A method and apparatus for detecting lost packages for an open RFID system

Abstract

This invention discloses a method and device for detecting lost packages in open RFID systems, applicable to inventory scenarios with unexpected tag interference. The method constructs virtual frames and calculates tag mapping time slots based on a frame time slot mechanism, classifies time slots, and dynamically selects representative tags for unconfirmed packages and a set of selected time slots. Based on the selected time slots, it constructs and broadcasts filtering-sorting instructions to filter out non-representative and unexpected tags and assign response order to representative tags. Each active tag determines its silent / response state according to the instructions and responds in a designated time slot. The actual response is compared with the expected response to confirm the existence of the package and determine its loss. Relevant parameters are adaptively updated, and the process is iteratively executed until reliability requirements are met or all packages are confirmed, outputting the lost package identification result. This invention improves the identification success rate and time efficiency.

Description

Technical Field

[0001] This invention relates to the field of radio frequency identification (RFID) item monitoring and anomaly detection, and in particular to a method for quickly identifying lost packages in open environments with unexpected tag interference, for RFID systems that are packaged in boxes or parcels. Background Technology

[0002] In warehousing and logistics, access control and inventory management, and inbound / outbound verification, RFID is widely used for automatic identification and tracking of items. In actual deployments, items are often stacked, packaged, and circulated in units of boxes / parcels, and loss events tend to occur at the parcel level. Most existing methods focus on tag-level identification, requiring the collection of all tag responses or monitoring of a fixed set of key tags, resulting in high communication overhead. Furthermore, in open scenarios, a large number of unexpected tags (unknown tags outside the monitoring range) compete for time slots with the tags to be monitored, causing collisions and leading to missed detections and reduced efficiency. To adapt to the characteristic of boxed systems where "one response per parcel confirms its existence," a highly efficient identification method is needed that can dynamically select representative tags, quickly suppress interfering tags, and allocate sequential responses to representative tags in consecutive time slots. Summary of the Invention

[0003] Purpose of the invention: The technical problem to be solved by the present invention is to address the shortcomings of the prior art by providing a lost package detection method and device for open RFID systems, thereby improving the efficiency and reliability of lost package detection.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0005] A method for detecting lost packages in an open RFID system includes:

[0006] Obtain pre-constructed package-tag membership information, which records the division relationship between the known tag set and packages in the system, wherein each package contains at least one known tag, and the known tags of different packages do not overlap;

[0007] Based on the known tag set, a hash function is used to calculate the time slot index of each known tag. According to the distribution of time slot indexes and the package-tag membership information, the calculated time slot indexes are divided into single-tag time slot indexes, homogeneous time slot indexes, and conflicting time slot indexes. The homogeneous time slot index is a time slot index calculated from multiple known tags that all come from the same package. The conflicting time slot index is a time slot index calculated from multiple known tags that come from at least two different packages. Based on the time slot index division results, a representative tag is selected from each package to form a representative tag set.

[0008] A filter-sorting indicator vector is constructed based on the slot index of the selected representative label. The filter-sorting indicator vector is used to filter out non-representative labels and unexpected labels, and to assign a response order to each representative label.

[0009] The filtering-sorting indication vector and its construction parameters are broadcast, and the actual responses of the RFID tags are received. The RFID tags within the coverage area determine their own status and response order according to the vector and its construction parameters, and respond in the allocated time slots.

[0010] The missing packages are identified based on the comparison between the actual response and the expected response, and it is determined whether the identification result meets the preset reliability threshold. If there is no response in the time slot corresponding to the tag, it is determined that the corresponding package is missing. If the reliability requirement is not met, the relevant parameters are updated and the process returns to the time slot index calculation step to continue processing until the reliability requirement is met or the preset round limit is reached.

[0011] Furthermore, the time slot index of each known tag is calculated using a hash function, as shown in the following formula:

[0012]

[0013] in, For time slot index, For known labels , For the specified hash seed, For the first The frame size of the wheel;

[0014] Based on the time slot index partitioning results, a representative tag is selected from each package to form a representative tag set, including:

[0015] In the initial round, all tags corresponding to single-tag slot indices and all tags corresponding to isomorphic slot indices are selected as representative tags;

[0016] In subsequent rounds, if a single-label time slot index exists, a representative label is randomly selected from the labels of the corresponding single-label time slot index; if no single-label time slot index exists, a label of the corresponding isomorphic time slot index is selected; if neither exists, a label is selected from the labels of the corresponding conflicting time slot indices, prioritizing the label corresponding to the time slot index with the fewest labels.

