Ultra-high frequency RFID system data transmission method, device, equipment and storage medium

Abstract

The present invention relates to the technical field of data transmission, and discloses a data transmission method, apparatus, device, and storage medium for an ultra-high frequency (UHF) RFID system. The method comprises: performing bandpass filtering and orthogonal demodulation on multi-tag backscatter signals received by an RFID reader to obtain a tag feature data sequence; performing multi-dimensional feature analysis and timing correlation calculation to obtain an independent tag signal frame; performing boundary identification to determine a starting point of a tag data frame; performing continuous sampling analysis and tag signal conflict identification to output a conflict position sequence; performing matrix interleaving transformation and position rearrangement on data marked in the conflict position sequence to generate an anti-collision data packet; constructing a tag access scheduling queue based on the transmission timing and priority attributes of the anti-collision data packet, and outputting a multi-tag concurrent transmission timing. The present invention improves the data transmission performance of the ultra-high frequency RFID system through multi-level signal processing and an intelligent scheduling mechanism.

Description

Technical Field

[0001] The present invention relates to the field of data transmission technology, and in particular to an ultra-high frequency RFID system data transmission method, device, equipment and storage medium. Background Art

[0002] Ultra-high frequency RFID technology, with its advantages of non-contact, multi-target, and long-range identification, has been widely used in logistics tracking, asset management, and industrial production. However, in practical applications, when multiple tags simultaneously respond to reader queries, signal interference and collisions between tags significantly reduce the system's recognition efficiency and reliability. This is especially true in densely deployed environments, where traditional anti-collision algorithms struggle to meet the demands for efficient data transmission.

[0003] With the rapid development of IoT technology, RFID systems are facing requirements for higher data throughput and lower transmission latency. Existing data transmission methods primarily rely on simple random time slot allocation mechanisms, lacking in-depth analysis of tag signal characteristics and optimized scheduling of transmission resources. This results in a sharp decline in system performance in multi-tag concurrent communication scenarios, and an inability to guarantee the real-time and reliability of data transmission. Furthermore, traditional RFID data transmission schemes often employ fixed frame structures and synchronization mechanisms, which are prone to frame synchronization failures and data loss in poor channel conditions or densely distributed tags. Furthermore, existing retransmission mechanisms lack effective error control and resource scheduling strategies, exhibiting significant performance fluctuations when handling sudden conflicts, making it difficult to meet the requirements of high-reliability application scenarios. Summary of the Invention

[0004] The present invention provides an ultra-high frequency RFID system data transmission method, device, equipment and storage medium. The present invention improves the performance of ultra-high frequency RFID system data transmission through multi-level signal processing and intelligent scheduling mechanism.

[0005] In a first aspect, the present invention provides a method for transmitting data in an ultra-high frequency RFID system, the method comprising:

[0006] Bandpass filtering and orthogonal demodulation are performed on the multi-tag backscattered signals received by the RFID reader to obtain the tag feature data sequence;

[0007] Performing multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame;

[0008] Performing boundary identification on the independent tag signal frame to determine the starting point of the tag data frame;

[0009] Performing continuous sampling analysis and tag signal conflict identification on the signal after the starting point of the tag data frame, and outputting a conflict position sequence;

[0010] Performing matrix interleaving transformation and position rearrangement on the data marked by the conflict position sequence to generate an anti-collision data packet;

[0011] A tag access scheduling queue is constructed based on the transmission timing and priority attributes of the anti-collision data packet, and a multi-tag concurrent transmission timing is output.

[0012] In a second aspect, the present invention provides an ultra-high frequency RFID system data transmission device, the ultra-high frequency RFID system data transmission device comprising:

[0013] The filtering module is used to perform bandpass filtering and orthogonal demodulation on the multi-tag backscattered signals received by the RFID reader to obtain a tag feature data sequence;

[0014] A calculation module, configured to perform multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame;

[0015] An identification module, configured to identify the boundary of the independent tag signal frame and determine the starting point of the tag data frame;

[0016] An analysis module, configured to perform continuous sampling analysis on the signal after the starting point of the tag data frame and identify tag signal conflicts, and output a conflict position sequence;

[0017] a rearrangement module, configured to perform matrix interleaving transformation and position rearrangement on the data marked by the conflict position sequence to generate an anti-collision data packet;

[0018] The output module is used to construct a label access scheduling queue based on the transmission timing and priority attributes of the anti-collision data packet, and output the multi-label concurrent transmission timing.

[0019] The third aspect of the present invention provides an ultra-high frequency RFID system data transmission device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the ultra-high frequency RFID system data transmission device executes the above-mentioned ultra-high frequency RFID system data transmission method.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned UHF RFID system data transmission method.

[0021] In the technical solution provided by the present invention, 1. the time-space separation signal feature analysis and multi-dimensional feature mapping technology are adopted, combined with multi-level feature extraction and correlation analysis, to effectively improve the recognition accuracy in a multi-tag environment; 2. the innovative introduction of adaptive interleaving coding and multiple staggered transmission mechanisms enhances the system's anti-interference ability and data transmission reliability in a high-density tag environment; 3. a priority-based multi-level scheduling algorithm is designed, combined with a dynamic time slot allocation strategy, to reduce tag response conflicts and reduce transmission delays; 4. through layered staggered control and performance adaptive optimization, the system resource utilization efficiency is optimized and the data transmission throughput is improved; 5. multiple anti-collision coding and intelligent retransmission control are adopted to enhance the transmission stability of the system in complex environments.

[0022] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or understood by practicing the present invention. The purposes and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0023] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A schematic diagram of an embodiment of a method for transmitting data in an ultra-high frequency RFID system according to an embodiment of the present invention;

[0025] Figure 2 A schematic diagram of an embodiment of a UHF RFID system data transmission device according to an embodiment of the present invention;

[0026] Figure 3 Schematic diagram of an embodiment of an ultra-high frequency RFID system data transmission device in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0028] The terms "including," "having," and any variations thereof, as used in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to the process, method, product, or device.

[0029] To facilitate understanding of this embodiment, a method for transmitting data in an ultra-high frequency RFID system disclosed in an embodiment of the present invention is first introduced in detail. Figure 1 As shown, this method includes the following steps:

[0030] 101. Perform bandpass filtering and orthogonal demodulation on the multi-tag backscatter signals received by the RFID reader to obtain a tag feature data sequence;

[0031] It is understandable that the execution subject of the present invention may be a UHF RFID system data transmission device, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0032] Specifically, an RFID reader receives backscattered signals from multiple tags and then filters these signals through a bandpass filter with a center frequency of 915 MHz to produce a primary filtered signal. The bandpass filter selects valid signals within a specific frequency band from the received broadband signal, effectively removing ambient noise and irrelevant interference signals while retaining the useful components of the backscattered signal, resulting in the primary filtered signal. The primary filtered signal is sampled at four times the carrier frequency and then frequency-converted by a digital down-conversion circuit to produce an intermediate frequency (IF) signal. Quadruple carrier frequency sampling improves signal resolution and better preserves signal details. The digital down-conversion circuit frequency-adjusts the signal to an IF range, making it easier to process. The IF signal is input into a quadrature demodulator and multiplied by the in-phase and quadrature branches of the local carrier, respectively, to produce the in-phase component I and the quadrature component Q. The in-phase component I and the quadrature component Q represent the projection of the signal in the in-phase and quadrature directions, respectively. The quadrature demodulation process enables the system to effectively separate and process the different frequency components of the signal, extracting more information. The in-phase component I and the quadrature component Q are band-limited using a low-pass filter to produce baseband I and Q signals. The low-pass filter removes high-frequency components from the intermediate frequency signal while retaining the baseband portion. The resulting baseband I and Q signals contain low-frequency features directly related to the tag information. The baseband I and Q signals are digitally sampled using a 16-bit analog-to-digital converter to produce digitized I and Q data streams. Timing recovery is performed on the digitized I and Q data streams, and a sampling clock sequence is established through this timing recovery process to prevent signal distortion or information loss caused by offset sampling time points. Based on the sampling clock sequence, the digitized I and Q data streams are segmented into 16-bit RN16 blocks to produce I and Q data blocks. RN16 is 16 bits long and is the basic unit of length for tag response frames. The I and Q data blocks are combined to form a complex tag signature data sequence while preserving the signal's amplitude and phase information.

