RFID target tag category information collection method based on double time slots

Abstract

The invention discloses an RFID target tag category information collection method based on double time slots, under unknown tag interference, target tags are classified into M possible categories according to environment state information collected by the target tags, in order to classify local tags (located in a coverage area of a reader-writer) in the target tags, a system needs to execute multiple rounds, and the classification efficiency is high. Each round comprises two stages of non-target label filtering and local target label category information collection; in a non-target label filtering stage, two test frames are introduced to allocate two time slots for a known label, on one hand, non-target labels can be iteratively eliminated from the time slots through the two test frames; on the other hand, unknown labels are identified through the 0 component in the double indication vectors; in a target label category information collection stage, category information of a target label is collected through two information frames; according to the method, the identification probability of the unknown tag is improved, and the collection efficiency of the category information of the local target tag is also improved.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) technology, specifically to a method for collecting RFID target tag category information based on dual time slots. Background Technology

[0002] Radio Frequency Identification (RFID) uses RFID tags embedded in the surface of objects to transmit identification information to a reader, enabling the identification, tracking, monitoring, and "transparent" management of the physical objects corresponding to the RFID tags. RFID has been widely applied in logistics, anti-counterfeiting, and retail warehouse management. RFID tags integrate miniature sensors to monitor physical parameters such as ambient temperature and humidity. Many goods in cold chain logistics, pharmaceutical transportation, and fresh food storage are highly sensitive to temperature changes. For example, a cold chain warehouse stores three types of goods: frozen food (-18℃), refrigerated medicines (2-8℃), and room temperature food (20-25℃). From the perspective of monitoring medicines (vaccines, insulin), the system only needs to determine in real time whether the temperature of refrigerated medicines is within a safe range; at the same time, it needs to exclude interference from frozen food, room temperature food, and unknown tags not registered in the system.

[0003] To address the above requirements, a protocol is needed that can eliminate interference from unknown tags and collect category information from a subset of targets of interest. Polling is the simplest identification method in RFID systems, where the reader actively queries tag IDs one by one, allowing only one tag to respond at a time to avoid conflicts; however, this method has a long identification time when there are many target tags. Previous techniques for collecting information from target tags include the TIC protocol (X. Liu, J. Yin, S. Zhang, B. Xiao and B. Ou, "Time-Efficient Target Tags Information Collection in Large-Scale RFID Systems," in IEEE Transactions on Mobile Computing, vol. 20, no. 9, pp. 2891-2905, 1 Sept. 2021). The advantage of this protocol is that it collects information from a subset of targets of interest, i.e., target tags, but it does not consider interference from unknown tags, nor does it optimize the efficiency of category information collection. To address interference from unknown tags, the RQ protocol (X. Liu et al., "Efficient Range Queries for Large-Scale Sensor-Augmented RFID Systems," in IEEE / ACM Transactions on Networking, vol. 27, no. 5, pp. 1873-1886, Oct. 2019) is a technique for collecting category information. Its advantage is that it collects tag category information considering interference from unknown tags; however, it targets all tags within the reader's range, rather than a subset of the target tags of interest. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to provide a method for collecting category information of RFID target tags based on dual time slots, which can efficiently collect category information of target tags in the presence of unknown tag interference.

[0005] Technical Solution: The present invention provides a method for collecting RFID target tag category information based on dual time slots, comprising the following steps: specifying a metric ε for the classification accuracy of local target tags and a metric δ for the misclassification rate of non-local target tags (the ratio of non-local target tags being incorrectly identified as local target tags). Selecting a round number R to execute the RFID target tag category information collection method based on dual time slots, each round consisting of the following two stages:

[0006] (1) Non-target tag filtering stage: Calculate two time slot indices for each known tag using a hash function; map the known tags to the corresponding time slots of the two test frames; construct two indicator vectors to represent the time slot status: if the time slot is an expected empty time slot or is only assigned to non-target known tags, the corresponding component of the indicator vector is recorded as 0; if the time slot is assigned to at least one target tag, the corresponding component of the indicator vector is recorded as 1; the reader broadcasts the indicator vector, and eliminates unknown tags and non-target tags through the components that are 0 in the indicator vector;

[0007] (2) Local target tag category information collection stage: The active local tag calculates and checks the corresponding components in the two indicator vectors based on the two time slot indices; if the components in the two indicator vectors are both 1, the active local tag sends its category information in the two information frames in sequence; the time slot status vector in each information frame is the result of bitwise OR operation on the category information of the tag selected in the time slot; the reader then performs a bitwise AND operation on the candidate category vector of each target tag and the time slot status vector in the corresponding information frame, and updates the candidate category vector of the target tag; if the candidate category vector becomes a zero vector, it is identified as a non-local tag and eliminated; after the predetermined rounds are completed, if the candidate category vector is a standard unit vector, it means that the category information of the target tag has been successfully acquired; the remaining target tags whose candidate category vectors are not standard unit vectors or zero vectors are collected by polling.

