would be sensitive to the end, then the enforced termination of following orders would avoid waste of resources.To verify the validity of the innovative scheme, four groups of simulation on MATLAB were conducted including adjustment method for frame size, hash method in each frame, transmission overhead, identification delay and throughput. Simulation result shows that the proposed scheme can increase the system throughput and consumes less power and fewer time slots than previous work, and provides a solution for data collection of Internet of Things which takes RFID tags as perceptive terminal nodes.