[0033] 102. Perform multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame;

[0034] Specifically, complex operations are performed on the tag feature data sequence, calculating the squared amplitude and phase angle of each data point to obtain the tag signal strength and phase sequences. These sequences represent the amplitude and phase characteristics of each tag signal in the complex plane and are crucial for identifying different tags. By sorting the tag signal strength sequence by signal energy, a valid signal point set is obtained, representing the most energetic portion of the received tag signal. This effectively removes interfering signals and noise. Arrival time information is extracted for each data point in the valid signal point set, capturing the specific temporal characteristics of the tag response. To improve the accuracy of this temporal information, a weighted averaging method based on signal strength is employed to weight the arrival time of each valid signal point, resulting in a tag response time sequence. This weighted averaging approach maximizes the impact of data points with higher signal strength on the response time, effectively reducing noise interference in the response time calculation and improving the accuracy of the tag response time sequence. The tag response time sequence is then input into a feature classifier, where time intervals are grouped using a K-means clustering algorithm to obtain a set of tag response cycles. The K-means clustering algorithm groups response times, classifying similar response times into categories. This effectively identifies the response cycles of individual tags and generates a set of cycles containing the response times of each tag. Based on the set of tag response cycles, an A×B-dimensional feature matrix is constructed, where A represents the number of tags and B represents the feature dimension. This matrix contains the amplitude, phase, and time information of each tag, resulting in a feature space mapping matrix. For each row of the feature space mapping matrix, the Pearson correlation coefficient is calculated with adjacent data frames. The Pearson correlation coefficient measures the similarity between two data frames. This quantifies the correlation between each tag signal frame to determine whether the response belongs to the same tag. A correlation threshold of 0.8 is set during the calculation of the Pearson correlation coefficient. Only frames with a correlation coefficient above 0.8 are considered highly correlated. This process effectively eliminates noise and irrelevant signals, retaining highly correlated frames for subsequent processing, resulting in a tag correlation matrix. Connected component analysis is performed on the tag correlation matrix, grouping frames with correlation coefficients above the set threshold to obtain a set of independent tag candidates. Connected component analysis (CCA) is a graph theory algorithm that combines highly correlated data frames into a connected component, thereby identifying and distinguishing signals with different labels. By combining highly correlated data frames, a set of multiple candidate label signals is generated, known as the independent label candidate set. Temporal consistency verification is performed on the independent label candidate set, checking whether each data frame meets a fixed response period. Data frames that do not meet the fixed period characteristics are eliminated to ensure that the output signal frames are independent label signal frames with good temporal consistency.

[0035] 103. Identify the boundary of the independent tag signal frame and determine the starting point of the tag data frame;

[0036] Specifically, a 16-bit cyclically shifted synchronization codeword is constructed for the independent tag signal frame, generating a standard synchronization codeword with a maximum and minimum Hamming distance to improve the synchronization process's anti-interference capability. The maximum and minimum Hamming distance is selected to ensure sufficient discrimination between synchronization codewords, effectively identifying the tag's starting boundary even in noisy environments. The standard synchronization codeword is cyclically shifted throughout its entire cycle to generate a set of 16 shifted versions of the synchronization codeword. This generation of multiple shifted versions increases matching flexibility, ensuring that a matching version can be found within the synchronization codeword set regardless of the signal's phase. The independent tag signal frame is segmented according to the length of the synchronization codeword, and each segment is normalized to produce the data sequence to be matched. Normalization eliminates the influence of intensity differences between different signal segments, ensuring that the correlation results rely solely on the relative shape of the signal and are unaffected by the absolute amplitude. The data sequence to be matched is correlated with each shifted version in the synchronization codeword set, and a correlation coefficient matrix is generated by calculating a 256-point sequence of correlation values. Peak detection is performed on each column of the correlation coefficient matrix to extract the location and amplitude information of the local maximum point to obtain a set of candidate synchronization points. Based on the amplitude distribution of the candidate synchronization point set, the optimal decision threshold is calculated to obtain a valid synchronization position sequence. By analyzing the amplitude distribution, a decision threshold is adaptively set to exclude pseudo-synchronization points caused by noise and only retain synchronization points with sufficiently high amplitude and high credibility. A majority vote is then performed on the synchronization points of three adjacent data frames in the valid synchronization position sequence to enhance the stability and robustness of the synchronization points. The most likely synchronization position is selected from multiple candidate positions, that is, the position with the highest frequency of occurrence is selected as the final frame synchronization position. The data frame is segmented and reorganized based on the frame synchronization position, and the independent label signal frame is segmented into data units of fixed length according to the synchronization position, and the starting point of the label data frame is output.

[0037] 104. Perform continuous sampling analysis and tag signal conflict identification on the signal after the starting point of the tag data frame, and output a conflict position sequence;

[0038] Specifically, the signal after the start of the tag data frame is segmented into sliding windows. The window length is set to the length of RN16, and the sliding step is set to one symbol, resulting in a continuous sequence of signal windows. This sliding window segmentation approach helps gradually extract signal features and capture subtle changes in the signal sequence, achieving more accurate collision detection. Complex envelope detection is performed on each window data in the continuous signal window sequence to extract the signal's amplitude and phase traces, generating a trajectory sequence containing signal features. Complex envelope detection aims to separate the signal's amplitude and phase characteristics, facilitating analysis of signal variation patterns and collision identification. A differential operation is performed on the signal feature trajectory sequence to calculate the amplitude change rate and phase jump amount between adjacent sampling points, resulting in a signal variation feature sequence. The differential operation reveals the signal's variation characteristics along the time axis. The amplitude change rate reflects sudden changes in signal strength, while the phase jump amount reveals discontinuous changes in signal phase. The signal variation feature sequence is input into a dual-threshold detector, with the amplitude change rate threshold set to 20% and the phase jump threshold set to 45 degrees. A preliminary collision decision result is obtained by comparing the signal feature sequence with these thresholds. Signal vector diagram analysis is performed on the primary collision decision results. The presence of a collision region is determined by counting the number of cluster centers in the signal space. In the signal vector space, if the number of cluster centers in a region is greater than one, it indicates the presence of multiple sets of signal features, caused by overlapping signals from multiple labels. These regions are then marked as potential collision points, resulting in a set of collision point candidates. Based on the collision point candidate set, a collision feature vector is constructed containing multiple features, including amplitude abruptness, phase jump, and cluster dispersion. These features describe the signal variation in the collision region, generating a collision feature sequence. Amplitude abruptness reflects the dramatic change in signal intensity at the collision point, phase jump describes discontinuous phase changes, and cluster dispersion assesses the concentration of signal clusters. Together, these features comprehensively describe the signal behavior in the collision region. The collision feature sequence is then input into a support vector machine classifier for collision event identification, resulting in a collision event label sequence. As a machine learning classification algorithm, the support vector machine uses the input feature vector to determine whether a signal segment represents a collision event, thereby classifying each signal segment and outputting a collision label sequence. Machine learning algorithms are used to accurately classify collision events, avoiding misjudgments or missed detections caused by simple threshold judgments, and improving the accuracy of overall collision identification. Time-domain correlation analysis is performed on the collision event marker sequence, and adjacent collision event markers are merged to generate the final collision location sequence.

[0039] 105. Perform matrix interleaving transformation and position rearrangement on the data marked with the conflict position sequence to generate an anti-collision data packet;

[0040] Specifically, the data marked with the collision position sequence is segmented into a fixed length of 8 bits to construct a data unit sequence, resulting in a set of data units to be retransmitted. The set of data units to be retransmitted is then populated into an M×N interleaving matrix in row-first order, where M represents the data unit length and N represents the currently detected collision frequency. Data is then sequentially populated into each row of the matrix in row-first order, resulting in an initial interleaving matrix. The construction of the interleaving matrix aims to improve collision resistance by rearranging the data. In particular, when multiple labels collide, this data scattering effectively reduces the probability of repeated collisions among certain data units, ensuring the stability and reliability of data transmission in the system. A column-by-column cyclic shift operation is performed on the initial interleaving matrix, with the shift amount linearly proportional to the column number. This generates a position-scattered interleaved data matrix and a rearranged data matrix. The column-by-column cyclic shift operation cyclically shifts the data in each column by a specific shift amount, scrambling the data positions so that data units originally in the same position are distributed throughout the matrix. This position scattering strategy reduces the likelihood of collisions by increasing the randomness of the data, thereby improving the anti-interference capability of the transmitted data. The rearranged data matrix is read out in column-major order, and the readout data is grouped into 4-bit groups to produce the data groups to be encoded. Hamming coding is performed on the data groups to generate a 3-bit parity bit for each data group. This parity bit is then appended to the original data to produce the encoded data sequence. Hamming coding is introduced to provide error correction capabilities. By adding redundant parity bits, errors can be detected and corrected during data transmission, improving data transmission reliability. A packet header is constructed based on the encoded data sequence. The packet header contains information about the interleaving depth, packet length, and original position identifier, providing essential auxiliary information for subsequent data decoding and recovery. The interleaving depth indicates the degree of data interleaving, the packet length helps the receiver determine data integrity, and the original position identifier provides a basis for data restoration during decoding. The packet header and the encoded data sequence are combined and encapsulated into a complete data packet according to a fixed format. A 32-bit cyclic redundancy check (CRC) is added to the packet. The CRC verifies the contents of the entire packet to detect errors during transmission. CRC is an error-checking mechanism used in communication systems. It performs specific mathematical operations on data to generate a checksum and append it to the packet. At the receiving end, the same calculation is used to check the received data, determine if there are any transmission errors, and take appropriate error correction measures. The above steps ultimately result in a collision-proof packet.

