Data processing method, device and equipment based on RFID reader middleware

Abstract

The application relates to a data processing method, device and equipment based on RFID reader middleware. The method comprises the following steps: acquiring tag radio frequency characteristic data collected by an RFID reader in a management and control area and target space motion sensing data synchronously collected by a millimeter wave radar, establishing a three-dimensional data correlation model after protocol adaptation and millisecond-level clock time sequence alignment processing of the two types of heterogeneous data, filtering misread data of the RFID data by adopting a double-factor verification mechanism combining radio frequency characteristics and space motion parameters based on the model, continuously fitting and analyzing the filtered radar motion trajectory, combining RFID tag identification trigger time sequence to determine a passing state type, uploading effective data to a background management and control platform after edge-side deduplication error correction and structured preprocessing of the effective data. The method can improve the accuracy and system real-time performance of the in-out management and control.

Description

Technical Field

[0001] This application relates to the field of radio frequency identification and multi-sensor fusion technology, and in particular to a data processing method, apparatus and device based on RFID reader middleware. Background Technology

[0002] With the widespread application of IoT technology in smart manufacturing, intelligent warehousing, and logistics management, Radio Frequency Identification (RFID) technology has become a core means of automatic identification and tracking management of materials. Current mainstream RFID systems typically deploy reader middleware, which acts as a data bridge between reader devices and upper-layer business systems, undertaking basic functions such as data acquisition, protocol parsing, and forwarding.

[0003] However, in typical application scenarios such as access control, existing RFID reader middleware generally suffers from the following technical defects.

[0004] First, the lack of multi-source data fusion and verification capabilities results in low identification accuracy. Traditional RFID middleware only supports a single radio frequency data acquisition channel and cannot link with spatial sensing sensors such as millimeter-wave radar for cross-verification. Due to the physical characteristics of RFID signal coverage generalization, cross-reading, and misreading, static tags inside and outside the channel and tags of nearby goods are easily mis-collected, resulting in false entry and exit records. Under the single radio frequency data dimension, the system struggles to accurately determine the movement status, spatial location, and direction of entry and exit of the controlled target, and is prone to misjudgment in complex passage scenarios such as lingering, turning back, and parallel passage. At the same time, radio frequency signals are susceptible to interference from factors such as metallic environments and antenna crosstalk, further reducing data reliability in complex electromagnetic environments such as industrial and warehousing facilities. Therefore, how to effectively fuse and verify RFID radio frequency data with millimeter-wave radar spatial sensing data has become the key to improving the accuracy of entry and exit control identification.

[0005] Second, device access security mechanisms are weak. Traditional middleware relies solely on IP addresses, port numbers, or simple passwords for reader access authentication, lacking strong verification based on the uniqueness of device hardware. Attackers can set up virtual readers or forge terminal identities to access the system and steal tag data. Most solutions only perform static verification once during the initial device registration, lacking a continuous dynamic authentication mechanism during online operation. If a device is hijacked or replaced, the system cannot promptly identify it. The lack of a unified registration verification standard for readers from multiple vendors necessitates the development of independent authentication logic for different devices on the gateway, resulting in high system coupling and poor scalability. Therefore, a secure access mechanism supporting device hardware feature binding and continuous dynamic authentication is urgently needed to build a trusted closed-loop management system from access to online operation.

[0006] Third, the data processing efficiency is insufficient in batch refresh scenarios for active RFID tags. In traditional single-reader serial processing schemes, wake-up broadcast, tag reporting, image transmission, screen refresh command, and sleep command are executed sequentially, resulting in significant time consumption for large-scale tag refreshes. When large-size dot matrix image data is transmitted, the large data volume transmission can block the wake-up command channel, causing dormant tags in the area to fail to be activated in time, leading to a high rate of missed refreshes. Even with a simple parallel scheme using dual readers, which only achieves physical separation of the two radio frequency paths at the hardware level, the data flow still lacks decoupling scheduling. The wake-up reporting data flow and the large image data flow compete for each other in terms of host computer buffer and channel resources, and the conflict of multiple tag concurrent return packets can easily cause screen distortion and frequent retransmissions. All tag tasks are queued in a unified queue, resulting in unreasonable channel time slot allocation and low channel utilization. Therefore, how to separate and schedule short-frame control data and large-capacity image data to avoid mutual blocking is an important direction for improving the efficiency of batch tag refreshes.

[0007] Fourth, the original data has high redundancy, resulting in a heavy system load. In intensive read / write scenarios, tags are scanned repeatedly at high frequencies. Coupled with environmental interference and signal fluctuations, this generates a large amount of duplicate and misread dirty data. Existing middleware only uses a fixed time window for simple deduplication, lacking layered noise reduction and precise filtering capabilities. The massive upload of invalid data leads to network congestion, wastes server computing power and storage resources, and significantly degrades system real-time performance. Therefore, middleware with layered noise reduction and precise filtering capabilities is urgently needed to reduce the amount of redundant data uploaded and ensure system real-time performance. Summary of the Invention

[0008] Therefore, it is necessary to provide a data processing method, apparatus, and device based on RFID reader middleware that can improve identification accuracy, enhance device access security, improve data processing efficiency, and reduce system load, in order to address the above-mentioned technical problems.

[0009] A data processing method based on RFID reader middleware, the method comprising: The system acquires RFID tag radio frequency feature data collected by RFID readers within the controlled area and target spatial motion sensing data synchronously collected by millimeter-wave radar. The RFID tag radio frequency feature data includes tag identification, signal phase, and RSSI signal strength. The target spatial motion sensing data includes spatial coordinates, distance, and movement speed. Heterogeneous data protocol adaptation and millisecond-level clock timing alignment are performed on the tag radio frequency feature data and the target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. Based on the aforementioned three-dimensional data association model, a two-factor verification mechanism combining radio frequency features and spatial motion parameters is adopted to filter RFID data for misreading based on preset spatial thresholds and motion thresholds in the passage area. The radar motion trajectory after the misread data filtering process is subjected to continuous trajectory fitting analysis, and the passage status type of the controlled target is determined by combining the RFID tag identification trigger timing. On the edge side, the valid data after determining the passage status is deduplicated, corrected for errors, and pre-processed into a structured format. The pre-processed access control data is then uploaded to the backend control platform.