[0017] Further, a filter-sorting indicator vector is constructed based on the slot index of the selected representative label, including:

[0018] Arrange the time slot indices of the selected representative tags in ascending order, take the first time slot index as the first element of the index vector, and calculate the difference between adjacent time slot indices in turn as the subsequent elements of the index vector to form the index vector.

[0019] Based on the time slot index type to which each representative tag belongs and the fingerprint distinguishability of the representative tag from other tags in the same time slot, an identifier value is generated for each representative tag's time slot index. All identifier values ​​are arranged in the order of the representative tag's time slot index to form a sorting vector. The sorting vector is used to indicate whether each representative tag is responding in this round and the order of the responses.

[0020] Sample labels are selected from the time slots corresponding to the time slot indices marked as responses in the sorting vector to calculate fingerprints. The calculated fingerprints are then arranged in the order of responses representing the labels to form a fingerprint vector.

[0021] The index vector, sorting vector, and fingerprint vector constitute the filter-sorting indicator vector.

[0022] Furthermore, in the sorting vector, if the time slot index representing the tag is a single tag time slot index or a homogeneous time slot index, then the corresponding identifier value in the sorting vector is set to the first value;

[0023] If the time slot index representing the tag is a conflicting time slot index, and the fingerprint of the tag cannot be distinguished from that of other tags in the same time slot, then the corresponding identifier value in the sorting vector is set to the second value; otherwise, it is set to the first value.

[0024] The first value indicates that the tag responds in this round, and the second value indicates that the tag does not respond in this round.

[0025] Furthermore, receiving the actual response from the RFID tag includes:

[0026] In the first round, a 10-bit checksum is received from the tag, and the number of unexpected tags is estimated based on the time slot status; in subsequent rounds, a 1-bit or a preset short bit is received from the tag for package existence confirmation.

[0027] The number of unexpected tags in the first round is estimated based on the number of empty time slots, the number of checksum consistent time slots, and the number of collision time slots, as shown in the following formula:

[0028]

[0029] In the formula, This represents an unexpected number of tags. The length of the first frame. For the total number of packages, Number of empty time slots To verify the consistency of the number of time slots, This is a preset constant; denoted as the number of collision time slots; e is the natural constant.

[0030] Furthermore, when the reliability requirements are not met, relevant parameters are updated, including the frame length, which is determined by minimizing the average recognition time.

[0031] Average recognition time Calculated using the following formula:

[0032]

[0033] in Let i be the total time for the i-th round. Let i be the number of packages confirmed to exist in the i-th round;

[0034] The Calculated using the following formula:

[0035]

[0036] in Let i be the number of packages that have not yet been confirmed in the i-th round. Let Euler's constant be 1. For a single bit transmission time, This refers to the time slot interval.

[0037] Finding the solution using numerical methods Minimum frame length As the optimal frame length.

[0038] Furthermore, when the reliability requirements are not met, relevant parameters are updated. The updated parameters include the fingerprint length and the fingerprint length in the i-th round. The calculation formula is as follows:

[0039]

[0040] in, Let i be the number of unexpected tags in round i. For frame length, This is a preset reliability threshold.

[0041] A lost package detection device for an open RFID system includes:

[0042] The information acquisition module is used to acquire pre-constructed package-tag membership information. The package-tag membership information records the division relationship between the known tag set and the package in the system. Each package contains at least one known tag, and the known tags of different packages do not overlap.

[0043] The representative tag acquisition module is used to calculate the time slot index of each known tag based on the known tag set using a hash function. According to the time slot index distribution and package-tag affiliation information, the calculated time slot index is divided into single-tag time slot index, homogeneous time slot index, and conflicting time slot index. A homogeneous time slot index is a time slot index calculated from multiple known tags that all come from the same package, while a conflicting time slot index is a time slot index calculated from multiple known tags that come from at least two different packages. Based on the time slot index division results, a representative tag is selected from each package to form a representative tag set.

[0044] A vector construction module is used to construct a filter-sorting indicator vector based on the slot index of the selected representative label. The filter-sorting indicator vector is used to filter out non-representative labels and unexpected labels, and to assign a response order to each representative label.

[0045] The communication module is used to broadcast the filtering-sorting indication vector and its construction parameters, and receive the actual response from the RFID tags. The RFID tags within the coverage area determine their own status and response order according to the vector and its construction parameters, and respond in the allocated time slot.