[0008] Furthermore, step (1) includes the following steps:

[0009] (1.1) Initialization: Initialize all category sets to an empty set; Let N0 be initialized as an empty set, used to record identified non-local target tags; for target tag id, the RFID reader initializes its candidate class vector w. id A vector of all 1s with θ bits;

[0010] (1.2) Enter the non-target tag filtering stage, and map each known tag ID to the test frame V using a hash function. i k Selecting time slots in The random number f selected for the RFID reader i Frame length;

[0011] (1.3) The RFID reader is based on V i k Selecting time slots Construct indicator vector

[0012] Furthermore, step (1.3) includes the following steps:

[0013] Let the test frame be V i 1 For test frame V i 1 Each selected time slot V i 1 [j], 0 ≤ j ≤ f i -1, RFID reader executes:

[0014] (1.3.1) If V i 1 [j] represents the expected spatial time slot, then the indicator vector... The corresponding component is recorded as 0.

[0015] (1.3.2) If V i 1 [j] Only known labels that are not the target are assigned to the indicator vector. If the corresponding component is recorded as 0, then At the same time, the RFID reader records and eliminates non-target tags from the known tag set;

[0016] (1.3.3) If V i 1 If at least one target label is contained in [j], then the indicator vector IV will be... i 1 The corresponding component is denoted as 1.

[0017] (1.3.4) Test Frame V i 1 The identified known labels of non-target objects were used to further update the indicator vector IV. i 2 The RFID reader / writer tests the V frame. i 2 Perform the same steps (1.3.1-1.3.3); in test frame V i 2 Non-target known tags that are recorded and removed will also be iteratively used by the RFID reader to refresh the tags assigned to test frame V. i 1 Known labels in the time slot.

[0018] Furthermore, step (2) includes the following steps:

[0019] (2.1) Entering the local target tag category information collection stage, the RFID reader broadcasts... Indicator Vector and frame length f i To initiate the category information collection phase of the i-th round;

[0020] (2.2) Active local tags calculate the slot index and check according to step (1.2). and like The corresponding local tag participates in the subsequent execution of the protocol and sends its own category information; if The corresponding non-target known labels or unknown labels will not participate in the subsequent category information collection;

[0021] (2.3) If the following conditions are met Then the label ID will be placed in the indicator vector. middle The number of previous '1's is denoted as The tag ID is placed in the indicator vector middle The number of '1's previously is denoted as Tag ID in information frame F i 1 of Time slots and information frames F i 2 of In the time slot, the class information (ran) is transmitted using on-off keying modulation. id ; among which, ran id =(ran id [0],…,ran id [θ-1]), is a standard unit vector e i If the information perceived by the label id is id.val∈S i Then ran id [i] = 1; the remaining components are ran id [j](0≤j≤θ-1,j≠i) are all 0;

[0022] (2.4) Using LIV i 1 Information frame F i 1 The number of time slots in the RFID reader / writer for the information frame F i 1 Each information slot F in i 1 [k], Perform the following processing: If information time slot F i 1 [k] is an empty time slot, indicating that the label selected for this time slot is a non-local label; if the information time slot F i 1 [k] represents the busy time slot, used as... Indicates in information time slot F i 1If the local tag responds in [k], the RFID reader will retrieve the information from time slot F. i 1 [k] Extracted θ-bit slot state vector info k , means as follows:

[0023]

[0024] Where ∨ represents all Corresponding category information ran id Perform a bitwise OR operation, and the result is the slot state vector of the corresponding slot in the information frame;

[0025] (2.5) For each selected information slot F i 1 The target label id of [k] is used by the reader to update its candidate class vector:

[0026] w id ←w id ∧info k

[0027] Where ∧ represents the logical AND operation; that is, the reader performs a bitwise AND operation between the current candidate category vector of each target tag and the slot state vector of the corresponding information frame, and updates the candidate category vector of the target tag.

