CN117938210A - Method and system for identifying lost tags in parallel in multi-category RFID system - Google Patents
Method and system for identifying lost tags in parallel in multi-category RFID system Download PDFInfo
- Publication number
- CN117938210A CN117938210A CN202410100948.8A CN202410100948A CN117938210A CN 117938210 A CN117938210 A CN 117938210A CN 202410100948 A CN202410100948 A CN 202410100948A CN 117938210 A CN117938210 A CN 117938210A
- Authority
- CN
- China
- Prior art keywords
- tag
- category
- time slot
- reader
- bit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 239000013598 vector Substances 0.000 claims abstract description 104
- 230000008707 rearrangement Effects 0.000 claims abstract description 30
- 238000012795 verification Methods 0.000 claims description 22
- 238000010276 construction Methods 0.000 claims description 14
- 238000012790 confirmation Methods 0.000 claims description 6
- 230000007704 transition Effects 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 17
- 238000010586 diagram Methods 0.000 description 11
- 206010010099 Combined immunodeficiency Diseases 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000004088 simulation Methods 0.000 description 5
- 238000009826 distribution Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000007423 decrease Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000009827 uniform distribution Methods 0.000 description 3
- 238000002679 ablation Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- JXASPPWQHFOWPL-UHFFFAOYSA-N Tamarixin Natural products C1=C(O)C(OC)=CC=C1C1=C(OC2C(C(O)C(O)C(CO)O2)O)C(=O)C2=C(O)C=C(O)C=C2O1 JXASPPWQHFOWPL-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000001360 collision-induced dissociation Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000005672 electromagnetic field Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000003334 potential effect Effects 0.000 description 1
- 238000007430 reference method Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Landscapes
- Near-Field Transmission Systems (AREA)
Abstract
The invention discloses a method and a system for identifying lost tags in parallel in a multi-category RFID system. Queuing the K label categories in the RFID system through a queuing vector V i, and distributing independent heat codes of a category layer for each label based on indexes; for candidate tab set C k with category index k, a sub-notification vector is constructed based on the preselected time slotsGet notification vectorThe broadcast V N performs tag reply time slot rearrangement; and decoding reply signals of different types of labels from the aggregate signals of each time slot, so as to realize parallel identification of a plurality of types of labels. The invention generates the category layer single thermal code for each tag through the category identifier, thereby avoiding the conflict between different category tags. By constructing the notification vector, the conflict labels are coordinated, so that the probability of label conflict is reduced, and meanwhile, the checked labels are converted into a silent state, so that redundant data transmission is avoided. The lost tag is also determined by comparing the difference between the actual and expected aggregate signals, effectively identifying the lost tag in parallel.
Description
Technical Field
The invention relates to the technical field of RFID (radio frequency identification) of the Internet of things, in particular to a method and a system for parallelly identifying lost tags in a multi-category RFID system.
Background
Radio Frequency Identification (RFID) is an advanced automatic identification technology that uses electromagnetic fields to identify and monitor RFID-tagged items. Typically, an RFID system consists of at least one RFID reader and thousands of low cost RFID tags. Each item in the system is attached with a unique 96-bit identifier (tag ID). The reader reads the data of the associated articles from the tag to automatically manage the system, greatly simplifying the heavy manual operation. Accordingly, RFID is widely deployed in various large-scale applications such as inventory control, warehouse management, and the like.
In large-scale systems, RFID-tagged items are classified into a number of categories; for example, supermarkets store different kinds of merchandise and libraries store books of different themes. Accurate locating of lost tags is a fundamental task to manage multi-category RFID systems. It is counted that the U.S. retail industry suffers from significant economic losses due to management errors and theft. Therefore, to improve quality of service and prevent potential property damage, it is important to report item loss events in time.
At present, there is no effective solution as to how to effectively identify missing tags in a multi-category system.
Disclosure of Invention
The method and the system for identifying the lost tags in parallel in the multi-category RFID system can identify the lost tags in parallel, completely avoid signal collision among different categories of tags, and effectively relieve the signal collision among the same category of tags.