[0041] 106. Based on the transmission timing and priority attributes of the anti-collision data packet, a tag access scheduling queue is constructed to output the multi-tag concurrent transmission timing.

[0042] Specifically, a first-level priority calculation is performed on anti-collision packets, assigning a weight of 0.4 to the tag data volume, a weight of 0.3 to the number of retransmissions, and a weight of 0.3 to the waiting time. A weighted sum function is then used to calculate the first priority sequence for each tag. A second-level priority calculation is then performed on this first priority sequence, applying exponential weighting based on transmission latency requirements to produce a new tag priority sequence. The introduction of exponential weighting amplifies the priority of latency-sensitive tags, prioritizing the transmission of these time-sensitive tags, ensuring system responsiveness and overall efficiency. The tag priority sequence is input into a dual-threshold selector, with the first threshold set at 80% of the system's processing capacity and the second threshold set at the average priority value. Only tags that meet both thresholds are selected into the first transmission queue; tags that meet only one of the thresholds are selected into the second transmission queue, resulting in a hierarchical set of tags to be scheduled. This dual-threshold selection approach allows for more detailed tag screening, rationally allocating system resources, preventing system overload, and improving overall transmission efficiency and fairness. For the hierarchical set of tags to be scheduled, a two-level clustering algorithm is applied to the tags, coarsely grouping them, and then finely dividing each coarse group. After two levels of clustering, a time slot protection factor β = 0.1 is set between groups to minimize the probability of conflicts between different groups, resulting in a multi-level transmission group structure. The multi-level grouping method can effectively manage transmission conflicts between a large number of tags and achieve optimal scheduling of system resources by rationally planning time slot allocation within and between tag groups. Based on the multi-level transmission group structure, a time slot resource allocation tree is constructed. The first-level nodes of the time slot resource allocation tree are used to allocate basic time slot resources, the second-level nodes are responsible for allocating off-peak intervals for transmission within the group, and the third-level nodes specifically allocate the transmission time for each tag, thus forming a hierarchical time slot allocation scheme. The hierarchical time slot allocation scheme undergoes triple peak-staggering optimization. The scheme optimizes inter-group peak-staggering to minimize conflicts caused by simultaneous transmissions from different groups by setting appropriate peak-staggering between groups. It then optimizes intra-group peak-staggering to minimize simultaneous transmissions between tags within a group. Finally, it optimizes inter-node peak-staggering to ensure that each specific tag transmits at the most appropriate time. This triple peak-staggering optimization generates a multidimensional peak-staggering control sequence that minimizes the probability of conflicts. Based on this multidimensional peak-staggering control sequence, a three-level scheduling queue is established: the first level is the urgent transmission queue, used to transmit tags with high latency requirements; the second level is the standard transmission queue, used for tags with general priority; and the third level is the low-priority transmission queue, used to transmit data with less stringent latency requirements. Differentiated scheduling periods are assigned to each queue, resulting in a hierarchical scheduling control sequence.Based on the hierarchical scheduling control sequence, a performance monitoring matrix is established, consisting of three main monitoring points: the first monitoring point is set at the queue switching location to monitor the smooth switching between queues of different priorities; the second monitoring point is set at the time slot boundary to ensure that each time slot is used as expected and there are no conflicts; and the third monitoring point is set at the transmission completion location to assess the data transmission success rate and overall latency. Through the combined analysis of these three monitoring points, multi-dimensional performance evaluation results are obtained. Based on the results of this multi-dimensional performance evaluation, a three-level dynamic adjustment mechanism is triggered: the first level adjusts the division of transmission groups to adapt to the changing system load by re-dividing the tag groups; the second level adjusts the allocation of time slots to optimize time slot usage based on transmission performance feedback; and the third level adjusts the peak shifting parameters to reduce conflicts by adjusting the peak shifting strategy. All adjustments are controlled by an exponential backoff algorithm. This allows for dynamic and flexible scheduling optimization based on the actual system operation, outputting a multi-tag concurrent transmission schedule to achieve efficient and stable data transmission.

[0043] In the embodiments of the present invention, 1. the time-space separation signal feature analysis and multi-dimensional feature mapping technology are adopted, combined with multi-level feature extraction and correlation analysis, to effectively improve the recognition accuracy in a multi-tag environment; 2. the innovative introduction of adaptive interleaving coding and multiple staggered transmission mechanisms enhances the system's anti-interference capability and data transmission reliability in a high-density tag environment; 3. a priority-based multi-level scheduling algorithm is designed, combined with a dynamic time slot allocation strategy, to reduce tag response conflicts and reduce transmission delays; 4. through layered staggered control and performance adaptive optimization, the system resource utilization efficiency is optimized and the data transmission throughput is improved; 5. multiple anti-collision coding and intelligent retransmission control are adopted to enhance the transmission stability of the system in complex environments.

[0044] In a specific embodiment, the process of executing step 101 may specifically include the following steps:

[0045] The multi-tag backscattered signals are received by an RFID reader and filtered by a bandpass filter with a center frequency of 915 MHz to obtain a primary filtered signal.

[0046] The primary filtered signal is sampled at four times the carrier frequency and frequency converted through a digital down-conversion circuit to obtain an intermediate frequency signal. The intermediate frequency signal is input into an orthogonal demodulator and multiplied with the in-phase branch and quadrature branch of the local carrier respectively to obtain the in-phase component I and the quadrature component Q;

[0047] The in-phase component I and the quadrature component Q are band-limited by a low-pass filter to obtain a baseband I signal and a baseband Q signal, and the baseband I signal and the baseband Q signal are digitally sampled by a 16-bit analog-to-digital converter to obtain a digital I data stream and a digital Q data stream;

[0048] Performing timing recovery on the digitized I-channel data stream and the digitized Q-channel data stream, establishing a sampling clock sequence, and segmenting the digitized I-channel data stream and the digitized Q-channel data stream according to the 16-bit RN16 length based on the sampling clock sequence to obtain I-channel data blocks and Q-channel data blocks;

[0049] The I-way data block and the Q-way data block are combined into a label feature data sequence in a complex form.

[0050] Specifically, when the RFID reader receives multiple tag signals, it will receive various clutter and interference signals. In order to effectively extract useful backscatter information, the center frequency of the bandpass filter is set to 915MHz, which is the operating frequency of the ultra-high frequency RFID system. The received original signal is processed by the bandpass filter to filter out noise and interference that are not within the operating frequency band, and only the effective frequency band signal containing the tag information is retained. The primary filtered signal is sampled at four times the carrier frequency. Assume that the carrier frequency is , its value is 915MHz, and the four times sampling frequency is . High-frequency sampling helps to retain the detailed characteristics of the signal and better capture the information in the tag backscattered signal. The sampling process is described by the Nyquist theorem, which requires that the sampling frequency be at least twice the highest frequency of the signal in order to fully restore the original signal. Therefore, four times the carrier frequency is used for sampling to ensure that no information is lost. The sampled signal is input into the digital down-conversion circuit, and the intermediate frequency signal is obtained through frequency conversion. Assume that the original received signal is , its expression is expressed as:

[0051] ;

[0052] in, represents the amplitude change of the signal, represents the phase change of the signal, The digital down-conversion process is to mix the signal with the local oscillator signal to reduce the frequency and convert the signal to an intermediate frequency range. Assume that the local oscillator signal is , the intermediate frequency signal obtained after down-conversion is:

[0053] ;

[0054] in, The intermediate frequency signal is input into the quadrature demodulator and multiplied by the in-phase branch and quadrature branch of the local carrier respectively to obtain the in-phase component. and orthogonal components The function of the quadrature demodulator is to decompose the intermediate frequency signal into two independent components. These two components describe the amplitude and phase changes of the signal on the complex plane. The demodulated signal of the in-phase branch is , the demodulated signal of the orthogonal branch is After quadrature demodulation, we get the in-phase component and the quadrature component:

[0055] ;