[0028] (2.6) The RFID reader / writer processes the information frame F i 2 Each information slot F in i 2 [k], Perform the calculations in steps (2.4) and (2.5);

[0029] (2.7) After the target tag category information collection protocol runs for R rounds, the RFID reader collects information on each target tag ID that is not identified as a non-local tag.

[0030] Furthermore, in step (2.5), if the candidate class vector w id Update to all-zero vector: Determine if the target label id is a non-local label; the candidate category vector of the non-local label is eventually updated to a zero vector during the iteration of the reader / writer.

[0031] Furthermore, in step (2.7), if the candidate class vector w id It is a unit vector e j ,0≤j≤θ-1: Classify the target label id into S j Otherwise, use a polling method to collect category information for these target labels.

[0032] Furthermore, if some non-local target label IDs have candidate category vectors w id If w accidentally becomes a unit vector in a certain round due to interference from other labels, then the logical AND operation in step (2.5) is used to update w. id w id It is subsequently updated to an all-zero vector.

[0033] Furthermore, let n be the number of known labels remaining in the i-th round of the non-target label filtering stage. i The number of remaining target labels is k. i Test Frame V i 1 The probability that one of the selected time slots is the expected empty time slot is:

[0034]

[0035] Test Frame V i 1 The probability that a selected time slot is allocated only to a non-target known label is:

[0036]

[0037] Analysis based on steps (1.3.1) and (1.3.2) The probability of occurrence is Similar analysis The probability of an unknown label being identified in the i-th round of the non-target label filtering stage in step (2.2) is:

[0038]

[0039] An electronic device according to the present invention includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the program to implement the steps of any of the methods described herein.

[0040] The present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the methods described herein.

[0041] Beneficial Effects: Compared with existing technologies, this invention has the following significant advantages: In the non-target label filtering stage, two test frames are introduced to allocate two time slots for known labels. On the one hand, non-target labels can be iteratively eliminated from the time slots through the two test frames; on the other hand, unknown labels are identified by using the 0 component in the dual indicator vector, thereby improving the identification probability of unknown labels. In the target label category information collection stage, the category information of target labels is collected through two information frames, thereby improving the efficiency of category information collection. Attached Figure Description

[0042] Figure 1 This is a label classification diagram of the present invention;

[0043] Figure 2 This is the construction indicator vector of the present invention;

[0044] Figure 3 This is the unknown tag response diagram of the present invention;

[0045] Figure 4 The information frame F of this invention i 1 ;

[0046] Figure 5 The information frame F of this invention i 2 ;

[0047] Figure 6 The graph showing the relationship between the local target tag collection time (seconds) and the known number of tags in this invention.

[0048] Figure 7 The relationship between the local target tag collection time (seconds) and the number of target tags (units) of this invention. Detailed Implementation

[0049] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0050] like Figure 1 As shown, this embodiment of the invention provides a method for collecting RFID target tag category information based on dual time slots, including the following steps:

[0051] In warehouse management, tags are attached to goods, and sensors collect ambient temperature data in real time for encoding. Goods stored at -5℃ to 5℃ are classified as R0, goods stored at 6℃ to 15℃ as R1, and goods stored at 16℃ to 25℃ as R2. That is, the tags are divided into three categories based on the environmental information collected. In practical applications, sometimes it is only necessary to know which category the goods belong to, which requires classifying and identifying the status information of the target tags.

[0052] Figure 1 In this paper, w1 and w2 are set to class R0 with a status identifier of 100; w3 is set to class R1 with a status identifier of 001; w4 is set to class R2 with a status identifier of 001; there are also unknown tags u1, u2 and u3 within the reader's recognition range. Let the status identifier corresponding to u1 be 100, the status identifier corresponding to u2 be 010, and the status identifier corresponding to u3 be 001.

[0053] The RFID reader initializes the three category sets S0, S1, and S2 as empty sets to record subsequently identified local target tags (e.g., Figure 1 In the set w1, w2, w3, w4, ..., N0 is also initialized to an empty set to record the identified non-local target labels (e.g., ...). Figure 1 w5, w6); for the target tag id ( Figure 1 In the array w1, w2, w3, w4, w5, w6, the reader initializes its class vector w. id It is 111.

[0054] The reader calculates a hash function for the known tags (t1, t2, t3, w1...w6) within the system and maps them to the test frame V respectively. i 1 and V i 2 The corresponding time slot is selected.