[0046] The identification and iteration module is used to identify missing packages based on the comparison between the actual response and the expected response, and to determine whether the identification result meets the preset reliability threshold. Specifically, if there is no response in the time slot corresponding to the tag, it is determined that the corresponding package is missing. If the reliability requirement is not met, the relevant parameters are updated and the process returns to the time slot index calculation step to continue processing until the reliability requirement is met or the preset round limit is reached.

[0047] The present invention also provides an electronic device comprising: one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, wherein when the programs are executed by the processors, they implement the steps of the lost package detection method for an open RFID system as described above.

[0048] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the lost package detection method for an open RFID system as described above.

[0049] Beneficial effects: Compared with tag-level methods that require responses from all tags, this invention makes full use of package-tag association information, enabling the system to complete package-level confirmation with fewer responses in each round of identification; compared with methods that only monitor fixed key tags, this invention can obtain more available single time slots or same package time slots through dynamic representative tag selection, improving the utilization rate of identification time slots; and by reducing interference from unexpected tags through fingerprint filtering and online estimation, it improves the identification success rate and time efficiency. Attached Figure Description

[0050] Figure 1 This is a flowchart of a lost package detection method based on an open RFID system.

[0051] Figure 2 This is a model diagram based on an open RFID system. Detailed Implementation

[0052] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading this invention, any modifications of the invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0053] This invention provides a lost package detection method for open RFID systems, comprising:

[0054] Obtain pre-constructed package-tag membership information, which records the division relationship between the known tag set and packages in the system, wherein each package contains at least one known tag, and the known tags of different packages do not overlap;

[0055] Based on the known tag set, a hash function is used to calculate the time slot index of each known tag. According to the distribution of time slot indexes and the package-tag membership information, the calculated time slot indexes are divided into single-tag time slot indexes, homogeneous time slot indexes, and conflicting time slot indexes. The homogeneous time slot index is a time slot index calculated from multiple known tags that all come from the same package. The conflicting time slot index is a time slot index calculated from multiple known tags that come from at least two different packages. Based on the time slot index division results, a representative tag is selected from each package to form a representative tag set.

[0056] A filter-sorting indicator vector is constructed based on the slot index of the selected representative label. The filter-sorting indicator vector is used to filter out non-representative labels and unexpected labels, and to assign a response order to each representative label.

[0057] The filtering-sorting indication vector and its construction parameters are broadcast, and the actual responses of the RFID tags are received. The RFID tags within the coverage area determine their own status and response order according to the vector and its construction parameters, and respond in the allocated time slots.

[0058] The missing packages are identified based on the comparison between the actual response and the expected response, and it is determined whether the identification result meets the preset reliability threshold. If there is no response in the time slot corresponding to the tag, it is determined that the corresponding package is missing. If the reliability requirement is not met, the relevant parameters are updated and the process returns to the time slot index calculation step to continue processing until the reliability requirement is met or the preset round limit is reached.

[0059] In the presence of interference from non-representative known tags and unexpected unknown tags, the above method achieves the following: ① Each package only requires one (or a few) representative tag responses to confirm its existence; ② Reduces empty and conflicting time slots through an integrated filtering and sorting mechanism, thereby improving the recognition speed; ③ Completes the estimation of the number of unexpected tags and guides the adaptive configuration of parameters without introducing additional online overhead, thus meeting the preset recognition reliability requirements.

[0060] In one specific embodiment, the method of the present invention is described in detail from the entire process of system modeling, problem definition, and solution, referring to... Figure 1 The method includes the following steps:

[0061] Step 1: Create an open RFID system model consisting of a backend server, RFID readers, and a large number of passive tags. Each tag has a unique identifier. It is used to uniquely identify the item to which it is attached. For example, Figure 2The open RFID system shown includes a backend server, RFID readers, and RFID tags in multiple packages. Each package contains one or more items, and each item is attached with a unique RFID tag. The backend server maintains the affiliation between packages and tags, recording which tags each package contains; for example, package 1 may contain tags t1 to t6, and package 2 may contain tags t7 to t10. The reader wirelessly interacts with the tags within its coverage area and obtains tag association information for the packages to be monitored from the backend server to support subsequent representative tag selection, mapping time slot calculation, and lost package identification. In this system, packages are typically stored and circulated in a stacked, packaged form; therefore, loss events tend to occur on a package-by-package basis. Confirming the presence of at least one representative tag in each package is sufficient to determine the existence of the corresponding package.