[0056] Through these two components and , describing the amplitude and phase characteristics of the signal. The in-phase component is filtered by a low-pass filter. and orthogonal components Perform band-limiting processing to obtain baseband Road signals and The purpose of low-pass filtering is to remove the high-frequency components in the signal and retain the baseband part related to the tag information. The signals after low-pass filtering are recorded as and , contains the main information of the signal. Use a 16-bit analog-to-digital converter to convert the baseband Road signals and The signal is digitally sampled to obtain digital Data flow and data stream. Assume the sampling frequency is , the sampled digital signals are:

[0057] ;

[0058] ;

[0059] in, is the sampling period, is the serial number of the sampling point. Through 16-bit analog-to-digital conversion, high quantization accuracy is ensured, signal details are retained, and the accuracy of signal processing is improved. Data flow and The timing of the data stream is recovered and the sampling clock sequence is established. The appropriate sampling time is extracted from the received signal to ensure that the sampling point of the data stream can be aligned with the modulation characteristics of the original signal to minimize the sampling error. Let the recovered sampling clock be , the corresponding digital data is expressed as:

[0060] ;

[0061] The sampling points obtained by timing recovery ensure stable data sampling, thereby accurately extracting the characteristics of the signal. Data flow and The data stream is divided into 16-bit RN16 lengths to obtain Road data block and RN16 is a basic response length unit in the RFID system, representing 16 bits of data. The data is divided according to the length of RN16 to ensure that each data block contains a complete tag response information. Road data block and The data blocks are combined into a complex tag feature data sequence. Each data block is represented as a complex number, where is the real part, is the imaginary part:

[0062] ;

[0063] in, Is the imaginary unit, representing the imaginary part of a complex number. This complex number representation preserves both the amplitude and phase information of the signal, making it easier to perform multidimensional analysis of label features. For example, the modulus and phase of a complex number are represented as:

[0064] ;

[0065] ;

[0066] in, represents the amplitude of the signal, Indicates the phase angle of the signal. This complex number tag feature data sequence provides complete amplitude and phase information for subsequent tag decoding and feature analysis.

[0067] In a specific embodiment, the process of executing step 102 may specifically include the following steps:

[0068] Perform complex operations on the tag feature data sequence, calculate the square of the amplitude and phase angle of each data point, obtain the tag signal strength sequence and phase sequence, and sort the tag signal strength sequence according to the signal energy to obtain the effective signal point set;

[0069] The arrival time information of each data point in the valid signal point set is extracted, and the time information is weighted averaged according to the signal strength to obtain the tag response time sequence;

[0070] The tag response time series is input into the feature classifier, and the time intervals are grouped based on the K-means clustering algorithm to obtain the tag response period set;

[0071] According to the tag response period set, an A×B dimensional feature matrix is constructed, where A is the number of tags and B is the feature dimension, which contains amplitude, phase and time information, to obtain the feature space mapping matrix;

[0072] Calculate the Pearson correlation coefficient of each row of the feature space mapping matrix with the adjacent data frame, and set the correlation threshold to 0.8 to obtain the label correlation matrix;

[0073] Connected component analysis is performed based on the label correlation matrix, and data frames with correlation coefficients higher than the threshold are combined to obtain an independent label candidate set. Temporal consistency verification is then performed on the independent label candidate set, and data frames that do not meet the fixed response period are eliminated, and independent label signal frames are output.

[0074] Specifically, the square of the amplitude and the phase angle of each data point are calculated to obtain the intensity sequence and phase sequence of the tag signal. Assume that the complex number representation of each data point is ,in and are the in-phase and quadrature components, is an imaginary unit. The square of the amplitude (i.e., the strength of the signal) is calculated using the following formula:

[0075] ;

[0076] in, Represents the square of the signal amplitude, that is, the strength of the signal, and are the real and imaginary parts of the data point respectively. By calculating the square of the amplitude of each data point, the tag signal strength sequence is obtained. The phase angle is calculated using the following formula:

[0077] ;

[0078] in, Represents the phase angle of the data point. Through these two formulas, we can get the signal strength and phase information respectively. Sort the tag signal strength sequence according to the signal energy to get the valid signal point set. Let the signal strength sequence be By sorting these signal intensities from large to small, a new sequence is obtained ,in ( ). By sorting, the signals with lower energy are filtered out, and only the valid signal points with higher signal energy are retained to form a valid signal point set. The arrival time information of each data point in the valid signal point set is extracted. Let the first data point in the valid signal point set be The arrival time of the data point is , and its corresponding signal strength is In order to improve the accuracy of the signal arrival time, the signal strength is used to perform a weighted average of the arrival time. The calculation formula is:

[0079] ;

[0080] in, is the weighted average arrival time, is the number of valid signal points, For the The strength of the signal point, is the corresponding arrival time. By weighted averaging, a more stable and accurate tag response time sequence is obtained, which reduces the impact of noise and unstable signals on the response time calculation. The tag response time sequence is input into the feature classifier, and the time intervals are grouped based on the K-means clustering algorithm to obtain the tag response period set. The tag response time is divided into several groups, each group represents a tag's characteristic response period. Let the tag response time sequence be , these response times are divided into clusters, the objective function of clustering is expressed as:

[0081] ;

[0082] in, is the objective function, For the clusters, For the By minimizing the objective function , dividing the response time into Clusters, each cluster represents the response period of a tag. According to the tag response period set, a dimensional feature matrix, where is the number of labels, is the dimension of the feature, including amplitude, phase and time information. Let the feature matrix be , then Row, No. The elements of a column are represented as:

[0083] ;

[0084] in, Indicates the of the tags eigenvalues, including amplitude , Phase and time information In this way, the constructed feature space mapping matrix can fully describe the multi-dimensional features of each label. The Pearson correlation coefficient with the adjacent data frame is calculated for each row of the feature space mapping matrix. The Pearson correlation coefficient is used to measure the correlation between two vectors. Let the first row of the feature space mapping matrix be Behavior , No. Behavior , then the Pearson correlation coefficient is expressed as:

[0085] ;

[0086] in, For the Row and Pearson correlation coefficient between rows, and Respectively Row and By calculating the Pearson correlation coefficient, we can get the label correlation matrix , the elements of the matrix Indicates the Tags and Based on the label correlation matrix, a connected component analysis is performed to combine the data frames with correlation coefficients higher than the threshold (set to 0.8) to obtain the independent label candidate set. , then it is considered that Tags and Tags belonging to the same connected component are merged into an independent tag candidate set. The independent tag candidate set is verified for temporal consistency, and data frames that do not meet the fixed response period are eliminated to obtain the final independent tag signal frame.

[0087] In a specific embodiment, the process of executing step 103 may specifically include the following steps:

[0088] Construct a 16-bit cyclic shift synchronization codeword for the independent tag signal frame to generate a standard synchronization codeword with the maximum and minimum Hamming distance, and perform a full-cycle cyclic shift operation on the standard synchronization codeword to generate a set of 16 shifted versions of the synchronization codeword;

[0089] The independent tag signal frame is segmented according to the synchronization codeword length, and each segment of the signal sequence is normalized to obtain the data sequence to be matched. The data sequence to be matched is then correlated with each shifted version in the synchronization codeword set to calculate a 256-point correlation value sequence and obtain a correlation coefficient matrix.

[0090] Perform peak detection on each column of the correlation coefficient matrix, extract the position and amplitude information of the local maximum point, obtain a set of candidate synchronization points, and calculate the optimal decision threshold based on the amplitude distribution of the candidate synchronization point set to obtain a valid synchronization position sequence;

[0091] A majority vote is performed on the synchronization points of three adjacent data frames in the valid synchronization position sequence, and the position with the highest frequency is selected to obtain the frame synchronization position. The data frame is divided and reorganized based on the frame synchronization position, and the independent label signal frame is divided into data units of fixed length according to the synchronization position, and the starting point of the label data frame is output.