[0055] exist Figure 2 In the construction of the indicator vector:

[0056] In (1.1), time slots 5, 7, and 8 are all expected space time slots, so the indicator vector...

[0057] (1.2) Due to test frame V i 1 Time slots 3 and 6 are assigned only to non-target labels t2 and t3, respectively, so the indicator vector... The RFID reader eliminates non-target known tags t2 and t3 from the known tag set;

[0058] (1.3) Due to test frame V i 1 At least one target label is contained in time slots 0, 1, 2, and 4, so the indicator vector...

[0059] (1.4) Test Frame V i 1 The recorded non-target known labels are used to further update the indicator vector IV. i 2 The RFID reader / writer tests the V frame. i 2 Similarly, steps (1.1) through (1.3) are performed. By allocating time slots for known labels using two test frames, an indicator vector is constructed, increasing the probability of filtering non-target labels and unknown labels. Non-target labels filtered in the first test frame can be eliminated from the second test frame, and further filtering of non-target labels can be performed on the second test frame. The non-target labels filtered in the second test frame can then be used to update the known labels allocated to the time slots of the first test frame. Figure 3 Middle test frame V i 1 The non-target known labels t2 and t3 identified in the test frame are first used to update the assignment to the test frame V. i 2 Known labels in time slots, then for V i 2 Filter out non-target tags; V i 2 The non-target known label t1 identified in the test frame can be used to update the assignment to the test frame V. i 1 Known tags in the time slots; therefore, during the non-target tag filtering phase, non-target known tags t1, t2, and t3 are recorded by the RFID reader and eliminated from the time slots of the two test frames:

[0060] (2.1) RFID reader broadcast Indicator Vector and frame length f i This initiates the category information collection phase of the i-th round.

[0061] (2.2) For active local tags t1, t2, t3, w1, w2, w3, w4, u1, u2, u3, calculate the slot index and check according to step 1.2). and like Figure 2 The tags w1, w2, w3, and w4 satisfy... That is, if the corresponding elements in both indicator vectors are both 1, the system can continue to participate in the subsequent execution of the protocol and send its own category information; for example... Figure 2 Labels t1, t2, and t3 do not satisfy the condition. Non-target labels t1, t2, and t3 will not participate in subsequent category information collection; among them, t1's... t2 t3 like Figure 3 Unknown tags u1 and u2 It can continue to participate in the subsequent execution of the protocol and send its own category information; while the unknown tag u3 Therefore, they will not participate in the subsequent stages of implementation.

[0062] (2.3) Constructing an information frame: For satisfying The labels w1, w2, w3, w4, u1, u2 are used to enter the target label category information collection stage. Taking label w1 as an example, w1 is calculated in the indicator vector. middle The number of '1's before was exist middle The number of '1's before was Tag w1 will be in information frame F i 1 F i 1 [0]Time slot and information frame F i 2 F i 2 [2] Send their category information ran in the time slot respectively. id1 =100.

[0063] (2.4) Update the time slot state vector: Figure 4 and Figure 5 Information slot F in the information frame corresponding to tag w1 i 1 [0] and F i 2 [2] These are all busy time slots, using Indicates in information time slot F i 1 The set of local tags in [0] that responds, using Indicates in information time slot F i 2 [2] The set of local tags in the response: Subsequently, the time slot status information (info) of the two information time slots was analyzed. j The bitwise OR operation is performed based on the tag's category information. Therefore, information frame F i 1 F in i 1 The time slot state vector of [0] is represented as:

[0064]

[0065] Information frame F i 2 China F i 2 The time slot state vector of [2] is represented as:

[0066]

[0067] Figure 4 and Figure 5 The middle belongs to information frame F i 1 [1] The local label in the time slot is In information frame F i 2 [1] Assigned to F i 1 [1] and F i 2[1] Non-local target label w5 in the time slot does not send its category information. Therefore, information frame F i 1 F in i 1 The time slot state vector of [1] is expressed as:

[0068]

[0069] Information frame F i 2 F in i 2 The time slot state vector of [1] is represented as:

[0070]

[0071] Figure 4 and Figure 5 The middle belongs to information frame F i 1 [2] The local label of the time slot is In information frame F i 2 [5] The local label for the time slot is Assigned to F i 1 [2] and F i 2 [5] Non-local target tags w6 in time slots do not send their category information. Therefore, information frame F i 1 F in i 1 The time slot state vector of [2] is represented as:

[0072]

[0073] Information frame F i 2 F in i 2 [5] The time slot state vector is represented as:

[0074]

[0075] (2.5) Update the candidate class vector:

[0076] Perform a bitwise AND operation between the current candidate category vector of each target tag and the slot state vector of the corresponding information frame, and update the candidate category vector of the target tag.