[0062] Specifically, during the packaging, warehousing, or initial inventory initialization phases, the reader performs batch readings of all RFID tags attached to each package, collecting the tag's unique identifier (e.g., EPC or ID) and necessary business fields. The backend server cleans and verifies the consistency of the collected data, then writes the set of tag IDs corresponding to each package into the package-tag affiliation database. This allows the system to quickly query "which tags a package contains" and "which package a certain tag belongs to" in subsequent rounds, thus supporting tag selection, virtual frame construction, and missing package location.

[0063] Step 2: Define the lost package detection problem for open RFID systems based on the system model in Step 1.

[0064] use Represents the known set of labels in the system, where Given the total number of labels; set The tags in are divided into A number of non-overlapping packages are represented as Among them, packages Include Each package is labeled. For each package, a representative label is selected, forming a set of representative labels. ,in Each represents a label. Unique Identifier Package The remaining known tags form a set of non-representative tags. ,in This indicates the number of non-representative tags. Additionally, the set of unexpected tags in the system is... Unexpected tags refer to tags that do not belong to the current set of monitored packages and whose identity information is not included in the current package-tag affiliation. However, because the items or people attached to these tags enter the reader's coverage area, they may still respond to the reader's queries, thus competing for time slots with known tags and causing interference. The set of lost packages is defined as... ,in The lost package detection problem for open RFID systems is defined as follows:

[0065]

[0066] in For the recognition results, This is the required recognition reliability threshold for the system.

[0067] Step 3: Based on all known tags, including representative and non-representative tags, the reader constructs a virtual frame using the frame slot Aloha protocol. This "virtual frame" refers to a set of slots logically constructed for the active known tags in the current round, based on the frame length and hash seed, used to predict the mapping position of each tag in this round. Representative tags are then selected based on the slot mapping results.

[0068] Step 3 specifically includes:

[0069] Step 3-1: The reader calculates the mapping time slot for each tag:

[0070]

[0071] in For tags , For hash seed, For the first The frame size of the round. The "mapped time slot" is a logical time slot index value calculated by the tag based on its identifier, hash seed, and frame length, and is not the actual physical time slot after the tag has actually been transmitted on the physical channel. Therefore, the subsequent single-tag time slot, homogeneous time slot, and conflicting time slot are usually classified according to the mapped time slot index; in the absence of ambiguity, they can also be simply referred to as the time slot of the corresponding type.

[0072] Step 3-2: Divide the calculated mapping slot index into three categories: single-tag slot (a slot is mapped to only one known tag), homogeneous slot (a slot is mapped to multiple known tags and all come from the same package), and conflicting slot (a slot is mapped to multiple known tags and comes from at least two different packages).

[0073] Step 3-3: The reader selects representative tags based on these time slot types, thereby obtaining the representative tag set for this round. and its mapped time slot set The strategy for selecting representative labels differs between the initial round and subsequent rounds.

[0074] Specifically, it includes:

[0075] Step 3-2-1: The initial round aims to estimate the number of unexpected labels, for which all labels mapped to single-label time slots and all labels mapped to isomorphic time slots are selected.

[0076] Step 3-2-2: Subsequent rounds are used to identify missing packages. Representative labels are selected according to the following rules: If a single-label time slot exists, a label is randomly selected from the labels mapped to the single-label time slot; if a single-label time slot does not exist, a label mapped to a homogeneous time slot is selected; if neither of the above two time slots exists, a label is selected from the labels mapped to the same conflicting time slot as other package labels, with priority given to the time slot with the fewest labels.

[0077] Step 4: Construct a filtering-sorting indicator vector based on the representative labels selected in Step 3. The filtering-sorting indicator vector is a set of synthesized indicator information, including an index vector, a sorting vector, and a fingerprint vector. The index vector and fingerprint vector respectively perform filtering functions, while the sorting vector indicates whether the representative label is in the current response round and its response order. In this paper, the index filter can be understood as the index vector, and the fingerprint filter can be understood as the fingerprint vector.

[0078] Step 4 specifically includes:

[0079] Step 4-1: Based on the mapping slot index set of the representative labels in this round, construct an index filter through adjacent index differential encoding. .

[0080] Specifically, it includes:

[0081] Step 4-1-1: For the first Wheel represents a set of tags Each of the following represents a label Calculate its mapping time slot To form an index set And sort them in ascending order to obtain the sequence: ,in This represents the number of time slots selected in this round.