The invention provides a method for identifying lost tags in parallel in a multi-category RFID system, which comprises the following steps:
Queuing the K unordered tag class identifiers in the RFID system by using a queuing vector V i, and distributing one class layer of one-hot codes for each tag based on the sequence index;
for a candidate tab set C k with a category index of K (K e 1., K) constructs a corresponding sub-notification vector based on its preselected time slots Finally get notification vector/>Broadcasting the notification vector V N to perform tag reply time slot rearrangement;
The reader decodes the independent thermal codes in the reply signals of different types of tags from the aggregate signal of each time slot, and realizes the parallel identification of a plurality of types of tags.
Specifically, the queuing of K unordered tag class identifiers in an RFID system by using a queuing vector V i assigns a class layer of one-hot codes to each tag based on a sequential index includes:
Assuming that K i classes are queued, in any ith round, starting conditions K 1 =k, constructing the queuing vector V i;
the reader passes through a hash function Calculating an index corresponding to the label category, and setting a bit corresponding to the index bit in V i to be 1 when only a single category is mapped to a certain index; if multiple categories map to the same index, the bit position corresponding to the index bit in V i is 0;
The reader broadcasts a query command with a parameter < v i,ri,K-Ki,Vi >, and the system tag calculates the hash function Obtaining an index, and checking a corresponding value in V i through the obtained index; where t CID is the tag class identifier, V i is the length of V i, and r i is the random seed; if the value is 0, continuing to participate in the next iteration; if the value is 1, the number of bits '1' before the index bit in V i is calculated, the number is marked as j, and a single-hot bit string S (K, K-K i +j+1) is generated until all the categories are queued, so that single-hot coding of a category layer is allocated to each tag; where S (K, K) represents a one-hot code of length K, with bit "1" at the kth bit.
In particular, the candidate tab set C k for category index K (K e1, constructing a corresponding sub-notification vector based on its preselected time slotsFinally get notification vector/>And broadcasting the notification vector V N to perform tag reply time slot rearrangement, including:
The reader generates an optimal frame size f 1 and a random seed r 1 using a hash function Calculating a preselected time slot of the candidate tag set C k;
The reader constructs a sub-notification vector For each time slot j, if |a k (j) |=1 or/>Will/>Setting 1, otherwise setting 0, and obtaining/>, through K iterationsWherein a k (j) is a set of tags of category k in slot j, |a k (j) | is the radix of the set, and U k is a set of tags of category k that are verified but in an active state;
the reader broadcasts V N and the system tag selects a sub-notification vector based on its tag class k And check/>Corresponding bit values of (a);
If the bit value is 1, then the number of "1" S before the bit is calculated, denoted by j, and in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th time slot; when the tag responds to the reader, the tag will transition to a silent state;
if the bit value is 0, the tag follows Recalculating the reply time slot and transmitting its one-time-code in the (f 1-f2 +j') th time slot bit of the tag verification step; where f 2 is vector/>In "0".
Specifically, the reader decodes the single thermal codes in reply signals of different types of tags from the aggregate signal of each time slot, so as to realize parallel identification of a plurality of types of tags, and the method comprises the following steps:
Assume that the aggregate signal received in slot j is represented as Selecting a time slot j to reply after the time slot rearrangement step, wherein the class is a label set of k;
For a label of category k, if
Or/>And/>The reader determines/>Tag set loss, will/>Removed from candidate tag set C k, added to missing tag set M k;
Or X and/> The reader determines/>Tag set already exists, will/>Added to the validated tag set U k;
For each time slot j, if all All verified, the reader sends a confirmation instruction to make the tag in a silent state, and meanwhile the reader will/>And deleting from C k and U k, otherwise, sending a continuing identification instruction so that the tag is kept in an activated state.
The invention also provides a system for identifying lost tags in parallel in the multi-category RFID system, which comprises:
the independent heat code generation module is used for queuing K unordered tag class identifiers in the RFID system by using a queuing vector V i, and distributing independent heat codes of a class layer to each tag based on a sequence index;
The time slot rearrangement module is used for constructing a corresponding sub-notification vector based on a preselected time slot of a candidate label set C k (K epsilon [1,.. The K ]) with the category index of K Finally get the notification vectorBroadcasting the notification vector V N to perform tag reply time slot rearrangement;
and the parallel identification module is used for decoding the independent thermal codes in the reply signals of the different types of tags from the aggregate signal of each time slot by the reader so as to realize the parallel identification of a plurality of types of tags.