[0092] Specifically, a 16-bit cyclic shift synchronization codeword is constructed for the independent tag signal frame to generate a standard synchronization codeword with the maximum and minimum Hamming distance. The Hamming distance is an indicator to measure the degree of difference between two binary sequences. It is defined as the number of different bits in two binary sequences of the same length. In order to ensure the robustness of the synchronization codeword in different signal scenarios, the maximum and minimum Hamming distance is selected to construct the synchronization codeword, which effectively improves the reliability of frame synchronization in a noisy environment. Assume that the synchronization codeword is ,in Represents each bit of the codeword. Through the optimization algorithm, a binary sequence of length 16 is found to maximize the minimum Hamming distance between it and other possible codewords, thereby improving the signal's anti-interference ability. Perform a full-cycle cyclic shift operation to generate a set of 16 shifted versions of the synchronization codeword. , shift all its bits to the right by one bit each time, and move the last bit to the first bit, repeat this operation 16 times, and get 16 shifted versions. The codeword after the shift is , get a set of synchronous code words , used for subsequent matching operations. The independent tag signal frame is segmented according to the length of the synchronization codeword, and the signal frame is divided into multiple segments with a length of 16 bits. Assume that the independent tag signal frame is , which is divided into several segments of length 16, and each segment signal is represented as ,in is the sequence number of the segment. For each signal sequence Normalization is performed to eliminate the influence of amplitude and make the amplitude of the signal comparable at the same scale. The normalized signal is expressed as:

[0093] ;

[0094] in, Signal sequence The Euclidean norm of is defined as:

[0095] ;

[0096] Normalized signal sequence The amplitude of is adjusted to 1 to facilitate correlation with the standard synchronization codeword. The normalized data sequence to be matched is correlated with each shifted version in the synchronization codeword set, and a 256-point correlation value sequence is calculated to form a correlation coefficient matrix. Suppose the normalized data sequence to be matched is , the first in the synchronization codeword set The shifted version is , then the result of the related operation is:

[0097] ;

[0098] in, Indicates the The signal sequence of the first The correlation between the synchronization codeword versions, and Represents the first By calculating the synchronization codewords for all segments and all shifted versions, a correlation coefficient matrix is obtained. , where each element Indicates the Segment signal and The correlation between synchronization codewords. Perform peak detection on each column of the correlation coefficient matrix, extract the position information and amplitude information of the local maximum point, and obtain a set of candidate synchronization points. Peak detection finds the signal position with the maximum correlation with the synchronization codeword. These positions are most likely to be the starting points of the tag data frame. The sequence of related values for the columns is By detecting the local maximum among these values, the position and amplitude information of the candidate synchronization points are extracted to obtain the candidate synchronization point set ,in Indicates the The location of candidate synchronization points, Represents the corresponding correlation amplitude. According to the amplitude distribution of the candidate synchronization point set, the optimal decision threshold is calculated to obtain a valid synchronization position sequence. Assume that the amplitude of the candidate synchronization point set is , by analyzing the distribution of these amplitudes, a decision threshold is set , only when the amplitude is higher than The synchronization point is considered to be a valid synchronization point, and a valid synchronization position sequence is obtained. A majority vote is performed on the synchronization points of three adjacent data frames in the valid synchronization position sequence to select the position with the highest frequency as the final frame synchronization position. The data frame is segmented and reassembled based on the frame synchronization position, and the independent tag signal frame is segmented into fixed-length data units according to the synchronization position, and the starting point of the tag data frame is finally output.

[0099] In a specific embodiment, the process of executing step 104 may specifically include the following steps:

[0100] The signal after the starting point of the tag data frame is segmented into sliding windows, the window length is set to RN16 length, and the sliding step is 1 code element to obtain a continuous signal window sequence;

[0101] Perform complex envelope detection on each window data in the continuous signal window sequence, extract the signal amplitude trajectory and phase trajectory, and obtain a signal feature trajectory sequence. Perform differential operation on the signal feature trajectory sequence, calculate the amplitude change rate and phase jump amount between adjacent sampling points, and obtain a signal change feature sequence.

[0102] The signal change feature sequence is input into the dual-threshold detector, the amplitude change rate threshold is set to 20%, and the phase jump threshold is set to 45 degrees to obtain the primary collision judgment result;

[0103] Perform signal vector diagram analysis on the primary collision judgment results, count the number of cluster centers in the signal space, mark the area with more than 1 cluster center as a potential collision point, and obtain the collision point candidate set;

[0104] Based on the collision point candidate set, a collision feature vector is constructed, which includes three feature quantities: amplitude mutation, phase jump and cluster dispersion, and a collision feature sequence is obtained;

[0105] The collision feature sequence is input into the support vector machine classifier for collision event recognition to obtain a collision event marker sequence. The collision event marker sequence is then subjected to time domain correlation analysis, and adjacent collision event markers are merged to generate a conflict position sequence.

[0106] Specifically, the signal after the starting point of the tag data frame is segmented into sliding windows, the window length is set to RN16 length, the sliding step is one code element, and a continuous signal window sequence is obtained. Assuming that the RN16 length is 16 code elements, the input signal sequence is , by sliding the window with a sliding step of 1 code element, we get a signal window of length 16 ,in Is the starting position of the sliding window, ranging from In this way, a continuous signal window sequence containing the complete characteristics of the signal is obtained. Complex envelope detection is performed on each window in the continuous signal window sequence to map the signal from the real domain to the complex domain. Assume that the data of each signal window is , through the complex envelope operation, each signal point is expressed as a complex form:

[0107] ;

[0108] in, Indicates the In the window The complex representation of the signal points, and are the in-phase and quadrature components of the signal, respectively. is the imaginary unit. By calculating the complex envelope, we can get the amplitude trajectory and phase trajectory of the signal, where the amplitude and counters They are expressed by the following formulas:

[0109] ;

[0110] ;

[0111] By performing complex envelope detection on each window, the amplitude trajectory sequence of the signal is obtained and phase trajectory sequence These features describe the change of the signal in the time domain. Perform differential operation on the signal feature trajectory sequence to calculate the amplitude change rate and phase jump between adjacent sampling points. Let the signal amplitude trajectory be And the phase trajectory is , amplitude change rate and phase jump variable Calculated by the following formulas:

[0112] ;

[0113] ;

[0114] in, Indicates the In the window Hedi The amplitude change rate between sampling points, Represents the amount of phase change. By performing differential operations between each sampling point, the signal change feature sequence is obtained to reflect the signal mutation. Especially in a multi-label environment, these features are used to identify signal conflicts and collisions. The signal change feature sequence is input into the dual-threshold detector, and the amplitude change rate threshold is set to 20%, and the phase jump threshold is set to 45 degrees. The dual-threshold detector determines whether the signal has changed dramatically by the preset amplitude and phase change thresholds, thereby preliminarily determining whether there is a collision of tag signals. Assume that the amplitude change rate threshold is , the phase jump threshold is If the amplitude change rate of a signal point or phase jump variable , then the point is considered to be a potential collision point, and the primary collision judgment result is obtained. For the primary collision judgment result, perform signal vector diagram analysis, and determine the collision area by counting the number of cluster centers in the signal space. Plot the signal change feature sequence as points in the feature space, analyze the distribution of these points, and use a clustering algorithm (such as K-means clustering) to count the number of cluster centers of the signal change features. If the number of cluster centers in a certain area is greater than 1, it is considered that there is a signal collision in the area, and it is marked as a potential collision point to obtain a collision point candidate set. Based on the collision point candidate set, a collision feature vector is constructed, which includes three feature quantities: amplitude mutation degree, phase jump degree, and cluster discreteness, to obtain a collision feature sequence. Suppose the amplitude mutation degree of a collision point is , the phase jump degree is , the cluster dispersion is These characteristics reflect the amplitude change degree, phase change degree and signal clustering concentration degree of the collision point. Through the comprehensive analysis of these characteristics, a complete collision feature vector is formed:

[0115] ;

[0116] The collision feature sequence is input into the support vector machine (SVM) classifier for collision event recognition to obtain the label sequence of the collision event. SVM is a machine learning classifier that is suitable for binary classification and multi-classification problems. Determine whether a collision event occurs in the current signal and output a sequence of markers containing whether each time window is a collision ,in , 0 means no collision, 1 means collision occurs. Time domain correlation analysis is performed on the collision event marker sequence to merge adjacent collision event markers and generate a conflict position sequence. Suppose the collision event marker sequence is If several adjacent marks are all 1, these positions are merged into a single conflict event, and the conflict position sequence is finally obtained , where each Indicates the starting position of a collision event.

[0117] In a specific embodiment, the process of executing step 105 may specifically include the following steps:

[0118] The data marked by the conflict position sequence is divided into 8-bit fixed lengths to construct a data unit sequence and obtain a set of data units to be retransmitted;

[0119] Fill the set of data units to be retransmitted into an M×N interleaving matrix in row-first order, where M is the data unit length and N is the currently detected collision frequency value, to obtain an initial interleaving matrix;

[0120] Perform a column cyclic shift operation on the initial interleaving matrix, with the shift amount being linearly related to the column number, to generate a position-scattered interleaving data matrix, obtaining a rearranged data matrix, reading the rearranged data matrix in column-priority order, and grouping the read data, with each group of data being 4 bits long, to obtain a data group to be encoded;

[0121] Performing Hamming coding on the coded data group to generate a 3-bit check bit, and adding the check bit to the original data to obtain a coded data sequence, constructing a data packet header based on the coded data sequence, including interleaving depth information, data packet length, and original position identifier, to obtain a data packet header;

[0122] Combine the data packet header with the encoded data sequence, encapsulate it into a data packet according to a fixed format, add a CRC check code, generate a 32-bit cyclic redundancy check code, and output an anti-collision data packet.