[0077] Through F i 1 [0] Update the candidate class vector of w1, w id1 ←w id1∧info0=111∧110=110, further through F i 2 [2] Update the candidate class vector of w1, w id1 ←w id1 ∧info2=110∧100=100.

[0078] Through F i 1 [1] Update the candidate class vector of w5: w id5 ←w id5 ∧info1=111∧110=110, further through F i 2 [1] Update the candidate class vector of w5, w id5 ←w id5 ∧info1=110∧000=000. The w stored in the reader / writer. id5 If the vector converges to all zeros, w5 can be determined to be a non-local target label.

[0079] Through F i 1 [2] Update the candidate class vector of w6: w id6 ←w id6 ∧info2=111∧100=100, further through F i 2 [5] Update the candidate class vector of w6, w id6 ←w id6 ∧info5=100∧001=000. The w stored in the reader / writer. id6 If the vector converges to all zeros, w6 can be determined to be a non-local target label.

[0080] To verify the performance advantages of the proposed BRQT scheme in local target label recognition, this invention conducts comparative experiments and analyses with the existing TIC, the new RQT protocol with the added target label collection function, and the Polling protocol.

[0081] BRQT with different numbers of known labels

[0082] The experimental variable was the number of known tags, which increased from 5000 to 10000 in increments of 1000. The number of locally known tags was half the total number of known tags. While maintaining consistent experimental parameters such as the number of target tags (3000), unknown tags (1000), and locally known target tags (1500), the experiment divided the tags into 8 categories, and the proportion of target tags among known tags decreased from 60% to 30%.

[0083] Table 1. Local target label recognition accuracy under different numbers of known labels

[0084]

[0085] As shown in Table 1, BRQT can consistently achieve a local target label recognition accuracy of 0.99 under the condition of a known increase in the number of labels.

[0086] Table 2. Misclassification rate (%) of non-local target labels under different numbers of known labels.

[0087]

[0088] The misclassification rate refers to the rate at which non-local target labels are identified as local target labels. The misclassification rates of non-local target labels listed in Table 2 show that BRQT performs excellently in resisting interference from non-target labels, with misclassification rates consistently below 1%, while TIC (which does not consider interference from unknown labels) and RQT (which does not consider interference from non-target labels) have misclassification rates as high as 4.93–9.07% and 17%–19%, respectively.

[0089] Table 3. Time cost (seconds) for identifying local target labels under different numbers of known labels.

[0090]

[0091] As shown in Table 3, when the number of known tags exceeds 6000, the BRQT recognition scheme demonstrates a more significant improvement in time performance compared to TIC and RQT. With 10000 known tags, BRQT takes 10.125 seconds, which is approximately 33% and 43% faster than TIC (15.377 seconds) and RQT (17.980 seconds), respectively.

[0092] Depend on Figure 6 It can be seen that the polling protocol maintains a constant time due to the unchanged number of target tags. The BRQT scheme shows a gradual increase in information collection time across different numbers of known tags; as the number of known tags increases from 5000 to 10000, the collection time only increases from 9.658 seconds to 10.125 seconds. When the number of known tags reaches 6000 or more, the BRQT scheme has a certain advantage over the other three protocols.

[0093] Combining the above three tables with Figure 6 It can be seen that BRQT has certain advantages in terms of local target label recognition accuracy, misclassification rate and recognition time in large-scale and complex scenarios.

[0094] BRQT with varying numbers of target tags: The experimental variable was the number of target tags, increasing from 1000 to 6000 in increments of 1000. Local target tags comprised half of the total target tags. While maintaining consistent experimental parameters such as the number of known tags (10000), unknown tags (1000), and locally known tags (5000), the tags were divided into 8 categories, and the proportion of target tags among known tags increased from 10% to 60%.

[0095] Table 4. Local target label recognition accuracy under different numbers of target labels

[0096]

[0097] As shown in Table 4, the BRQT scheme maintains a very high local target label recognition accuracy under various target label quantity conditions, which can meet the requirement that the local target label recognition accuracy is greater than 0.99.

[0098] Table 5. Misclassification rate (%) of non-local target labels under different numbers of target labels.