Specifically, the single thermal encoding generation module includes:
A queuing vector construction unit, configured to construct the queuing vector V i on the assumption that K i classes are queued, and in any ith round, the starting condition K 1 =k; the reader passes through a hash function Calculating an index corresponding to the label category, and setting a bit corresponding to the index bit in V i to be 1 when only a single category is mapped to a certain index; if multiple categories map to the same index, the bit position corresponding to the index bit in V i is 0;
A single thermal code generating unit for broadcasting inquiry command with < v i,ri,K-Ki,Vi > parameter by the reader, and calculating the hash function by the system tag Obtaining an index, and checking a corresponding value in V i through the obtained index; where t CID is the tag class identifier, V i is the length of V i, and r i is the random seed; if the value is 0, continuing to participate in the next iteration; if the value is 1, the number of bits '1' before the index bit in V i is calculated, the number is marked as j, and a single-hot bit string S (K, K-K i +j+1) is generated until all the categories are queued, so that single-hot coding of a category layer is allocated to each tag; where S (K, K) represents a one-hot code of length K, with bit "1" at the kth bit.
Specifically, the time slot reordering module includes:
A preselected time slot calculating unit for generating an optimal frame size f 1 and a random seed r 1 by the reader using a hash function Calculating a preselected time slot of the candidate tag set C k;
A notification vector construction unit for constructing sub-notification vectors by the reader For each time slot j, if |a k (j) |=1 or/>Will/>Setting 1, otherwise setting 0, and obtaining/>, through K iterationsWherein a k (j) is a set of tags of category k in slot j, |a k (j) | is the radix of the set, and U k is a set of tags of category k that are verified but in an active state;
A time slot rearrangement unit for broadcasting V N by the reader, and selecting sub-notification vector by the system tag based on the tag class k And check/>Corresponding bit values of (a); if the bit value is 1, then the number of "1" S before the bit is calculated, denoted by j, and in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th time slot; when the tag responds to the reader, the tag will transition to a silent state; if the bit value is 0, the tag follows/>Recalculating the reply time slot and transmitting its one-time-code in the (f 1-f2 +j') th time slot bit of the tag verification step; where f 2 is vector/>In "0".
In particular, the parallel recognition module is in particular configured to assume that the aggregate signal received in time slot j is represented as Selecting a time slot j to reply after the time slot rearrangement step, wherein the class is a label set of k; for a tag of category k, if/>Or/>And/>The reader determines/>Tag set is lost, willRemoved from candidate tag set C k, added to missing tag set M k; if/>Or X and/>The reader determines/>Tag set already exists, will/>Added to the validated tag set U k; for each time slot j, if all/>All verified, the reader sends a confirmation instruction to make the tag in a silent state, and meanwhile the reader will/>And deleting from C k and U k, otherwise, sending a continuing identification instruction so that the tag is kept in an activated state.
One or more technical schemes provided by the invention have at least the following technical effects or advantages:
Queuing the K unordered tag categories in the RFID system by using a queuing vector V i, and distributing one category layer of single-hot codes for each tag based on the sequence index; for a candidate tab set C k with a category index of K (K e 1., K) constructs a corresponding sub-notification vector based on its preselected time slots Finally get notification vector/>And broadcasting V N to perform tag reply time slot rearrangement; combining the advantages of the single-heat encoding and the manchester encoding, the reader decodes the reply signals of different types of tags from the aggregate signal of each time slot, thereby realizing the parallel identification of a plurality of types of tags, and repeating the steps until the candidate tag set C k is empty, and finally obtaining the lost tag set M k (K epsilon [1, the..K ]). The invention generates category layer single-heat codes for each tag through the category identifier by utilizing the hash function, and completely avoids the conflict between different category tags by utilizing the characteristics of single-heat codes and Manchester codes. The invention further coordinates the conflict label by constructing the notification vector, reduces the probability of label conflict, and simultaneously converts the checked label into a silent state, thereby avoiding the transmission of redundant data. In addition, the invention determines the lost tag set by comparing the difference between the actual aggregate signal and the expected aggregate signal, so that the reader can identify the lost tag in parallel.