[0010] In one embodiment, heterogeneous data protocol adaptation and millisecond-level clock timing alignment are performed on the tag radio frequency feature data and the target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. This includes: distinguishing between RFID private protocols and general standard protocols through a message feature code protocol automatic identification engine; uniformly converting the underlying raw hexadecimal radio frequency data into standard event data through a bidirectional standardized message conversion mechanism; dynamically adjusting the message aggregation window size according to business latency requirements; and clock-aligning and matching the converted RFID standard data and radar sensing data according to millisecond-level timestamps to form a three-dimensional data association structure of tag identification, spatiotemporal location and motion state.

[0011] In one embodiment, based on the three-dimensional data association model, a two-factor verification mechanism combining radio frequency characteristics and spatial motion parameters is adopted to filter RFID data for misreading according to preset passage area spatial thresholds and motion thresholds. This includes: determining the real-time spatial position of the controlled target relative to the passage area using distance and spatial coordinate data collected by millimeter-wave radar; filtering out RFID data corresponding to controlled targets whose spatial positions are outside the preset passage area spatial thresholds as out-of-area cross-read tag data; and filtering out RFID data corresponding to controlled targets whose moving speed is continuously lower than the preset motion threshold as static stationary tag data within the area.

[0012] In one embodiment, the radar motion trajectory after the misread data filtering process is subjected to continuous trajectory fitting analysis, and the passage status type of the controlled target is determined by combining the RFID tag identification trigger timing. This includes: performing trajectory fitting on the continuous spatial coordinate data collected by the millimeter-wave radar to obtain the motion trajectory curve of the controlled target, and combining the timestamp of the first identification and the timestamp of the last identification of the RFID tag of the controlled target to determine the passage status of the controlled target as an entry state, exit state, turn-back state, lingering state, or traversing state.

[0013] In one embodiment, before acquiring the tag radio frequency feature data collected by the RFID reader within the controlled area, the method further includes: performing multi-factor composite registration verification on the RFID reader; the multi-factor composite registration verification includes network layer access pre-verification, static verification of device whitelist and service permissions, two-way identity authentication of device hardware fingerprint, and dynamic heartbeat continuous online verification based on session key; if any verification fails, registration is rejected and connection is blocked.

[0014] In one embodiment, the device hardware fingerprint in the two-way authentication of device hardware fingerprint is generated by concatenating the device SN, main control chip ID, radio frequency module hardware ID, network port MAC address, and factory calibration radio frequency parameters, and then hashing them using SM3. The two-way authentication of device hardware fingerprint includes the RFID reader uploading the device SN and the device's local random number R1 when initiating registration. After the middleware verifies the existence of the SN, it generates a platform random number R2 and encrypts and sends it using the public key corresponding to the RFID reader. The RFID reader decrypts the platform random number R2 with its local fixed private key, concatenates the device's local random number R1, the platform random number R2, and the device hardware fingerprint, generates a digest using SM3, and signs and sends it back. The middleware uses the public key to verify the signature and verify the consistency of the hardware fingerprint. If both the signature and the fingerprint match, it is determined to be a legitimate device and a session key SM4 is allocated.

[0015] In one embodiment, the dynamic heartbeat continuous online verification based on the session key includes initiating a lightweight heartbeat verification every preset period after the device is successfully registered and a long connection is established. The heartbeat packet carries the device fingerprint and temporary session signature that are updated in real time. The middleware verifies the legality of the session key SM4 and the consistency of the device fingerprint. If the verification fails for a preset number of consecutive times, it is determined that the device has been hijacked or replaced and the connection is automatically disconnected. The session key SM4 is automatically rotated at preset periods.

[0016] A data processing device based on RFID reader middleware, the device comprising: The RFID data acquisition module is used to collect the tag identification, signal phase, and RSSI signal strength of RFID tags within the controlled area. The radar sensing module is used to synchronously collect the spatial coordinates, distance, and movement speed of targets within the control area; The timing alignment unit is used to perform heterogeneous data protocol adaptation and millisecond-level clock timing alignment processing on the collected tag radio frequency feature data and target spatial motion sensing data, and to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. The two-factor verification unit is used to perform misread data filtering on RFID data based on the three-dimensional data association model and a two-factor verification mechanism that combines radio frequency features and spatial motion parameters, according to preset passage area spatial thresholds and motion thresholds. The trajectory analysis and discrimination unit is used to perform continuous trajectory fitting analysis on the radar motion trajectory after the misread data filtering process, and to determine the passage status type of the controlled target by combining the RFID tag identification trigger timing. The edge preprocessing unit is used to perform deduplication, error correction, and structured preprocessing on the valid data after determining the passage status at the edge, and upload the preprocessed access control data to the backend control platform.

[0017] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps: The system acquires RFID tag radio frequency feature data collected by RFID readers within the controlled area and target spatial motion sensing data synchronously collected by millimeter-wave radar. The RFID tag radio frequency feature data includes tag identification, signal phase, and RSSI signal strength. The target spatial motion sensing data includes spatial coordinates, distance, and movement speed. Heterogeneous data protocol adaptation and millisecond-level clock timing alignment are performed on the tag radio frequency feature data and the target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. Based on the aforementioned three-dimensional data association model, a two-factor verification mechanism combining radio frequency features and spatial motion parameters is adopted to filter RFID data for misreading based on preset spatial thresholds and motion thresholds in the passage area. The radar motion trajectory after the misread data filtering process is subjected to continuous trajectory fitting analysis, and the passage status type of the controlled target is determined by combining the RFID tag identification trigger timing. On the edge side, the valid data after determining the passage status is deduplicated, corrected for errors, and pre-processed into a structured format. The pre-processed access control data is then uploaded to the backend control platform.