[0123] Specifically, the data marked by the conflict position sequence is divided into 8-bit fixed lengths to construct a data unit sequence and obtain a set of data units to be retransmitted. ,in Represents each bit. The data unit obtained by dividing into groups of 8 bits is expressed as ,in is the index of the data unit, ranging from By this segmentation method, the original conflicting data is divided into multiple 8-bit data units, ensuring that subsequent data processing is performed at a fixed length, thereby simplifying the encoding and interleaving operations. The data units to be retransmitted are filled into a row-priority order. in the interwoven matrix. Indicates the length of the data unit, that is, the number of bits contained in each data unit, and its value is 8; Indicates the currently detected collision frequency value, reflecting the frequency of tag collision during data transmission. The initial interleaving matrix is obtained by filling the data unit into the interleaving matrix row by row in the order of row priority. , each element of which Indicates the Row, No. The bit value in the data unit of the column. The padding operation is expressed as:

[0124] ;

[0125] in, , By this filling method, the data is filled into the matrix in row order to form an initial interleaving matrix. The initial interleaving matrix is subjected to column circular shift operation. The shift amount is linearly related to the column number, which means that the first The column shift amount is ,in is a constant, usually less than The value of is to ensure that the column data after shifting is not lost. The column shift amount is , the shifted matrix elements are expressed as:

[0126] ;

[0127] in, represents the rearranged matrix elements, Represents the new row index after circular shift. Through this column circular shift, the order of data in the original matrix is disrupted, generating an interleaved data matrix with scattered positions, and obtaining a rearranged data matrix. The position scattering operation aims to disperse the data in the original order to reduce the impact of continuous collisions and improve the success rate of data retransmission. The rearranged data matrix is read out in column priority order. That is, starting from the first column, each element is read column by column, and a new data sequence is obtained after all the data are read out. Let the read data be , these data are then grouped into groups of 4 bits each to obtain the data groups to be encoded ,in Represents the index of the data group. For each data group to be encoded, perform Hamming coding. Hamming coding is a coding method that can detect and correct single-bit errors. By adding redundant check bits, the receiving end can detect and correct single-bit errors that occur during data transmission. Suppose the data group to be encoded is , the process of Hamming encoding is to generate 3 check bits , these check digits are calculated using the following linear equations:

[0128] ;

[0129] ;

[0130] ;

[0131] in, Represents a bitwise exclusive OR operation. Added to the original data group to obtain a 7-bit coded data sequence The data sequence after Hamming coding can effectively improve the reliability of data transmission and ensure that even if a single-bit error occurs during transmission, it can be corrected by the check bit. The data packet header is constructed based on the coded data sequence. The data packet header contains information such as interleaving depth information, data packet length, and original position identifier. Assume that the interleaving depth is , the length of the data packet is , the original position is marked as , then the packet header is expressed as Interweaving Depth Indicates the level at which data is interleaved, and the length of the data packet Indicates the number of bits in the entire data packet, while the original position identifier It is used to indicate the location of the data in the original conflicting data so that the receiving end can restore the data to the correct location. With the encoded data sequence Combined together, encapsulated into a complete data packet according to a fixed format, and a cyclic redundancy check code (CRC) is added to the end of the data packet. CRC is an error detection method used in digital communications. It generates a check code by performing a specific polynomial operation on the data packet and appends it to the end of the data packet so that the receiving end can verify the integrity of the data. Suppose the data packet is , the generation of CRC check code is expressed as:

[0132] ;

[0133] in, represents the polynomial corresponding to the data packet, is a predefined generating polynomial, The 32-bit CRC checksum is generated. The CRC checksum is added to the end of the data packet to form the final anti-collision data packet. .

[0134] In a specific embodiment, the process of executing step 106 may specifically include the following steps:

[0135] Perform a first-level priority calculation on the anti-collision packets, assigning a weighting factor of 0.4 to the label data volume, a weighting factor of 0.3 to the number of retransmissions, and a weighting factor of 0.3 to the waiting time. Based on the weighted sum function, a first priority sequence is obtained. Perform a second-level priority calculation on the first priority sequence, applying exponential weighting based on the transmission delay requirement to obtain a label priority sequence.

[0136] The tag priority sequence is input into a dual-threshold selector. The first threshold is set to 80% of the system processing capacity, and the second threshold is set to the priority average. Tags that meet both thresholds are selected into the first transmission queue, and tags that meet only a single threshold are selected into the second transmission queue. This results in a hierarchical set of tags to be scheduled.

[0137] A two-level clustering algorithm is performed on the hierarchical set of labels to be scheduled. The first level performs coarse grouping of the labels, and the second level performs fine division of each coarse group. The time slot protection factor β = 0.1 is set between groups to obtain a multi-level transmission group structure.

[0138] A time slot resource allocation tree is constructed based on a multi-level transmission group structure. The first-level nodes allocate basic time slot resources, the second-level nodes allocate peak-shifting intervals within the group, and the third-level nodes allocate specific transmission times, thus obtaining a hierarchical time slot allocation scheme.

[0139] A triple peak-shifting optimization is performed on the hierarchical time slot allocation scheme: the first optimization is inter-group peak-shifting, the second optimization is intra-group peak-shifting, and the third optimization is node peak-shifting, generating a multi-dimensional peak-shifting control sequence that minimizes the probability of conflict.

[0140] Based on the multi-dimensional peak-shifting control sequence, a three-level scheduling queue is established: the first level is the emergency transmission queue, the second level is the standard transmission queue, and the third level is the low-priority transmission queue. Differentiated scheduling cycles are assigned to each queue, resulting in a hierarchical scheduling control sequence.

[0141] A performance monitoring point matrix is set based on a hierarchical scheduling control sequence, including: the first monitoring point is set at the queue switching, the second monitoring point is set at the time slot boundary, and the third monitoring point is set at the transmission completion point. A multi-dimensional performance evaluation result is obtained through the joint analysis of the three points;

[0142] A three-level dynamic adjustment mechanism is triggered based on the results of multi-dimensional performance evaluation: the first level adjusts the transmission group division, the second level adjusts the time slot allocation, and the third level adjusts the peak shifting parameters. An exponential backoff algorithm is used to control the adjustment period and output the multi-tag concurrent transmission timing.

[0143] Specifically, the data volume, number of retransmissions, and waiting time of each tag are comprehensively evaluated, and different weights are assigned to these factors to obtain the first-level priority sequence. , retransmission times , waiting time , set the weight coefficient of data volume to 0.4, the weight coefficient of retransmission times to 0.3, and the weight coefficient of waiting time to 0.3, then the first-level priority value Calculated by weighted sum function, the formula is:

[0144] ;

[0145] in, Indicates the By assigning different weights to each factor, the tags with higher data volume, more frequent retransmissions and longer waiting times are given higher priority, thus being given more priority in transmission scheduling. By performing corresponding calculations on all tags, the first priority sequence is obtained. ,in Represents the total number of tags. Perform the second-level priority calculation on the first priority sequence and perform exponential weighting based on the transmission delay requirement to refine the priority differences of tags. Assume that the transmission delay requirement is , an exponential weighting function is applied to the first-level priority value to obtain a new tag priority sequence , the formula is:

[0146] ;

[0147] in, is a weighting factor that determines the degree of influence of transmission delay on priority value. Through exponential weighting, the priority of tags that do not require high transmission delay can be significantly reduced, while the priority of tags that are more sensitive to transmission delay can be increased, so that the system can better adapt to application scenarios with strong delay constraints. The tag priority sequence is input into the dual threshold selector for screening. The first threshold of the dual threshold selector is set to 80% of the system processing capacity, and the second threshold is set to the mean of the current tag priority sequence. Let the first threshold be ,in Indicates the system processing capacity, the second threshold is If the priority value of a tag satisfies and , then the tag is selected into the first transmission queue, otherwise if only a single threshold is met, it is selected into the second transmission queue. Through this step, the tags are hierarchically screened to obtain a hierarchical set of tags to be scheduled. For the hierarchical set of tags to be scheduled, a two-level clustering algorithm is executed to build a multi-level transmission group structure. In the two-level clustering algorithm, the tags are roughly grouped and divided into several groups according to priority. Each group contains tags with similar priority; then each coarse group is finely divided, and the tags in each group are further subdivided into smaller subgroups. A time slot protection factor is set between groups. , to ensure that the transmission time slots between different groups do not interfere with each other, and obtain a multi-level transmission group structure. Based on the multi-level transmission group structure, a time slot resource allocation tree is constructed. The first-level nodes of the time slot resource allocation tree are used to allocate basic time slot resources, the second-level nodes are responsible for allocating staggered intervals for transmission within the group, and the third-level nodes specifically allocate the transmission time of each tag. Let the basic time slot resource be , then the allocation of the first layer nodes is expressed as:

[0148] ;

[0149] in, is the number of groups; the second-layer nodes allocate staggered intervals for transmission within the group, and the staggered interval within the group is , the allocated time slots are:

[0150] ;

[0151] in, is the index of the group; the third-layer node assigns a specific transmission time to each label to obtain a hierarchical time slot allocation scheme. Through hierarchical allocation, it is ensured that the transmission of different groups and different labels within the group can be carried out in an orderly manner, thereby reducing conflicts. A triple peak-shifting optimization is performed on the hierarchical time slot allocation scheme to reduce the probability of conflicts. The first optimization is inter-group peak-shifting optimization, which makes the transmission between different groups more staggered by adjusting the time slot difference between groups; the second optimization is intra-group peak-shifting optimization, which adjusts the transmission time of the label within the group to reduce the mutual interference of the transmission within the group; the third optimization is node peak-shifting optimization, which makes the transmission time more dispersed by further staggering the specific nodes. Through these three optimizations, a multi-dimensional peak-shifting control sequence that minimizes the probability of conflicts is generated, thereby effectively improving the success rate of data transmission. According to the multi-dimensional peak-shifting control sequence, three levels of scheduling queues are established, namely, the emergency transmission queue, the standard transmission queue, and the low-priority transmission queue. For each queue, a differentiated scheduling period is assigned according to its urgency. Let the scheduling period of the emergency transmission queue be , the scheduling period of the standard transmission queue is , the scheduling period of the low priority transmission queue is ,and Through a hierarchical scheduling approach, the transmission needs of urgent tags are prioritized while also taking into account the transmission of standard and low-priority tags, resulting in a hierarchical scheduling control sequence. Based on this hierarchical scheduling control sequence, a performance monitoring matrix is established to monitor system performance in real time. This performance monitoring matrix consists of three main monitoring points: the first monitoring point, set at the queue switch, monitors the smooth switching between queues of different priority levels; the second monitoring point, set at the time slot boundary, ensures that each time slot is used as expected and there is no time slot overlap; and the third monitoring point, set at the transmission completion point, assesses the transmission success rate and latency. Through the combined analysis of these three monitoring points, a multi-dimensional performance evaluation is generated, ensuring the effectiveness of the entire transmission scheduling process. Based on these multi-dimensional performance evaluation results, a three-level dynamic adjustment mechanism is triggered to adapt to changing conditions during system operation. The first level of dynamic adjustment involves adjusting the division of transmission groups, re-dividing tag groups to adapt to changing system loads; the second level of adjustment involves adjusting the allocation of time slots, optimizing time slot usage based on real-time performance evaluation results; and the third level of adjustment involves adjusting the peak shifting parameters, further reducing conflicts by varying the peak shifting parameters between and within groups. All adjustments use exponential backoff algorithm to control the adjustment period. Let the adjustment period be , dynamically adjust the period through exponential backoff to adapt to the actual situation of the system:

[0152] ;

[0153] in, is the initial adjustment period, is the number of adjustments. In this way, the adjustment frequency changes dynamically according to the stability of the system, thereby outputting a multi-tag concurrent transmission sequence.

[0154] The above describes the UHF RFID system data transmission method according to the embodiment of the present invention. The following describes the UHF RFID system data transmission device according to the embodiment of the present invention. Figure 2 An embodiment of the ultra-high frequency RFID system data transmission device in the embodiment of the present invention includes:

[0155] The filtering module 201 is used to perform bandpass filtering and orthogonal demodulation on the multi-tag backscattered signals received by the RFID reader to obtain a tag feature data sequence;

[0156] The calculation module 202 is used to perform multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame;

[0157] Identification module 203, used to identify the boundary of the independent tag signal frame and determine the starting point of the tag data frame;

[0158] The analysis module 204 is used to continuously sample and analyze the signal after the starting point of the tag data frame and identify tag signal conflicts, and output a conflict position sequence;

[0159] A rearrangement module 205 is configured to perform matrix interleaving transformation and position rearrangement on the data marked with the conflict position sequence to generate an anti-collision data packet;

[0160] The output module 206 is configured to construct a tag access scheduling queue based on the transmission timing and priority attributes of the anti-collision data packet, and output a multi-tag concurrent transmission timing.

[0161] Through the collaborative cooperation of the above components, 1. The use of time-space separation signal feature analysis and multi-dimensional feature mapping technology, combined with multi-level feature extraction and correlation analysis, effectively improves the recognition accuracy in multi-tag environments; 2. The innovative introduction of adaptive interleaving coding and multiple staggered transmission mechanisms enhances the system's anti-interference ability and data transmission reliability in high-density tag environments; 3. A priority-based multi-level scheduling algorithm is designed, combined with a dynamic time slot allocation strategy, to reduce tag response conflicts and reduce transmission delays; 4. Through layered staggered control and performance adaptive optimization, the system resource utilization efficiency is optimized and the data transmission throughput is improved; 5. The use of multiple anti-collision coding and intelligent retransmission control enhances the system's transmission stability in complex environments.

[0162] above Figure 2 The medium and ultra-high frequency RFID system data transmission device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The ultra-high frequency RFID system data transmission equipment in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0163] Figure 3This is a schematic diagram of the structure of an ultra-high frequency RFID system data transmission device provided by an embodiment of the present invention. The ultra-high frequency RFID system data transmission device 300 may vary significantly depending on configuration or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors), memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) storing application programs 333 or data 332. The memory 320 and storage medium 330 may be either transient or persistent storage. The program stored in the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions and operations within the ultra-high frequency RFID system data transmission device 300. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, executing the series of instructions and operations stored in the storage medium 330 on the ultra-high frequency RFID system data transmission device 300 to implement the steps of the ultra-high frequency RFID system data transmission method described above.

[0164] The UHF RFID system data transmission device 300 may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3 The illustrated structure of the UHF RFID system data transmission device does not limit the UHF RFID system data transmission device provided by the present invention and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0165] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions. When the instructions are executed on a computer, the computer executes the steps of the ultra-high frequency RFID system data transmission method.

[0166] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0167] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0168] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for transmitting data in an ultra-high frequency RFID system, characterized in that: The method comprises: Bandpass filtering and orthogonal demodulation are performed on the multi-tag backscattered signals received by the RFID reader to obtain the tag feature data sequence; Performing multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame; Performing boundary identification on the independent tag signal frame to determine the starting point of the tag data frame; Performing continuous sampling analysis and tag signal conflict identification on the signal after the starting point of the tag data frame, and outputting a conflict position sequence; Performing matrix interleaving transformation and position rearrangement on the data marked by the conflict position sequence to generate an anti-collision data packet; Based on the transmission timing and priority attributes of the anti-collision data packet, a label access scheduling queue is constructed, and a multi-label concurrent transmission timing is output; specifically, the method includes: performing a first-level priority calculation on the anti-collision data packet, assigning a weight coefficient of 0.4 to the label data volume, a weight coefficient of 0.3 to the number of retransmissions, and a weight coefficient of 0.3 to the waiting time, and obtaining a first priority sequence according to a weighted sum function; performing a second-level priority calculation on the first priority sequence, performing exponential weighting based on the transmission delay requirement, and obtaining a label priority sequence; inputting the label priority sequence into a dual-threshold selector, wherein the first threshold is set to 80% of the system processing capacity, and the second threshold is set to the priority mean, and labels that meet the dual-threshold conditions are selected into the first transmission queue, and labels that only meet the single threshold conditions are selected into the second transmission queue, and a hierarchical set of labels to be scheduled is obtained; performing a two-level clustering algorithm on the hierarchical set of labels to be scheduled, wherein the first level performs a coarse grouping of the labels, and the second level performs a fine division of each coarse group, and a time slot protection factor β=0.1 is set between groups to obtain a multi-level transmission group structure; constructing a time slot resource allocation based on the multi-level transmission group structure A tree is configured, in which the first-layer nodes allocate basic time slot resources, the second-layer nodes allocate intra-group staggered intervals, and the third-layer nodes allocate specific transmission times, to obtain a hierarchical time slot allocation scheme; a triple staggered optimization is performed on the hierarchical time slot allocation scheme: the first optimization is inter-group staggered, the second optimization is intra-group staggered, and the third optimization is node staggered, to generate a multi-dimensional staggered control sequence that minimizes the probability of conflict; a three-level scheduling queue is established according to the multi-dimensional staggered control sequence: the first level is the emergency transmission queue, the second level is the standard transmission queue, and the third level is the low-priority transmission queue, and a differentiated scheduling cycle is assigned to each queue to obtain a hierarchical scheduling control sequence; a performance monitoring point matrix is set based on the hierarchical scheduling control sequence, including: the first monitoring point is set at the queue switching point, the second monitoring point is set at the time slot boundary, and the third monitoring point is set at the transmission completion point, and a multi-dimensional performance evaluation result is obtained through a joint analysis of the three points; a three-level dynamic adjustment mechanism is triggered according to the multi-dimensional performance evaluation result: the first level adjusts the transmission group division, the second level adjusts the time slot allocation, and the third level adjusts the staggered parameters, and an exponential backoff algorithm is used to control the adjustment cycle, and output the multi-label concurrent transmission timing.