Drawings
FIG. 1 is a flowchart of a method for identifying lost tags in parallel in a multi-category RFID system according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating parallel recognition according to an embodiment of the present invention;
FIG. 3 is a flow chart of a parallel recognition phase in an embodiment of the present invention;
FIG. 4 is a graph illustrating η for different frame sizes and class sizes according to an embodiment of the present invention;
FIG. 5 is an experimental graph of ablation with the number of candidate labels maintained at 50000 and K=10:10:90 in an embodiment of the present invention;
fig. 6 is a diagram of time overhead per protocol for k=10:10:90 in an embodiment of the invention;
fig. 7 shows the time for each category to be executed for k=10:10:90 in the embodiment of the present invention;
fig. 8 shows the time overhead per protocol for k=1:1:10 in the embodiment of the invention;
FIG. 9 is a diagram illustrating the time overhead of each protocol when changing the size of each category (subject to uniform distribution) in an embodiment of the present invention;
FIG. 10 is a diagram illustrating the time overhead of each protocol when changing the size of each category (following Gaussian distribution) in an embodiment of the invention;
Fig. 11 is a block diagram of a system for identifying missing tags in parallel in a multi-category RFID system according to an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides a method and a system for identifying lost tags in parallel in a multi-category RFID system, which can identify the lost tags in parallel and completely avoid signal collision between different categories of tags, thereby effectively relieving the signal collision between the same category of tags.
The technical scheme in the embodiment of the invention aims to achieve the technical effects, and the overall thought is as follows:
Problem definition, consider that there are K categories of candidate tag sets C 1,C2,....,CK to monitor, where Class C k is the candidate tag set for k. The reader has a priori ID knowledge of these candidate tags, including tag ID and CID. Some tags may be lost from the RFID system due to theft and management errors. The embodiment of the invention represents M k as a lost tag set in C k (K is more than or equal to 1 and less than or equal to K). L k is a set of tags of category k, also referred to as an RFID system tag set, that reside within the range of the reader's query. Thus, L k∪Mk＝Ck, wherein (1.ltoreq.k.ltoreq.K). The reader does not have ID knowledge of L k (1. Ltoreq.k. Ltoreq.K), i.e., it is unknown which tags already exist. The problems to be solved by the invention are defined as follows: given candidate set C 1,C2,...,CK, an object of an embodiment of the present invention is to design a valid protocol to identify all missing tags M k (1. Ltoreq.k. Ltoreq.K) at a minimum time cost.
Specifically, the method for identifying the lost tag in parallel in the multi-category RFID system provided by the embodiment of the invention comprises the following steps:
Step S1, queuing K unordered tag categories in an RFID system by using a queuing vector V i, and distributing one-hot coding of a category layer to each tag based on the sequence index;
step S2, for candidate tab set C k (K E [ 1.,. The K ]) with category index of K, constructing corresponding sub-notification vector based on the pre-selected time slot thereof Finally get notification vector/>And broadcasting V N to perform tag reply time slot rearrangement;
And S3, combining the advantages of the single-heat coding and the Manchester coding, and decoding reply signals of different types of tags from the aggregate signal of each time slot by the reader so as to realize parallel identification of a plurality of types of tags. And repeating the step S2 and the step S3 until all the candidate tag sets C k are empty, obtaining a lost tag set M k with the category of k, and terminating the method.