[0018] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor: The system acquires RFID tag radio frequency feature data collected by RFID readers within the controlled area and target spatial motion sensing data synchronously collected by millimeter-wave radar. The RFID tag radio frequency feature data includes tag identification, signal phase, and RSSI signal strength. The target spatial motion sensing data includes spatial coordinates, distance, and movement speed. Heterogeneous data protocol adaptation and millisecond-level clock timing alignment are performed on the tag radio frequency feature data and the target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. Based on the aforementioned three-dimensional data association model, a two-factor verification mechanism combining radio frequency features and spatial motion parameters is adopted to filter RFID data for misreading based on preset spatial thresholds and motion thresholds in the passage area. The radar motion trajectory after the misread data filtering process is subjected to continuous trajectory fitting analysis, and the passage status type of the controlled target is determined by combining the RFID tag identification trigger timing. On the edge side, the valid data after determining the passage status is deduplicated, corrected for errors, and pre-processed into a structured format. The pre-processed access control data is then uploaded to the backend control platform.

[0019] The aforementioned data processing method, apparatus, and equipment based on RFID reader middleware simultaneously acquire RFID radio frequency feature data and millimeter-wave radar spatial motion sensing data at the edge. After protocol adaptation and millisecond-level clock timing alignment of the two types of heterogeneous data, a three-dimensional data association model is established, enabling deep integration of RFID identification capabilities and radar spatial sensing capabilities under a unified data model. Based on this, a two-factor verification mechanism combining radio frequency features and spatial motion parameters is used to filter misread data, and continuous fitting analysis of radar motion trajectories is performed to determine the passage status. This effectively eliminates the inherent problems of cross-reading, misreading, and direction misjudgment of single RFID sensors, improving the recognition accuracy to a reliable level of multi-dimensional cross-verification. Simultaneously, a multi-factor composite registration verification mechanism achieves full lifecycle security management of the device from access to online operation; a dual-channel pipeline parallel processing mechanism significantly improves the batch refresh efficiency of active tags and reduces tag power consumption; and edge-side data cleaning and structured preprocessing effectively reduce the load on the backend system. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating a data processing method based on RFID reader middleware in one embodiment; Figure 2 This is a flowchart illustrating the steps of heterogeneous data protocol adaptation and 3D data association model establishment in one embodiment. Figure 3 This is a flowchart illustrating the steps for filtering misread data in one embodiment; Figure 4 This is a structural block diagram of a data processing device based on RFID reader middleware in one embodiment; Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0022] In one embodiment, such as Figure 1 As shown, a data processing method based on RFID reader middleware is provided, including the following steps: Step 102: Obtain the tag radio frequency feature data collected by the RFID reader within the controlled area and the target spatial motion perception data collected synchronously by the millimeter-wave radar.

[0023] The tag radio frequency characteristic data includes tag identification, signal phase and RSSI signal strength, and the target spatial motion sensing data includes spatial coordinates, distance and movement speed.

[0024] RSSI, or Received Signal Strength Indication, is a physical quantity that measures the power of the tag's reflected signal received by the RFID reader, measured in dBm. Its value reflects the relative distance between the tag and the reader's antenna and the signal propagation environment. Signal phase refers to the phase offset of the RFID carrier signal during its propagation path, which can be used to help determine minute displacement changes of the tag.

[0025] Specifically, the RFID reader middleware establishes a data channel with UHF RFID readers within the controlled area through the RFID data acquisition module, continuously acquiring tag scanning events reported by each reader. Each tag scanning event includes the tag's unique identifier (EPC code), the signal phase value at the time of the scan, and the RSSI value. Simultaneously, the middleware establishes a data channel with millimeter-wave radar equipment deployed within the same controlled area through the radar sensing module, synchronously acquiring target point cloud data output by the radar at a fixed frequency. This includes the three-dimensional spatial coordinates of each detected target in the controlled area coordinate system, the radial distance between the target and the radar, and the target's instantaneous speed.

[0026] Step 104: Perform heterogeneous data protocol adaptation and millisecond-level clock timing alignment processing on the tag radio frequency feature data and target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters.

[0027] Heterogeneous data protocol adaptation refers to the process of converting the original hexadecimal messages output by RFID readers under proprietary protocols or standard protocols of various manufacturers, and the sensing data messages output by millimeter-wave radar, into a unified standard event data format. Millisecond-level clock timing alignment refers to aligning the timestamps of RFID data and radar data to the same millisecond-level precision time axis, using the middleware's local high-precision clock as a reference, thus eliminating timing misalignments caused by sampling clock deviations and transmission delays between the two types of sensors.

[0028] Specifically, the middleware incorporates an automatic message signature protocol identification engine. This engine automatically determines the protocol type used by the RFID reader by matching the header signature field of the data frame reported by the reader, such as the ISO18000-6C standard protocol or a vendor's proprietary protocol, without requiring manual configuration. After identifying the protocol type, the middleware calls the corresponding bidirectional standardized message conversion mechanism to uniformly convert the underlying raw hexadecimal RF data into standard JSON event data. The standard event data includes tag EPC fields, reader identifier fields, millisecond-precision timestamp fields, event type fields, protocol type fields, and custom data fields. Simultaneously, the middleware receives spatial coordinates, distance, and velocity data reported by the millimeter-wave radar through a standard interface. For both types of converted standard data, the middleware performs clock matching with millisecond-level timestamps as the alignment dimension, associating and binding RFID tag identification events within the same moment or adjacent very short time windows with the spatial location of the radar target, forming a unified three-dimensional data association structure with tag identifier as the index, spatial location as the attribute, and motion state as the parameter.

[0029] This embodiment solves the technical barrier of protocol incompatibility between heterogeneous RFID devices and radar devices from multiple manufacturers by automatically identifying protocols and performing bidirectional standardized conversion, avoiding the cumbersome work of developing separate adaptation programs for each device; by aligning millisecond-level clock timing, it resolves the fundamental contradiction of asynchronous sampling in the time dimension between the two types of sensors, enabling radio frequency identification information and spatial motion information to be associated under a unified time reference, providing a reliable data foundation for subsequent two-factor cross-validation.