[0099]

[0100] As shown in Table 5, the misclassification rate of non-target labels in the BRQT scheme is consistently kept below 1%, with a minimum of 0.5% and a maximum of only 0.87%, which is significantly lower than that of TIC (3.00%–4.88%) and RQT (15.40%–18.70%). This indicates that BRQT can eliminate interference from non-local target labels.

[0101] Table 6. Time cost (seconds) for identifying local target labels under different numbers of target labels.

[0102]

[0103] As shown in Table 6, BRQT performs well in terms of recognition time, especially when the number of target labels is 1000-5000, it is better than TIC and RQT. It is only slightly better than TIC and RQT when the number of target labels is 6000. This is due to the scheduling cost that BRQT pays to ensure that the local target label recognition accuracy is greater than 0.99 and the non-target label misclassification rate is less than 1%.

[0104] Depend on Figure 7It is evident that BRQT consistently outperforms the Polling protocol. When the number of target labels is between 1000 and 4000, BRQT shows a certain advantage over TIC and RQT. When the number of target labels exceeds 5000, BRQT maintains a local target label recognition accuracy greater than 0.99 and a non-local target label misclassification rate less than 1%, although its time performance slightly decreases.

[0105] In summary, BRQT can meet the given recognition accuracy and misclassification rate requirements during the collection of classification information for target tags, while also exhibiting good time characteristics. Compared to existing solutions such as TIC, RQT, and Polling, BRQT's overall performance makes it suitable for IoT application environments with a large number of known tags, a moderate proportion of target tags among the known tags, and the presence of unknown tags, demonstrating significant practical and promotional value.

Claims

1. A method for collecting RFID target tag category information based on dual time slots, characterized in that, Includes the following steps: In the presence of unknown labels causing interference, perform the following operations: (1) Non-target tag filtering stage: Calculate two slot indices for each known tag using a hash function; map the known tags to the corresponding slots of the two test frames; Construct two indicator vectors to represent the slot status: if the slot is a expected empty slot or is only assigned to a known tag other than the target, the corresponding component of the indicator vector is recorded as 0; if the slot is assigned to at least one target tag, the corresponding component of the indicator vector is recorded as 1. The reader broadcasts an indication vector, eliminating unknown and non-target tags by using the components in the indication vector that are 0. (2) Local target label category information collection stage: Active local labels check the corresponding components in the two indicator vectors based on the calculated two time slot indices; If both components in the two indicator vectors are 1, the active local tag sends its category information sequentially in the two information frames; the state vector of each time slot in the information frame is the result of a bitwise OR operation on the category information of the tag selected for that time slot; the reader then performs a bitwise AND operation on the candidate category vector (initialized as an all-one vector) of each target tag with the time slot state vector in the corresponding information frame, and updates the candidate category vector of the target tag; when the candidate category vector of a tag becomes a zero vector, it is identified as a non-local tag and eliminated; after the predetermined number of rounds is completed, if the candidate category vector is a standard unit vector, it means that the category information of the target tag has been successfully acquired; The remaining target labels for candidate category vectors that are not standard unit vectors or zero vectors are collected using a polling method.

2. The method for collecting RFID target tag category information based on dual time slots according to claim 1, characterized in that, Step (1) includes the following steps: (1.1) Initialization: Initialize all category sets to an empty set; Let N0 be initialized as an empty set to record identified non-local target tags; for each target tag id, the RFID reader initializes its candidate class vector w. id A vector of all 1s with θ positions; (1.2) Enter the non-target tag filtering stage, and map each known tag ID to the test frame V using a hash function. i k Selecting time slots in The random number f selected for the RFID reader i Frame length; (1.3) The RFID reader is based on V i k Selecting time slots Construct indicator vector 3. The method for collecting RFID target tag category information based on dual time slots according to claim 2, characterized in that, Step (1.3) includes the following steps: Let the test frame be V i 1 For test frame V i 1 Each selected time slot V i 1 [j], 0 ≤ j ≤ f i -1, RFID reader executes: (1.3.1) If V i 1 [j] represents the expected spatial time slot, then the indicator vector... The corresponding component is recorded as 0. (1.3.2) If V i 1 [j] Only known labels that are not the target are assigned to the indicator vector. If the corresponding component is recorded as 0, then At the same time, the RFID reader records and eliminates non-target tags from the known tag set; (1.3.3) If If the vector contains at least one target label, then the indicator vector will be... The corresponding component is denoted as 1. (1.3.4) Test Frame V i 1 The identified known labels of non-target objects were used to further update the indicator vector. RFID reader for test frame V i 2 Perform the same steps (1.3.1-1.3.3); in test frame V i 2 Non-target known tags that are recorded and removed will also be iteratively used by the RFID reader to refresh the tags assigned to test frame V. i 1 Known labels in the time slot.