Step S1 queues the tag class identifiers in the RFID system through queuing vectors, and allocates one class layer of independent heat codes for each tag, step S2 rearranges the reply time slots of the tags based on the constructed notification vectors, so that the collision probability among the same class of tags is reduced, and step S3 avoids the collision among different class of tags based on the class layer independent heat codes of step S1. Step S1 provides a precondition for the parallel loss verification recognition of different types of tags in step S3, and step S2 provides possibility for the parallel loss verification recognition of the same type of tags in step S3.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Referring to fig. 1, a method for identifying missing tags in parallel in a multi-category RFID system according to an embodiment of the present invention includes:
Step S110: queuing the K unordered tag class identifiers in the RFID system by using a queuing vector V i, and distributing one class layer of one-hot codes for each tag based on the sequence index;
describing this step in detail, by queuing K unordered tag class identifiers in the RFID system using a queuing vector V i, each tag is assigned a class layer of one-hot codes based on a sequential index, including:
Assuming that K i classes are queued, in any ith round, starting conditions K 1 =k, constructing a queuing vector V i;
The reader passes through the hash function Calculating an index corresponding to the label category, and setting a bit corresponding to the index bit in V i to be 1 when only a single category is mapped to a certain index; if multiple categories map to the same index, the bit position corresponding to the index bit in V i is 0;
the reader broadcasts a query command with a parameter < v i,ri,K-Ki,Vi >, and the system tag calculates a hash function Obtaining an index, and checking a corresponding value in V i through the obtained index; where t CID is the tag class identifier, V i is the length of V i, and r i is the random seed; if the value is 0, continuing to participate in the next iteration; if the value is 1, the number of bits '1' before the index bit in V i is calculated, the number is marked as j, and a single-hot bit string S (K, K-K i +j+1) is generated until all the categories are queued, so that single-hot coding of a category layer is allocated to each tag; where S (K, K) represents a one-hot code of length K, with bit "1" at the kth bit.
Step S120: for a candidate tab set C k with a category index of K (K e 1., K) constructs a corresponding sub-notification vector based on its preselected time slotsFinally get notification vector/>And broadcasting a notification vector V N to perform tag reply time slot rearrangement;
to specifically describe this step, for a candidate tab set C k (K e [1,., K ]) with a category index of K, a corresponding sub-notification vector is constructed based on its preselected time slots Finally get the notification vectorAnd broadcast notification vector V N for tag reply time slot reordering, comprising:
The reader generates an optimal frame size f 1 and a random seed r 1 using a hash function Calculating a preselected time slot of the candidate tag set C k;
Reader constructs sub-notification vectors For each time slot j, if |a k (j) |=1 or/>Will beSetting 1, otherwise setting 0, and obtaining/>, through K iterationsWherein a k (j) is a set of tags of category k in slot j, |a k (j) | is the radix of the set, and U k is a set of tags of category k that are verified but in an active state;
The reader broadcasts V N and the system tag selects a sub-notification vector based on its tag class k And check/>Corresponding bit values of (a);
If the bit value is 1, then the number of "1" S before the bit is counted, denoted by j, and in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th slot; when the tag responds to the reader, the tag will transition to a silent state;
If the bit value is 0, the tag follows Recalculating the reply time slot and transmitting its one-time-code in the (f 1-f2 +j') th time slot bit of the tag verification step; where f 2 is vector/>In "0".
Step S130: the reader decodes the independent thermal codes in the reply signals of different types of tags from the aggregate signal of each time slot, and realizes the parallel identification of a plurality of types of tags.
To specifically explain this step, the reader decodes the single thermal codes in the reply signals of different class labels from the aggregate signal of each time slot, so as to realize parallel identification of multiple class labels, including:
Assume that the aggregate signal received in slot j is represented as Selecting a time slot j to reply after the time slot rearrangement step, wherein the class is a label set of k;
For a label of category k, if
Or/>And/>The reader judges/>Tag set loss, will/>Removed from candidate tag set C k, added to missing tag set M k;
Or X and/> The reader judges/>Tag set already exists, will/>Added to the validated tag set U k;
For each time slot j, if all All verified, the reader sends a confirmation instruction to make the tag in a silent state, and meanwhile the reader will/>And deleting from C k and U k, otherwise, sending a continuing identification instruction so that the tag is kept in an activated state.