[0030] Step 106: Based on the three-dimensional data association model, a two-factor verification mechanism combining radio frequency features and spatial motion parameters is adopted to filter RFID data for misreading based on preset spatial thresholds and motion thresholds in the passage area.

[0031] The two-factor verification mechanism refers to using the RFID radio frequency dimension and the radar spatial dimension as two independent and complementary verification factors to determine the credibility of each RFID tag scan record. The passage area spatial threshold refers to the geometric boundary range of the entrance channel to the controlled area in the spatial coordinate system, used to define which spatial locations belong to the valid passage area. The motion threshold is the minimum moving speed threshold used to distinguish whether the controlled target is in motion or stationary.

[0032] Specifically, the middleware uses distance and spatial coordinate data collected by millimeter-wave radar to determine the real-time spatial position of the controlled target relative to the passage area. RFID data corresponding to controlled targets whose spatial positions are outside a preset passage area spatial threshold are filtered out as out-of-area cross-read tag data. RFID data corresponding to controlled targets whose moving speed is consistently below a preset motion threshold are filtered out as static, stationary tag data within the area. Furthermore, in multi-target parallel passage scenarios, when the radar detects multiple targets with similar spatial positions simultaneously, the middleware combines the RSSI signal strength gradient changes and signal phase change trends of each target to distinguish the tag signal corresponding to each target, eliminating cross-matching errors caused by radio frequency signal reflection or multipath effects.

[0033] This embodiment overcomes the technical limitation of conventional single RFID sensors, which can only rely on signal strength for simple judgment, by using a two-factor cross-verification of radio frequency and spatial factors. In traditional solutions, when tags outside the area are misread due to radio frequency signal reflection or antenna sidelobe effects, the system lacks effective spatial dimension verification methods; when stationary tags within the area generate false entry and exit records due to continuous scanning, the system cannot exclude them through the speed dimension. This solution uses radar spatial position and movement speed as independent verification dimensions, fundamentally making up for the inherent shortcomings of RFID in spatial perception capabilities and achieving accurate filtering of misread data.

[0034] Step 108: Perform continuous trajectory fitting analysis on the radar motion trajectory after the misread data filtering process, and determine the passage status type of the controlled target by combining the RFID tag identification trigger timing.

[0035] Continuous trajectory fitting analysis refers to using a kinematic model to fit the target's spatial coordinate sequence output by millimeter-wave radar over multiple consecutive sampling periods to obtain the target's complete trajectory within the controlled area. RFID tag identification trigger timing refers to the identification time window formed by the moment the tag is first identified by any RFID reader within the area and the moment it is last identified.

[0036] The middleware extracts the continuous spatial coordinate sequence of the valid radar targets retained after filtering in step 106 within a preset time period, and generates a smooth motion trajectory curve for the target using a polynomial fitting method. Simultaneously, it queries the first and last identification timestamps of the target's corresponding tag in the 3D data association model. Based on the positional relationship between the start and end points of the motion trajectory curve relative to the entrance boundary of the passage area, the trajectory direction, and the identification time window, it comprehensively determines the passage status of the controlled target as one of the following: entering, exiting, turning back, remaining, or crossing. Specifically, entering indicates that the target moves from outside the area to inside; exiting indicates that the target moves from inside the area to outside; turning back indicates that the target enters the area but does not complete its passage and returns in the opposite direction; remaining indicates that the target stays continuously within the area; and crossing indicates that the target crosses the area but does not enter or exit through a regular passage.

[0037] This embodiment solves the misjudgment problem caused by traditional solutions that rely solely on the presence or absence of RFID signals to determine the direction of entry and exit by combining trajectory fitting and timing analysis. The continuous spatial trajectory provided by the radar contains information on the entire process of the target's movement. Combined with the precise timing of the first and last identifications of the RFID tag, it can accurately distinguish complex passage behaviors such as turning back and lingering, which cannot be identified by traditional solutions.

[0038] Step 110: On the edge side, the valid data after determining the passage status is deduplicated, corrected for errors, and pre-processed into a structured format, and the pre-processed access control data is uploaded to the backend control platform.

[0039] In this context, "edge processing" refers to data processing being performed on a local computing node located near the RFID reader middleware, close to the data acquisition source, rather than uploading all raw data to a remote server or cloud platform for further processing. Deduplication eliminates redundant records caused by the same tag being repeatedly scanned by multiple reader antennas within a short period. Error correction detects and corrects abnormal data caused by instantaneous signal fluctuations. Structured preprocessing encapsulates the processed data according to a predefined access control data format, generating structured records containing tag identification, access status, access time, access direction, and associated spatial trajectory summaries.

[0040] The middleware's edge preprocessing unit completes all data cleaning and processing locally. For compliant entry and exit events determined to be in an entry or exit state, a standard entry / exit record is generated; for turnaround, lingering, and traversing states, corresponding abnormal event records are generated along with trajectory summary information. The structured and encapsulated data is uploaded to the backend management platform through a standard interface. The backend platform is only responsible for persistent data storage, query display, and business alarm push notifications, without bearing the computational load of real-time data cleaning and status determination.

[0041] This embodiment deduplicatizes, corrects errors, determines status, and encapsulates data at the edge, ensuring that all data uploaded to the backend is high-value and valid data that has undergone step-by-step cleaning and judgment. Massive amounts of raw, redundant, and misread data are filtered at the edge, significantly reducing network bandwidth consumption and the storage and computing pressure on the backend platform. Simultaneously, millisecond-level local processing at the edge ensures the stringent real-time requirements of access control scenarios, eliminating backend communication latency and supporting complex passage scenarios involving high-speed movement and multiple parallel targets.