2. The UHF RFID system data transmission method according to claim 1, characterized in that: The method of performing bandpass filtering and orthogonal demodulation on the multi-tag backscatter signals received by the RFID reader to obtain a tag feature data sequence includes: Receive multi-tag backscatter signals through an RFID reader, and filter the multi-tag backscatter signals through a bandpass filter with a center frequency of 915 MHz to obtain a primary filtered signal; The primary filtered signal is sampled at four times the carrier frequency and frequency converted by a digital down-conversion circuit to obtain an intermediate frequency signal, the intermediate frequency signal is input into an orthogonal demodulator, and multiplied by the in-phase branch and the quadrature branch of the local carrier respectively to obtain an in-phase component I and a quadrature component Q; Performing band-limiting processing on the in-phase component I and the quadrature component Q through a low-pass filter to obtain a baseband I signal and a baseband Q signal, and digitally sampling the baseband I signal and the baseband Q signal through a 16-bit analog-to-digital converter to obtain a digitized I data stream and a digitized Q data stream; Performing timing recovery on the digitized I data stream and the digitized Q data stream, establishing a sampling clock sequence, and segmenting the digitized I data stream and the digitized Q data stream according to a 16-bit RN16 length based on the sampling clock sequence to obtain I data blocks and Q data blocks; The I-path data blocks and the Q-path data blocks are combined into a label feature data sequence in a plural form.

3. The UHF RFID system data transmission method according to claim 2, characterized in that: The performing multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame includes: Performing complex operations on the tag feature data sequence, calculating the square of the amplitude and the phase angle of each data point to obtain a tag signal strength sequence and a phase sequence, and sorting the tag signal strength sequence according to signal energy to obtain a valid signal point set; Extracting arrival time information for each data point in the valid signal point set, and performing weighted averaging on the time information according to the signal strength to obtain a tag response time sequence; Inputting the tag response time sequence into a feature classifier, grouping the time intervals based on a K-means clustering algorithm to obtain a tag response period set; According to the tag response period set, an A×B dimensional feature matrix is constructed, where A is the number of tags and B is the feature dimension, including amplitude, phase and time information, to obtain a feature space mapping matrix; Calculate the Pearson correlation coefficient of each row of the feature space mapping matrix with the adjacent data frame, and set the correlation threshold to 0.8 to obtain the label correlation matrix; Connected component analysis is performed based on the label correlation matrix, and data frames with correlation coefficients higher than a threshold are combined to obtain an independent label candidate set. Temporal consistency verification is performed on the independent label candidate set, and data frames that do not meet the fixed response period are eliminated, and independent label signal frames are output.

4. The UHF RFID system data transmission method according to claim 3, characterized in that: The step of identifying the boundary of the independent tag signal frame and determining the starting point of the tag data frame includes: Constructing a 16-bit cyclic shift synchronization codeword for the independent tag signal frame to generate a standard synchronization codeword with a maximum and minimum Hamming distance, and performing a full-cycle cyclic shift operation on the standard synchronization codeword to generate a set of 16 shifted versions of the synchronization codeword; Segmenting the independent tag signal frame according to the synchronization codeword length, performing normalization processing on each signal sequence to obtain a data sequence to be matched, and performing a correlation operation on the data sequence to be matched with each shifted version in the synchronization codeword set, calculating a 256-point correlation value sequence, and obtaining a correlation coefficient matrix; Performing peak detection on each column of the correlation coefficient matrix, extracting position information and amplitude information of the local maximum point, obtaining a set of candidate synchronization points, and calculating an optimal decision threshold based on the amplitude distribution of the set of candidate synchronization points to obtain a valid synchronization position sequence; A majority vote is performed on the synchronization points of the three adjacent data frames in the valid synchronization position sequence, and the position with the highest frequency of occurrence is selected to obtain the frame synchronization position. The data frame is divided and reorganized based on the frame synchronization position, and the independent label signal frame is divided into data units of fixed length according to the synchronization position, and the starting point of the label data frame is output.

5. The UHF RFID system data transmission method according to claim 4, characterized in that: The continuous sampling and analysis of the signal after the starting point of the tag data frame and tag signal conflict identification, and outputting a conflict position sequence, include: The signal after the starting point of the tag data frame is segmented into sliding windows, the window length is set to RN16 length, the sliding step is 1 code element, and a continuous signal window sequence is obtained; Performing complex envelope detection on each window data in the continuous signal window sequence, extracting the signal amplitude trajectory and phase trajectory to obtain a signal feature trajectory sequence, performing a differential operation on the signal feature trajectory sequence, calculating the amplitude change rate and phase jump amount between adjacent sampling points, and obtaining a signal change feature sequence; Input the signal change characteristic sequence into a dual-threshold detector, set the amplitude change rate threshold to 20%, and the phase jump threshold to 45 degrees, to obtain a primary collision judgment result; Performing signal vector diagram analysis on the primary collision judgment result, counting the number of cluster centers in the signal space, marking areas where the number of cluster centers is greater than 1 as potential collision points, and obtaining a candidate set of collision points; Constructing a collision feature vector based on the collision point candidate set, which includes three feature quantities: amplitude mutation degree, phase jump degree and cluster dispersion, to obtain a collision feature sequence; The collision feature sequence is input into a support vector machine classifier for collision event recognition to obtain a collision event marker sequence, and a time domain correlation analysis is performed on the collision event marker sequence, and adjacent collision event markers are merged to generate a conflict position sequence.

6. The UHF RFID system data transmission method according to claim 5, characterized in that: The performing matrix interleaving transformation and position rearrangement on the data marked by the conflict position sequence to generate an anti-collision data packet includes: The data marked by the conflict position sequence is divided into 8-bit fixed lengths to construct a data unit sequence to obtain a set of data units to be retransmitted; Filling the set of data units to be retransmitted into an M×N interleaving matrix in row-first order, where M is the data unit length and N is the currently detected collision frequency value, to obtain an initial interleaving matrix; Performing a column cyclic shift operation on the initial interleaving matrix, where the shift amount is linearly related to the column sequence number, to generate a positionally scattered interleaved data matrix to obtain a rearranged data matrix, reading the rearranged data matrix in column priority order, and grouping the read data, each group of data having a length of 4 bits, to obtain a data group to be encoded; Performing a Hamming coding operation on the data group to be encoded to generate a 3-bit check bit, and adding the check bit to the original data to obtain an encoded data sequence, and constructing a data packet header based on the encoded data sequence, which includes interleaving depth information, data packet length, and original position identifier, to obtain a data packet header; The data packet header is combined with the coded data sequence, encapsulated into a data packet according to a fixed format, and a CRC check code is added to generate a 32-bit cyclic redundancy check code, and an anti-collision data packet is output.

7. An ultra-high frequency RFID system data transmission device, characterized in that: The device is used to execute the UHF RFID system data transmission method according to any one of claims 1 to 6, comprising: The filtering module is used to perform bandpass filtering and orthogonal demodulation on the multi-tag backscattered signals received by the RFID reader to obtain a tag feature data sequence; A calculation module, configured to perform multi-dimensional feature analysis and time series correlation calculation on the tag feature data sequence to obtain an independent tag signal frame; An identification module, configured to identify the boundary of the independent tag signal frame and determine the starting point of the tag data frame; An analysis module, configured to perform continuous sampling analysis on the signal after the starting point of the tag data frame and identify tag signal conflicts, and output a conflict position sequence; a rearrangement module, configured to perform matrix interleaving transformation and position rearrangement on the data marked by the conflict position sequence to generate an anti-collision data packet; The output module is used to construct a label access scheduling queue based on the transmission timing and priority attributes of the anti-collision data packet, and output the multi-label concurrent transmission timing.

8. An ultra-high frequency RFID system data transmission device, characterized in that: The UHF RFID system data transmission device includes: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory to enable the UHF RFID system data transmission device to execute the UHF RFID system data transmission method according to any one of claims 1 to 6.

9. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the UHF RFID system data transmission method according to any one of claims 1 to 6 is implemented.

Images