The method for identifying lost tags in parallel in the multi-category RFID system provided by the embodiment of the present invention is further described below:
The method (PaMI) for identifying the lost tag in parallel in the multi-category RFID system provided by the embodiment of the invention comprises two stages: a category queuing stage and a parallel recognition stage. In the category queuing stage PaMI broadcasts a lightweight queuing vector, all K categories are ordered. In the parallel identification phase, the reader verifies whether the tag is lost by means of an aggregate signal collected from the system tag. To further reduce the time overhead required for PaMI execution, the parameters involved in the two phases PaMI are optimized separately, as follows:
1) Class queuing stage
The category queuing phase includes multiple iterations, in any ith round, the reader uses the queuing vector V i for category queuing. Suppose there is a K i class to be queued at the beginning of the ith round, where the first round K 1 =k
Constructing a queuing vector V i: the reader first performs K i categoriesAnd calculating to obtain an index corresponding to the category, wherein CIDs are tag category identifiers, r i is a random seed, and V i is the length of V i. The queuing vector construction rules are: if only a single category hashes to a certain index, the value of the index bit in V i is set to 1, otherwise, set to 0;
After constructing V i, the reader broadcasts a Query command containing the parameter < V i,ri,K-Ki,Vi). After each system tag receives the Query command, the tag firstly performs the following steps And obtaining the corresponding index of the category. Where t CID is the CID of the tag. The index bit in tag check V i: if the value is 0, the tag continues to participate in the next round of queuing process; if the value is 1, the tag will count the number of bits "1" before its index bit.
Assuming j "1' S" preceding its index bit, the tag concludes that j categories have been queued in the round, and the tag injects a single hot bit string S (K, K-K i +j+1) into its memory, where K-K i is the number of categories that have been queued prior to the ith round. The tag generates a one-time hot code of the tag from tuple < K, K-K i +j+1>, and the reader does not need to broadcast the string. Repeating the above process until all the categories are queued.
After the class queuing phase, signal collision between different class labels can be avoided because a bit tracking technique based on Manchester encoding is used, in Manchester encoding, a jump exists in the middle of each bit, the jump in the middle of each bit is used as a clock signal and a data signal, the lower bits to the higher bits represent the value of '1', and the upper bits to the lower bits represent the value of '0'. When multiple tags transmit the same data at the same time, the reader can successfully retrieve its data, otherwise a one bit collision signal "X" is detected, based on which any change in the tag can be found from the aggregate signal differences received by the reader if multiple tags reply with different one-time codes S (K, K) in the same time slot. Thus, different classes of tags reply with different single thermal codes, and the reader is able to avoid signal interference between them.
By way of illustration, in fig. 2 the dashed line represents a missing tag, the solid line represents a presence tag, and fig. 2 (a) shows an example of parallel missing tag identification from different classes of tags, where tag 1, tag 2 and tag 3 are expected to transmit the one-hot codes "100", "010" and "001" simultaneously. The actual received aggregate signalIs "XOX". Thus, it is determined that tag 2 is not present, and tag 1 and tag 3 are present.
An example of parallel missing tag identification from the same class for tag 1, tag 2 is shown in fig. 2 (b), where the reader predicts that tag 1, tag 2, tag 3 respond to data "100", "001" in the same time slot. Aggregate signal actually received by readerIs "XOX". In this case, the reader verifies that tag 3 is present because/>Although/>In this case, the reader may determine that at least one of these tags is present in the system, but since it is not possible to distinguish which of tag 1 and tag 2 is responding, the reader is not able to determine which of them are present.
The class queuing stage has the significance that different independent thermal codes are generated for different classes of labels by using lightweight queuing vectors, and theoretical premises are provided for realizing parallel identification of multiple classes of labels.
2) Parallel recognition phase
The parallel recognition stage is divided into a plurality of rounds, and each round is divided into three steps: (a) A notification vector construction step, wherein the reader constructs a notification vector V N according to the candidate tag preselect time slot state; (b) A time slot rearrangement step, wherein the reader broadcasts V N to reorganize the response time slots of the candidate tags; (c) And a tag verification step, wherein the tag responds to the self-independent thermal coding in the rearranged time slot, and the reader verifies the existence or the loss of the candidate tag according to the received signal.