[0042] The aforementioned data processing method based on RFID reader middleware simultaneously acquires RFID radio frequency feature data and millimeter-wave radar spatial motion sensing data at the edge, constructing a three-dimensional correlation structure between the two types of heterogeneous data under a unified time reference and data model. A dual-factor cross-validation mechanism of radio frequency and spatial data is used for misread filtering. Furthermore, a comprehensive judgment of passage status is made by combining radar continuous trajectory fitting and RFID identification timing. Finally, the entire dataset is cleaned and uploaded in a structured manner at the edge. This method solves the technical problems of frequent misreads and serious directional misjudgments caused by the lack of spatial sensing capabilities in traditional single RFID solutions, achieving precise access control through multi-dimensional cross-validation.

[0043] In one embodiment, such as Figure 2 As shown, heterogeneous data protocol adaptation and millisecond-level clock timing alignment are performed on tag RFID data and target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information, and motion state parameters. Specifically, this includes the following sub-steps: Step 202: The message feature code protocol automatic identification engine distinguishes between RFID private protocols and general standard protocols, and converts the underlying raw hexadecimal radio frequency data into standard event data through a two-way standardized message conversion mechanism.

[0044] The bidirectional standardized message conversion mechanism includes uplink conversion and downlink conversion. Uplink conversion parses the raw hexadecimal radio frequency data frames reported by the RFID reader into standard JSON event data containing tag EPC, reader identifier, timestamp, event type, protocol type, and carried data. Downlink conversion automatically translates control commands issued by the upper-layer business system into private execution commands corresponding to the target reader, realizing unified management and control of heterogeneous RFID devices.

[0045] Step 204: Dynamically adjust the message aggregation window size according to business latency requirements, and perform clock alignment and matching between the converted RFID standard data and radar sensing data according to millisecond-level timestamps to form a three-dimensional data association structure of tag identification, spatiotemporal location and motion status.

[0046] The message aggregation window refers to the maximum time window allowed to wait within a single data processing cycle. It is used to aggregate multiple tag scan events arriving within the window for batch processing. For scenarios with high real-time requirements, such as access control, the aggregation window is set to a smaller value; for scenarios with high data integrity requirements but low latency sensitivity, such as batch inventory, the aggregation window is set to a larger value to improve data aggregation and processing efficiency.

[0047] In this embodiment, through automatic protocol identification and bidirectional standardized conversion, universal access and unified management of heterogeneous RFID readers and radar devices from multiple vendors are achieved, eliminating the high coupling pain point of traditional solutions that require developing separate adaptation programs for each device. Through a dynamic aggregation window mechanism, the same data processing pipeline can adapt to the latency differences required by different business scenarios such as access control and batch inventory.

[0048] In one embodiment, such as Figure 3 As shown, based on a three-dimensional data association model, a two-factor verification mechanism combining radio frequency characteristics and spatial motion parameters is adopted. Misreading data is filtered from RFID data according to preset spatial and motion thresholds for the passage area. Specifically, the process includes the following sub-steps: Step 302: Determine the real-time spatial position of the controlled target relative to the passage area using the distance and spatial coordinate data collected by the millimeter-wave radar, and filter out the RFID data corresponding to the controlled target whose spatial position is outside the preset passage area spatial threshold as the data of the tag to be read outside the area.

[0049] Cross-reading outside the designated area refers to the phenomenon where tags in adjacent areas outside the entrance channel of a controlled area are accidentally read due to RFID radio frequency signals covering them. Because RFID radio frequency signals exhibit reflection, diffraction, and sidelobe effects during spatial propagation, the effective reading range of the reader often exceeds the physical boundary of the preset controlled channel, causing static tags outside the channel to be misidentified as passage targets. This embodiment utilizes precise spatial coordinate information provided by radar to perform spatial location verification on each RFID tag scanning event, directly filtering out tag data that does not belong to the spatial range of the passage area, thus eliminating cross-reading interference from a spatial perspective.

[0050] Step 304: RFID data corresponding to control targets whose moving speed is consistently below a preset motion threshold are filtered out as static lingering tag data within the area.

[0051] Among them, static loitering tags within the controlled area refer to tags located within the controlled area but in a stationary state, such as inventory tags placed near passageways or equipment tags fixed to walls. Because these tags are fixed in position within the area, they are continuously scanned by readers and cannot be distinguished from actual passing targets based solely on spatial location. This embodiment utilizes the movement speed information provided by radar to effectively distinguish between moving targets and stationary loitering targets by determining whether the target's speed remains below a motion threshold over multiple consecutive sampling periods.

[0052] This embodiment uses dual filtering logic of spatial position verification and motion speed verification to accurately filter two typical misread scenarios: cross-reading outside the area and lingering within the area. This solves the problem that traditional solutions cannot distinguish between these scenarios by relying solely on signal strength thresholds or fixed time windows.

[0053] In one embodiment, continuous trajectory fitting analysis is performed on the radar motion trajectory after the misread data filtering process, and the passage status type of the controlled target is determined by combining the RFID tag identification trigger timing. This includes: performing trajectory fitting on the continuous spatial coordinate data collected by the millimeter-wave radar to obtain the motion trajectory curve of the controlled target, and combining the timestamp of the first identification and the timestamp of the last identification of the RFID tag of the controlled target to determine the passage status of the controlled target as an entry state, exit state, turnaround state, lingering state, or traversing state.

[0054] The determination logic for the five passage states is as follows: When the starting point of the movement trajectory curve is outside the passage area and the ending point is inside the passage area, and the first identification timestamp of the tag matches the time window of the trajectory entering the area, it is determined to be in the entry state; when the starting point of the movement trajectory curve is inside the passage area and the ending point is outside the passage area, and the last identification timestamp of the tag matches the time window of the trajectory leaving the area, it is determined to be in the exit state; when the movement trajectory curve shows that the target's movement direction reverses after entering the area and eventually leaves from the entry side, it is determined to be in the return state; when the target's stay time in the area exceeds the preset time threshold and there is no trajectory change that crosses the area boundary, it is determined to be in the lingering state; when the movement trajectory curve crosses the passage area but the starting point and ending point of the trajectory are not within the boundary of the regular passage, it is determined to be in the traversing state.