4. The method for collecting RFID target tag category information based on dual time slots according to claim 1, characterized in that, Step (2) includes the following steps: (2.1) Entering the local target tag category information collection stage, the RFID reader broadcasts... Indicator Vector IV i 2 and frame length f i To initiate the category information collection phase of the i-th round; (2.2) Active local tags calculate the slot index and check according to step (1.2). and like The corresponding local tag participates in the subsequent execution of the protocol and sends its own category information; if The corresponding non-target known labels or unknown labels will not participate in the subsequent category information collection; (2.3) If the following conditions are met Then the label ID will be placed in the indicator vector middle The number of previous '1's is denoted as The tag ID is placed in the indicator vector middle The number of '1's previously is denoted as Tag ID in information frame F i 1 of Time slots and information frames F i 2 of In the time slot, the class information (ran) is transmitted using on-off keying modulation. id ; among which, ran id =(ran id [0],…,ran id [θ-1]), is a standard unit vector e i If the information perceived by the label id is id.val∈S i Then ran id [i] = 1; the remaining components are ran id [j](0≤j≤θ-1,j≠i) are all 0; (2.4) Using Information frame F i 1 The number of time slots in the RFID reader / writer for the information frame F i 1 Each information slot in Perform the following processing: If information time slot F i 1 [k] is an empty time slot, indicating that the label selected for this time slot is a non-local label; if the information time slot F i 1 [k] represents the busy time slot, used as... Indicates in information time slot F i 1 If the local tag responds in [k], the RFID reader will retrieve the information from time slot F. i 1 [k] Extracted θ-bit slot state vector info k , means as follows: Where ∨ represents all Corresponding category information ran id Perform a bitwise OR operation, and the result is the slot state vector of the corresponding slot in the information frame; (2.5) For each selected information slot F i 1 The target label id of [k] is used by the reader to update its candidate class vector: In id ←in id ∧info k Where ∧ represents the logical AND operation; that is, the reader performs a bitwise AND operation between the current candidate category vector of each target tag and the slot state vector of the corresponding information frame, and updates the candidate category vector of the target tag. (2.6) The RFID reader / writer processes the information frame F i 2 Each information slot in Perform the calculations in steps (2.4) and (2.5); (2.7) After the target label category information collection protocol runs for R rounds, if the candidate category vector of the target label is a standard unit vector, it means that the category information of the target label has been successfully obtained; for the remaining target labels whose candidate category vectors are not standard unit vectors or zero vectors, the category information is collected by polling.

5. The method for collecting RFID target tag category information based on dual time slots according to claim 4, characterized in that, In step (2.5), if the candidate class vector w id Update to all-zero vector: Determine if the target label id is a non-local label; the candidate category vector of non-local labels is eventually updated to a zero vector during the iterative information collection process of the reader.

6. The method for collecting RFID target tag category information based on dual time slots according to claim 4, characterized in that, In step (2.7), if the candidate class vector w id It is a unit vector e j ,0≤j≤θ-1: Classify the target label id into S j Otherwise, use a polling method to collect category information for these target labels.

7. The method for collecting RFID target tag category information based on dual time slots according to claim 6, characterized in that, If some non-local target label IDs are candidate category vectors w id If w accidentally becomes a unit vector in a certain round due to interference from other labels, then the logical AND operation in step (2.5) is used to update w. id w id It is subsequently updated to an all-zero vector.

8. The method for collecting RFID target tag category information based on dual time slots according to claim 7, characterized in that, Let n be the number of known labels remaining in the i-th round of the non-target label filtering stage. i The number of remaining target labels is k. i Test frame V i 1 The probability that one of the selected time slots is the expected empty time slot is: Test Frame V i 1 The probability that a selected time slot is allocated only to a non-target known label is: Analysis based on steps (1.3.1) and (1.3.2) The probability of occurrence is Similar analysis The probability of an unknown label being identified in the i-th round of the non-target label filtering stage in step (2.2) is:

9. An electronic device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the program to implement the steps of the method according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The device contains a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1-8.

Images