[0082] Step 4-1-2: Set the first element of the index filter to Through calculation Index filters are constructed using the differences between consecutive indexes. .

[0083] Step 4-2: Construct a sorting vector based on the selected time slot set and the expected response or silent state and response order of its corresponding representative tags. For the first Selected time slots :

[0084] If the time slot is a single-tag time slot or a time slot for the same package, then .

[0085] If the time slot is a anticipated conflict time slot (i.e., the representative label overlaps with a non-representative label in the same mapping time slot), then when the representative label's fingerprint is identical to that of an interfering label in the same mapping time slot, making it impossible to distinguish the time slot, then... Otherwise .

[0086] And define the number of verifiable packages in this round:

[0087]

[0088] Step 4-3: Construct a fingerprint filter based on the sample label fingerprint sequence and fingerprint length or hash parameters obtained from the selected time slot. .

[0089] Specifically, it includes:

[0090] Step 4-3-1: When At that time, for all those that satisfy For the selected time slot, randomly select one label from that time slot as the sample label and calculate its fingerprint:

[0091]

[0092] Write sequentially according to the selected time slots. ,make .

[0093] Step 5: Reader Broadcast Filtering - Sort Indicator Vector and its construction parameters, the system receives The active tags calculate and determine their own state and response order based on the construction parameters and respond in the scheduled time slots.

[0094] Step 5 specifically includes:

[0095] Step 5-1: The reader generates a filter-sorting indicator vector and its construction parameters based on the construction results of this round, and broadcasts the indicator vector and its construction parameters to the tags within the coverage area.

[0096] Step 5-2: When the first round ( Reader Broadcast ⟨ , > and construction parameters In the second round and beyond Reader Broadcast ⟨ , , >and construction parameters⟨ , , , >. Among them, For the first Frame length, For the first The hash seed used for round-robin time slot calculation. The hash seed used for fingerprint generation, For the first Fingerprint length.

[0097] Step 5-3: The tag calculates its own mapped time slot. And accumulate the difference to recover the selected time slot sequence:

[0098]

[0099] If it exists make Then the label determines that it has been selected and obtains an index. Otherwise, remain silent.

[0100] Step 5-4: If the label is selected and The tag remains active but does not respond in this round (the tagging instruction is sent by the reader). (Make it participate in the next round), if the tag is selected and Then the tag enters the response set, and in " The response in the consecutive time slots corresponding to the number of 1s preceding it.

[0101] Steps 5-6: During the second round and beyond and At that time, the tag calculates its own fingerprint. ;like It will respond in a continuous time slot s, otherwise it will remain silent.

[0102] Steps 5-7: In round 1, the response tag sends a 10-bit checksum, which the reader uses to determine single tag / collision in the time slot and to estimate the number of unexpected tags. In round 2 and thereafter, the response tag sends 1 bit or a preset short bit for package existence confirmation.

[0103] Step 6: The reader compares the actual response with the expected response, identifies the missing package, and determines whether the missing package identification result meets the reliability requirements described in Step 2.

[0104] Step 6 specifically includes:

[0105] Step 6-1: The reader, according to... Obtain the "expected response time slot set and sequence" for this round, and collect the actual response.

[0106] Step 6-2: For any assigned response time slot, if there is no response in the corresponding time slot, determine that the corresponding package is missing and add it to the identification result. .

[0107] Step 6-3: In the first round, due to the presence of unexpected label interference, "only missing packages whose labels are mapped to empty time slots" can be accurately identified; the remaining packages need to be confirmed in subsequent rounds after fingerprint filtering.

[0108] Step 6-4: The reader determines the reliability threshold based on step 2. Determine if the current recognition result meets the constraints; if not, update the parameters for the next round (e.g., ...). , , (etc.) and return to step 3 to enter the next round, until the reliability requirements are met or the preset number of rounds is reached.

[0109] The steps performed by the reader in the method described in this invention can be calculated by the reader itself or by a backend server that communicates with the reader.