[0055] This embodiment achieves accurate classification of five complex passage states by jointly determining the continuous radar trajectory and the discrete RFID timing. In particular, it provides a reliable determination method for states that cannot be identified by traditional single RFID solutions, such as turning back and lingering, which significantly improves the precision of access control.

[0056] In one embodiment, before acquiring the RFID tag radio frequency feature data collected by the RFID reader within the controlled area, a multi-factor composite registration verification is performed on the RFID reader. The multi-factor composite registration verification includes, in sequence, network layer access pre-verification, static verification of device whitelist and service permissions, two-way identity authentication of device hardware fingerprint, and continuous online verification of dynamic heartbeat based on session key. If any verification fails, registration is rejected and connection is blocked.

[0057] Specifically, the complete process of multi-factor composite registration verification is as follows: The first verification step is the network layer access pre-verification. When the RFID reader initiates a TCP connection or serial port connection request to the middleware, the middleware obtains basic information such as the device's MAC address, SN serial number, and model through a private protocol, sets the working mode parameters, verifies the allowed access port numbers, and verifies the binding relationship between the device number and the MAC address to prevent impersonating devices from accessing through legitimate ports.

[0058] The second layer of verification involves static checks of the device whitelist and business permissions. After the reader uploads its own serial number (SN), the middleware searches the pre-configured device whitelist database. If the SN does not exist in the whitelist, registration is rejected and an alarm is recorded. If the SN exists but the corresponding device is in a disabled, expired, or scrapped state, registration also fails and an alarm log is recorded. This whitelist supports batch import and full lifecycle management of devices, including operations such as adding, disabling, deregistering, and reclaiming devices.

[0059] The third verification step is two-way authentication using the device's hardware fingerprint. Legitimate RFID devices have their hardware fingerprint and private key burned into the reader's internal secure storage area before leaving the factory; the private key cannot be read or exported externally. The device's hardware fingerprint is generated by concatenating the device's serial number (SN), main control chip ID, RF module hardware ID, network port MAC address, and factory-calibrated RF parameters, and then using the national cryptographic SM3 hash algorithm to generate a fixed-length fingerprint value.

[0060] The specific interaction process for two-way authentication of device hardware fingerprint is as follows: When the reader initiates registration, it uploads the device SN and a locally generated random number R1; after verifying the existence of the SN, the middleware generates a platform-side random number R2, encrypts R2 using the reader's corresponding SM2 public key, and sends it to the reader; the reader decrypts R2 using its locally stored SM2 private key, concatenates R1 and R2 with the locally calculated device hardware fingerprint, generates a digest using SM3 hashing, digitally signs the digest using its own SM2 private key, and sends it back to the middleware; the middleware uses the reader's SM2 public key to verify the received signature, and simultaneously compares the hardware fingerprint carried in the digest with the device hardware fingerprint pre-stored in the database; after successful verification and a complete match of the hardware fingerprint, the middleware determines that the reader is a legitimate device, completes the registration, and allocates an SM4 symmetric session key for this session.

[0061] Among them, SM2 is an asymmetric encryption algorithm standard based on elliptic curve cryptography, used for digital signatures and public-key encryption. SM3 is a cryptographic hash algorithm standard used to generate fixed-length message digests, possessing collision resistance and irreversibility properties. SM4 is a block symmetric encryption algorithm standard used for encrypted transmission of session data.

[0062] The fourth verification step is a dynamic heartbeat-based continuous online verification based on the session key. After successful device registration and the establishment of a long connection, a lightweight heartbeat verification is initiated every preset period, such as 60 seconds. The heartbeat packet carries a real-time updated device fingerprint and a temporary session signature. The middleware verifies the validity of the SM4 key for this session and whether the device fingerprint is consistent with that at the time of registration, confirming that the device has not been replaced or tampered with during operation. If the heartbeat verification fails for a preset number of consecutive times, such as 3 times, the middleware determines that the device has been hijacked or replaced, automatically disconnects the TCP or serial port connection, marks the device as abnormal in the system, and pushes a security alarm to the background management platform. At the same time, the session key SM4 is automatically rotated at preset periods to prevent brute-force attacks on session packets caused by long-term fixed keys.

[0063] This embodiment constructs a comprehensive security protection chain from device access to online operation through four serial verification checkpoints. Compared with traditional one-time static verification schemes, the technical breakthroughs of this scheme are as follows: First, it expands the verification factor from a single dimension of the network layer to four dimensions: network layer, ledger layer, hardware layer, and encryption layer, achieving multi-layered defense in depth; second, the hardware fingerprint integrates underlying hardware features such as the main control chip ID and the factory calibration parameters of the RF module, possessing physical uniqueness that cannot be cloned, making it impossible for attackers to forge legitimate hardware fingerprints even if they obtain the device's SN and MAC address; third, the continuous heartbeat dynamic authentication and the SM4 key timed rotation mechanism ensure that the device remains under uninterrupted security monitoring after registration, solving the security blind spot in traditional schemes where devices can be replaced and hijacked after registration and going online.

[0064] In one embodiment, the acquisition of tag radio frequency feature data collected by RFID readers within the controlled area is also achieved through the following parallel processing mechanism: setting up two independent data processing pipelines, namely the wake-up data stream channel and the screen refresh data stream channel, and the middleware has a built-in dual-queue isolated business scheduling engine.

[0065] The wake-up data stream channel processes only short frame control data, including wake-up broadcast frames, tag ID reporting frames, and battery status frames. After data parsing, it pushes the online tag index to the shared online tag mapping pool in real time. The screen refresh data stream channel independently caches image dot matrix data, partial refresh bitmaps, CRC check modules, and screen update instructions, and transmits image data in segments and packets. The dual channels share a tag online status mapping queue, enabling the pipelined parallel execution of the two tasks: wake-up tag discovery and screen refresh synchronous image transmission. The two data streams do not block each other or preempt cache resources.