[0110] According to an embodiment of the present invention, the number of unexpected labels in steps 5-7 above The estimation and subsequent parameter updates are solved using the following method:

[0111] Let the probabilities of the first and second scenarios be . and The corresponding expected number of time slots is and After the first round, tags in single-label, homogeneous, or specific conflict slots (excluding those mapped to) (Each time slot participant) remains active. For packages , No. wheel The probabilities that the representative label is mapped to a single-label time slot and a homogeneous time slot are respectively and .in:

[0112]

[0113] For the first Wheel package Number of tags still active in . For the first The total number of active and non-representative package tags, and The probability of homogeneous time slots is:

[0114]

[0115] The first scenario occurs when only the representative label of the missing package is mapped to a single-label or homogeneous time slot, and no unexpected labels are mapped to these time slots. Since a missing package means that all labels within the package are missing, therefore:

[0116]

[0117] The second scenario occurs when: the representative tag of a non-missing package is mapped to a single-tag time slot; or the representative tag of a missing package is mapped to a single-tag time slot and an unexpected tag happens to match its 10-bit checksum. The latter scenario has a very low probability and can be ignored; therefore:

[0118]

[0119] thereby:

[0120]

[0121] when season , We can solve it first:

[0122]

[0123] Then by have to:

[0124]

[0125] when (This means either there are no missing packages or there are many unexpected tags), so the slot count estimation for case 4 is used. This situation occurs when: a time slot is mapped to by "1 missing representative label + at least 2 unexpected labels"; or by "1 non-missing representative label + at least 1 unexpected label". Therefore:

[0126]

[0127] From this, we can conclude that The expression (omitting the expansion in the middle) is approximated as:

[0128]

[0129] when Large or When, the second term in the above equation can be ignored, let have to:

[0130]

[0131] therefore :

[0132]

[0133] After the first round, unexpected labels mapped to single-label, homogeneous, or specific conflicting time slots remain active. The maximum number of such conflicting time slots is [number missing]. Let's consider this worst-case scenario. Assume the number of unexpected labels in the second round is... And initially ,but:

[0134]

[0135] Will Substituting, we can obtain for:

[0136]

[0137] e is a natural constant.

[0138] Consider the first wheel( ).make The number of packages that have not yet been confirmed. The total number of activated package tags and non-representative tags. This represents the number of unexpected labels. The number of packages confirmed to exist in the wheel is denoted as The initial conditions are: , , .

[0139] The expected value of the time slot selected for identifying missing packages is:

[0140]

[0141] then In the first Wheel, parameters updated to:

[0142]

[0143] In fact, the AMPI method described in this invention only requires selecting a very small portion of representative labels to verify package integrity.

[0144] According to an embodiment of the present invention, the maximum index difference length in step 4-1-2 above is... and fingerprint length The solution method is as follows:

[0145] This represents the bit length of the largest adjacent difference among the slot indices (sorted) to which the tag is mapped. Equivalently, this is equivalent to a bit length of... The rope was randomly cut into pieces. Given a segment, find the length of the longest segment. .when At that time, by the principle of inclusion-exclusion, we get Exceed The probability is:

[0146]

[0147] Its expectation is:

[0148]

[0149] in For the first A harmonic number. Using... approximate and with Substitution ,have to:

[0150]

[0151] Unexpected tags average approximately [number missing] per time slot There are 100 fingerprints. The probability that their fingerprints differ from the representative labels is 100. Therefore, in order to meet reliability requirements , No. Wheel fingerprint length Should meet:

[0152]

[0153] Solving for:

[0154]

[0155] According to an embodiment of the present invention, the optimal frame length for calculating the mapping time slot in step 3 above is... The solution method is as follows:

[0156] For the wheel Frame length Optimize to reduce average label recognition time Minimum. (For derivation) First, analyze the first Total time of round It consists of the time of the reader broadcast filter-sorting indicator and the time when the tag sends 1 bit to the reader, therefore:

[0157]

[0158]

[0159] Based on the foregoing analysis, the first Round confirmed There are [number] packages. Therefore, the average recognition time is:

[0160]

[0161] Use the above parameters Substituting the function form into the equation yields the following result. The expression is too long to be expanded further. Because it is difficult to obtain... The closed-form solution is found using numerical methods. Minimum optimal frame length .beg The core of the numerical method is to minimize the average recognition time. To achieve the goal, iterate through and optimize the frame size values. The value is the total time of a single round. This round uses fingerprint vectors The quotient of the target tag count. Traverse within a reasonable range. Values ​​are used for calculation , take Minimum For this round .

[0162] The present invention also provides a lost package detection device for an open RFID system, comprising:

[0163] The information acquisition module is used to acquire pre-constructed package-tag membership information. The package-tag membership information records the division relationship between the known tag set and the package in the system. Each package contains at least one known tag, and the known tags of different packages do not overlap.

[0164] The representative tag acquisition module is used to calculate the time slot index of each known tag based on the known tag set using a hash function. According to the time slot index distribution and package-tag affiliation information, the calculated time slot index is divided into single-tag time slot index, homogeneous time slot index, and conflicting time slot index. A homogeneous time slot index is a time slot index calculated from multiple known tags that all come from the same package, while a conflicting time slot index is a time slot index calculated from multiple known tags that come from at least two different packages. Based on the time slot index division results, a representative tag is selected from each package to form a representative tag set.

[0165] A vector construction module is used to construct a filter-sorting indicator vector based on the slot index of the selected representative label. The filter-sorting indicator vector is used to filter out non-representative labels and unexpected labels, and to assign a response order to each representative label.

[0166] The communication module is used to broadcast the filtering-sorting indication vector and its construction parameters, and receive the actual response from the RFID tags. The RFID tags within the coverage area determine their own status and response order according to the vector and its construction parameters, and respond in the allocated time slot.

[0167] The identification and iteration module is used to identify missing packages based on the comparison between the actual response and the expected response, and to determine whether the identification result meets the preset reliability threshold. Specifically, if there is no response in the time slot corresponding to the tag, it is determined that the corresponding package is missing. If the reliability requirement is not met, the relevant parameters are updated and the process returns to the time slot index calculation step to continue processing until the reliability requirement is met or the preset round limit is reached.

[0168] The present invention also provides an electronic device comprising: one or more processors; a memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, wherein when the programs are executed by the processors, they implement the steps of the lost package detection method for an open RFID system as described above.

[0169] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the lost package detection method for an open RFID system as described above.

[0170] The above are merely preferred embodiments of the present invention. Those skilled in the art can make equivalent substitutions or modifications to the step sequence, parameter update strategy, and vector encoding method without departing from the spirit of the present invention, and all such substitutions or modifications should fall within the protection scope of the present invention.

Claims

1. A method for detecting lost packages in an open RFID system, characterized in that, The method includes: Obtain pre-constructed package-tag membership information, which records the division relationship between the known tag set and packages in the system, wherein each package contains at least one known tag, and the known tags of different packages do not overlap; Based on the known tag set, a hash function is used to calculate the time slot index of each known tag. According to the distribution of time slot indexes and the package-tag membership information, the calculated time slot indexes are divided into single-tag time slot indexes, homogeneous time slot indexes, and conflicting time slot indexes. The homogeneous time slot index is a time slot index calculated from multiple known tags that all come from the same package. The conflicting time slot index is a time slot index calculated from multiple known tags that come from at least two different packages. Based on the time slot index division results, a representative tag is selected from each package to form a representative tag set. A filter-sorting indicator vector is constructed based on the slot index of the selected representative label. The filter-sorting indicator vector is used to filter out non-representative labels and unexpected labels, and to assign a response order to each representative label. The filtering-sorting indication vector and its construction parameters are broadcast, and the actual responses of the RFID tags are received. The RFID tags within the coverage area determine their own status and response order according to the vector and its construction parameters, and respond in the allocated time slots. The missing packages are identified based on the comparison between the actual response and the expected response, and it is determined whether the identification result meets the preset reliability threshold. If there is no response in the time slot corresponding to the tag, it is determined that the corresponding package is missing. If the reliability requirement is not met, the relevant parameters are updated and the process returns to the time slot index calculation step to continue processing until the reliability requirement is met or the preset round limit is reached.

2. The method according to claim 1, characterized in that, The time slot index of each known tag is calculated using a hash function, and the calculation formula is as follows: in, For time slot index, For known labels , For the specified hash seed, For the first The frame size of the wheel; Based on the time slot index partitioning results, a representative tag is selected from each package to form a representative tag set, including: In the initial round, all tags corresponding to single-tag slot indices and all tags corresponding to isomorphic slot indices are selected as representative tags; In subsequent rounds, if a single-label time slot index exists, a representative label is randomly selected from the labels of the corresponding single-label time slot index; if no single-label time slot index exists, a label of the corresponding isomorphic time slot index is selected; if neither exists, a label is selected from the labels of the corresponding conflicting time slot indices, prioritizing the label corresponding to the time slot index with the fewest labels.