[0066] In terms of data coordination and matching, the middleware's main control chip coordinates the processing of uplink and downlink command response data, sends tag wake-up commands to the wake-up module to execute the wake-up operation, and writes the successfully woken-up tag identifiers into the shared online tag dynamic mapping index pool in real time. The screen refresh module reads the online tag list in the mapping pool, filters out offline tags, and only allocates image delivery time slots to online tags that have been confirmed to be woken up. The middleware also sets up bidirectional data synchronization verification logic: the tag receives all image fragments and caches them, writes them into the frame buffer after completing the global CRC check and executes the screen refresh, and sends back a successful response to the mapping pool after the refresh is completed. The mapping pool automatically marks the tag as refreshed. If the tag verification is abnormal, only the sequence number of the missing fragment is returned, and the screen refresh module resends the missing fragment at the designated time. For tags that do not send back a response within a preset time, the mapping pool marks them as refresh failures and sends feedback to the wake-up channel, triggering a retry operation for wake-up and refresh. The mapping pool adopts an incremental update mechanism, only synchronizing the changes in the newly added tag identifiers and offline tag identifiers, reducing the amount of interactive data between the two channels and improving coordination efficiency.

[0067] In actual testing of a batch refresh scenario of 100 tags in a warehouse environment, the overall refresh time of the above-mentioned dual-channel pipeline parallel processing solution was reduced by more than 55% compared with the traditional single-reader serial processing solution, the average standby power consumption of the tags was reduced by 40%, and the dual-channel data collaborative matching reduced the tag wake-up time by more than 50%. The test results are shown in Table 1 below: Table 1 Test Results

[0068] This embodiment solves the technical bottleneck of blocking control command channels during large-volume image transmission in traditional serial processing schemes by using separate independent pipelines for wake-up and screen-scrolling data streams and an incremental synchronization and coordination mechanism with a shared mapping pool. It also overcomes the problems of buffer contention and packet conflicts caused by the lack of decoupling in simple dual-reader parallel schemes. The dual-channel parallel pipeline eliminates the waiting time between wake-up and screen-scrolling, significantly improving batch processing throughput. Segmented transmission and dual-layer CRC check, combined with a fixed-point retransmission mechanism, fundamentally eliminate screen flickering, ghosting, and packet loss during data transmission. The incremental mapping pool synchronization mechanism minimizes inter-channel communication overhead and ensures efficient coordination. The tag end effectively extends battery life through a strategy of layered data parsing and disabling high-power parsing units during sleep phases.

[0069] In one embodiment, the dual-channel pipeline parallel processing mechanism of the wake-up data stream channel and the screen refresh data stream channel can be further extended to a multi-module cluster architecture. The middleware sets up a central scheduling center, under which multiple partition channel scheduling threads are configured. Each partition channel scheduling thread independently manages the wake-up module and at least one screen refresh module for its corresponding database tag group. The partition channel scheduling threads run in parallel, with data channels isolated between partitions, and they do not preempt each other's cache and channel resources. Taking a typical three-database cluster scenario as an example, the middleware central scheduling center has a first partition channel scheduling thread that manages the first wake-up module and multiple first screen refresh modules, covering database tag group A; a second partition channel scheduling thread that manages the second wake-up module and the second screen refresh module, covering database tag group B; and a third partition channel scheduling thread that manages the third wake-up module and the third screen refresh module, covering database tag group C. When business scale grows and more warehouse areas need to be covered, new partition channel scheduling threads can be added under the central scheduling center and corresponding wake-up modules and screen-scrolling modules can be attached to achieve linear expansion from dual modules to multiple modules, adapting to the cluster distributed scheduling needs of ultra-large warehouse areas.

[0070] This embodiment uses a modular design of partitioned channel scheduling threads, which allows the dual-channel parallel processing capability of a single database area to be independently replicated to multiple database areas for parallel operation. No coordination or synchronization is required between partitions, and the overall system throughput increases linearly with the number of partitions, meeting the synchronous and rapid refresh requirements of hundreds or even thousands of tags in large-scale dense deployment scenarios.

[0071] It should be understood that, although Figure 1-3 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1-3 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0072] In one embodiment, such as Figure 4 As shown, a data processing device based on RFID reader middleware is provided, including an RFID data acquisition module 402, a timing alignment unit 404, a two-factor verification unit 406, a trajectory analysis and discrimination unit 408, and an edge preprocessing unit 410.

[0073] The RFID data acquisition module 402 is used to acquire the tag radio frequency feature data collected by the RFID reader within the controlled area and the target spatial motion perception data synchronously collected by the millimeter-wave radar; the tag radio frequency feature data includes tag identification, signal phase and RSSI signal strength, and the target spatial motion perception data includes spatial coordinates, distance and moving speed. The timing alignment unit 404 is used to perform heterogeneous data protocol adaptation and clock timing alignment processing on the tag radio frequency feature data and the target spatial motion sensing data, and to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. The two-factor verification unit 406 is used to perform misread data filtering on RFID data based on the three-dimensional data association model and a two-factor verification mechanism that combines radio frequency features and spatial motion parameters, according to preset passage area spatial thresholds and motion thresholds. The trajectory analysis and discrimination unit 408 is used to perform continuous trajectory fitting analysis on the radar motion trajectory after the misread data filtering process, and to determine the passage status type of the controlled target by combining the RFID tag identification trigger timing. The edge preprocessing unit 410 is used to perform deduplication, error correction and structured preprocessing on the valid data after determining the passage status at the edge, and upload the preprocessed access control data to the background control platform.

[0074] Specific limitations regarding the data processing device based on RFID reader middleware can be found in the limitations of the data processing method based on RFID reader middleware mentioned above, and will not be repeated here. Each module in the aforementioned data processing device based on RFID reader middleware can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0075] In one embodiment, a computer device is provided, which may be a server or an edge computing gateway, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores RFID tag data, radar sensing data, device whitelists, and access control records. The network interface communicates with external RFID readers, millimeter-wave radar devices, and a backend management platform via a network connection. When executed by the processor, the computer program implements a data processing method based on RFID reader middleware.

[0076] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0077] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the method described above.