3. The method according to claim 1, characterized in that, Construct a filter-sorting indicator vector based on the slot index of the selected representative label, including: Arrange the time slot indices of the selected representative tags in ascending order, take the first time slot index as the first element of the index vector, and calculate the difference between adjacent time slot indices in turn as the subsequent elements of the index vector to form the index vector. Based on the time slot index type to which each representative tag belongs and the fingerprint distinguishability of the representative tag from other tags in the same time slot, an identifier value is generated for each representative tag's time slot index. All identifier values ​​are arranged in the order of the representative tag's time slot index to form a sorting vector. The sorting vector is used to indicate whether each representative tag is responding in this round and the order of the responses. Sample labels are selected from the time slots corresponding to the time slot indices marked as responses in the sorting vector to calculate fingerprints. The calculated fingerprints are then arranged in the order of responses representing the labels to form a fingerprint vector. The index vector, sorting vector, and fingerprint vector constitute the filter-sorting indicator vector.

4. The method according to claim 3, characterized in that, In the sorting vector, if the time slot index representing the tag is a single tag time slot index or a homogeneous time slot index, then the corresponding identifier value in the sorting vector is set to the first value. If the time slot index representing the tag is a conflicting time slot index, and the fingerprint of the tag cannot be distinguished from that of other tags in the same time slot, then the corresponding identifier value in the sorting vector is set to the second value; otherwise, it is set to the first value. The first value indicates that the tag responds in this round, and the second value indicates that the tag does not respond in this round.

5. The method according to claim 1, characterized in that, The actual response received from the RFID tag includes: In the first round, a 10-bit checksum is received from the tag, and the number of unexpected tags is estimated based on the time slot status; in subsequent rounds, a 1-bit or a preset short bit is received from the tag for package existence confirmation. The number of unexpected tags in the first round is estimated based on the number of empty time slots, the number of checksum consistent time slots, and the number of collision time slots, as shown in the following formula: In the formula, This represents an unexpected number of tags. The length of the first frame. For the total number of packages, Number of empty time slots To verify the consistency of the number of time slots, This is a preset constant; denoted as the number of collision time slots; e is the natural constant.

6. The method according to claim 1, characterized in that, When reliability requirements are not met, relevant parameters are updated, including the frame length, which is determined by minimizing the average recognition time. Average recognition time Calculated using the following formula: in Let i be the total time for the i-th round. Let i be the number of packages confirmed to exist in the i-th round; The Calculated using the following formula: in Let i be the number of packages that have not yet been confirmed in the i-th round. Let Euler's constant be 1. For a single bit transmission time, This refers to the time slot interval. Finding the solution using numerical methods Minimum frame length As the optimal frame length.

7. The method according to claim 1, characterized in that, When reliability requirements are not met, relevant parameters are updated, including fingerprint length and the length of the fingerprint in the i-th round. The calculation formula is as follows: in, Let i be the number of unexpected tags in round i. For frame length, This is a preset reliability threshold.

8. A lost package detection device for an open RFID system, characterized in that, include: The information acquisition module is used to acquire pre-constructed package-tag membership information. The package-tag membership information records the division relationship between the known tag set and the package in the system. Each package contains at least one known tag, and the known tags of different packages do not overlap. The representative tag acquisition module is used to calculate the time slot index of each known tag based on the known tag set using a hash function. According to the time slot index distribution and package-tag affiliation information, the calculated time slot index is divided into single-tag time slot index, homogeneous time slot index, and conflicting time slot index. A homogeneous time slot index is a time slot index calculated from multiple known tags that all come from the same package, while a conflicting time slot index is a time slot index calculated from multiple known tags that come from at least two different packages. Based on the time slot index division results, a representative tag is selected from each package to form a representative tag set. A vector construction module is used to construct a filter-sorting indicator vector based on the slot index of the selected representative label. The filter-sorting indicator vector is used to filter out non-representative labels and unexpected labels, and to assign a response order to each representative label. The communication module is used to broadcast the filtering-sorting indication vector and its construction parameters, and receive the actual response from the RFID tags. The RFID tags within the coverage area determine their own status and response order according to the vector and its construction parameters, and respond in the allocated time slot. The identification and iteration module is used to identify missing packages based on the comparison between the actual response and the expected response, and to determine whether the identification result meets the preset reliability threshold. Specifically, if there is no response in the time slot corresponding to the tag, it is determined that the corresponding package is missing. If the reliability requirement is not met, the relevant parameters are updated and the process returns to the time slot index calculation step to continue processing until the reliability requirement is met or the preset round limit is reached.

9. An electronic device, characterized in that, include: One or more processors; Memory; And one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, wherein when the programs are executed by the processors, they implement the steps of the lost package detection method for an open RFID system as claimed in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the lost package detection method for an open RFID system as described in any one of claims 1-7.

Images