[0078] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0079] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A data processing method based on RFID reader middleware, characterized in that, The method includes: The system acquires RFID tag radio frequency characteristic data collected by RFID readers within the controlled area and target spatial motion sensing data synchronously collected by millimeter-wave radar. The RFID tag characteristic data includes tag identification, signal phase, and RSSI signal strength, while the target spatial motion sensing data includes spatial coordinates, distance, and movement speed. Heterogeneous data protocol adaptation and clock timing alignment are performed on the tag radio frequency feature data and the target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. Based on the aforementioned three-dimensional data association model, a two-factor verification mechanism combining radio frequency features and spatial motion parameters is adopted to filter RFID data for misreading based on preset spatial thresholds and motion thresholds in the passage area. The radar motion trajectory after the misread data filtering process is subjected to continuous trajectory fitting analysis, and the passage status type of the controlled target is determined by combining the RFID tag identification trigger timing. On the edge side, the valid data after determining the passage status is deduplicated, corrected for errors, and pre-processed into a structured format. The pre-processed access control data is then uploaded to the backend control platform.

2. The method according to claim 1, characterized in that, Heterogeneous data protocol adaptation and millisecond-level clock timing alignment are performed on the tag radio frequency feature data and the target spatial motion sensing data to establish a three-dimensional data association model between tag identification information, spatial location information, and motion state parameters, including: The message feature code protocol automatic identification engine distinguishes between RFID private protocols and general standard protocols, and converts the underlying raw hexadecimal radio frequency data into standard event data through a bidirectional standardized message conversion mechanism. The message aggregation window size is dynamically adjusted according to business latency requirements. The converted RFID standard data and radar sensing data are clock-aligned and matched with millisecond-level timestamps to form a three-dimensional data association structure of tag identification, spatiotemporal location and motion status.

3. The method according to claim 1, characterized in that, Based on the aforementioned three-dimensional data association model, a two-factor authentication mechanism combining radio frequency characteristics and spatial motion parameters is employed. This mechanism filters RFID data for misreading based on preset spatial and motion thresholds for the passage area, including: The real-time spatial position of the controlled target relative to the passage area is determined by the distance and spatial coordinate data collected by millimeter-wave radar. The RFID data corresponding to the controlled target whose spatial position is outside the preset passage area spatial threshold is filtered out as cross-read tag data outside the area. RFID data corresponding to control targets whose movement speed is consistently below a preset movement threshold will be filtered out as static lingering tag data within the area.

4. The method according to claim 1, characterized in that, The radar trajectory after the misread data filtering process is subjected to continuous trajectory fitting analysis. Combined with the RFID tag identification trigger timing sequence, the passage status type of the controlled target is determined, including: The trajectory curve of the controlled target is obtained by performing trajectory fitting on the continuous spatial coordinate data collected by millimeter-wave radar. Combined with the timestamp of the first and last identification of the RFID tag of the controlled target, the passage status of the controlled target is determined to be the entry status, exit status, turn-back status, lingering status, or traversing status.

5. The method according to claim 1, characterized in that, Before acquiring the RFID tag radio frequency characteristic data collected by RFID readers within the controlled area, the following steps are also included: Multi-factor composite registration verification is performed on RFID readers; the multi-factor composite registration verification includes network layer access pre-verification, device whitelist and service permission static verification, device hardware fingerprint two-way identity authentication, and dynamic heartbeat continuous online verification based on session key. If any verification fails, registration is rejected and connection is blocked.

6. The method according to claim 5, characterized in that, The device hardware fingerprint in the two-way authentication of device hardware fingerprint is generated by concatenating the device SN, main control chip ID, radio frequency module hardware ID, network port MAC address and factory calibration radio frequency parameters and then hashing them using SM3. The device hardware fingerprint two-way identity authentication includes the RFID reader / writer uploading the device SN and the device local random number R1 when initiating registration. After the middleware verifies the existence of the SN, it generates a platform random number R2 and uses the public key corresponding to the RFID reader / writer to encrypt and send it. The RFID reader decrypts the platform random number R2 using its local fixed private key, then concatenates the device's local random number R1, the platform random number R2, and the device's hardware fingerprint, generates a digest via SM3, signs it, and sends it back. The middleware uses a public key to verify the signature and check the consistency of the hardware fingerprint. Once the signature and fingerprint match, the device is determined to be legitimate and a session key SM4 is allocated.

7. The method according to claim 6, characterized in that, The dynamic heartbeat continuous online verification based on session keys includes initiating lightweight heartbeat verification every preset period after the device is successfully registered and a long connection is established. The heartbeat packet carries the device fingerprint and temporary session signature that are updated in real time. The middleware verifies the validity of the session key SM4 and the consistency of the device fingerprint. If the verification fails for a preset number of consecutive times, it determines that the device has been hijacked or replaced and automatically disconnects. The session key SM4 is automatically rotated at a preset period.

8. A data processing device based on RFID reader middleware, characterized in that, The method includes: The RFID data acquisition module is used to acquire tag radio frequency feature data collected by RFID readers within the controlled area and target spatial motion sensing data synchronously collected by millimeter-wave radar; the tag radio frequency feature data includes tag identification, signal phase and RSSI signal strength, and the target spatial motion sensing data includes spatial coordinates, distance and movement speed. The timing alignment unit is used to perform heterogeneous data protocol adaptation and clock timing alignment processing on the tag radio frequency feature data and the target spatial motion sensing data, and to establish a three-dimensional data association model between tag identification information, spatial location information and motion state parameters. The two-factor verification unit is used to perform misread data filtering on RFID data based on the three-dimensional data association model and a two-factor verification mechanism that combines radio frequency features and spatial motion parameters, according to preset passage area spatial thresholds and motion thresholds. The trajectory analysis and discrimination unit is used to perform continuous trajectory fitting analysis on the radar motion trajectory after the misread data filtering process, and to determine the passage status type of the controlled target by combining the RFID tag identification trigger timing. The edge preprocessing unit is used to perform deduplication, error correction, and structured preprocessing on the valid data after determining the passage status at the edge, and upload the preprocessed access control data to the backend control platform.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images