A. notification vector construction
A notification vector construction step: by usingAnd the notification vector corresponding to the k label is represented. To facilitate understanding, first set forth/>Based on all/>V N is generated, where K has a value in the range of 1 to K.
The reader stores three tag sets for each class k tag: candidate tag set C k stores tags to be verified, U k stores tags that are verified but still active, and M k stores identified lost tags. Initial condition, C k, contains all candidate tags for category k, U k and M k are empty sets.
The reader first generates an optimal frame size f 1 and a random seed r 1, and then generates a hash functionA preselected time slot for candidate tag set C k is calculated. Let a k (j) represent the labelset for preselected slot j in C k. Thus, there areAnd/>Where || represents the cardinality of the collection. Reader construction according to A k (j)The following are provided: /(I)Contains f 1 bits, each bit corresponding to a slot. For each time slot j, if |a k (j) |=1 orThen/>The value of (1) is set to 1, otherwise set to 0. Finally, a notification vector/>, with a label class of k, is obtainedRepeating the process to obtain K sub-notification vectors/>The subvectors are concatenated to form V N.
B. Time slot rearrangement
After V N is constructed, the reader broadcasts a Query command with a parameter < f 1,r1,VN,r2 >. The system tag receives the Query command as follows: each label belonging to k classes is first based onCalculate the index and check it at/>And a value of the index bit, where t ID is the tag's ID.
If the corresponding bit value is 1, the tag records the number of bits "1" preceding its index bit. If the number of bits "1" preceding the index bit of the tag is j, then in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th slot. When the tag responds to the reader, the tag will be in a silent state.
If the corresponding bit value is 0, the tag followsRecalculating the reply time slot and responding to its one-hot code in the (f 1-f2 +j') time slot bits of the tag verification step, where f 2 is vector/>The total number of "0" bits.
After the slot rearrangement step is completed, rearrangement of all candidate tags has been completed. By usingRepresenting a set of tags that respond in slot j after a slot bit reordering step, wherein/>Is/>Which contains only the tag set of category k. Since the reader has ID knowledge of all candidate tags and all hash parameters, the reader storesAnd collecting, wherein j is more than or equal to 1 and less than or equal to f 1, and K is more than or equal to 1 and less than or equal to K.
C. Label verification
All tags from different categories respond to their own one-hot codes in rearranged timeslots. Assume that the aggregate signal received per slot j isThe reader verifies the tag as follows: for each class K (1.ltoreq.k.ltoreq.K),
1. If it isOr/>And/>The reader concludes/>Tag set loss, will/>Removed from collection C k, added to M k;
2. If it is Or X and/>The reader judges/>Tag set already exists, will/>Added to U k;
3. If it is Or X and/>The reader cannot determine/>Whether present or lost;
For each time slot j, if all tag sets are verified The reader sends an "ACK" command to put these tags in silence and will/>Removed from C k and U k, otherwise a "NAK" command is sent to keep the tags active. Furthermore, for each tag in category k, if it is in/>If the index bit value of (1) is 1, the tag automatically enters a silent state after the one-hot code response returns to the reader, and the tag is deleted in C k and u k.
If all candidate sets C k (1.ltoreq.k.ltoreq.K) are empty, the reader will terminate PaMI protocol and set M k will be the final loss result for category K.
For example, as shown in FIG. 3, FIG. 3 details the workflow of the parallel recognition stage. There are a total of 4 categories and 13 candidate tags. The relationship between the tag and the category is shown in the upper left corner of fig. 3. In the figure, the dashed arrow indicates the mapping relationship between the candidate tag and the slot, and the solid arrow indicates the actual communication between the reader and the system tag.
In the notification vector construction step, the frame length size f 1 is set to 4. First, the reader is based onPredicting the preselected time slot of the candidate label, and constructing a notification vector V N according to the state of the preselected time slot. For better understanding, the present embodiment interprets notification vectors on a class-by-class basis. Each sub-notification vector/>Is 4, where each bit represents a time slot. For class 1 tags, tags t 1 and t 2, both preselected slots are slot 1, then will/>The value of bit 1 in (1) is set to 0, i.e. >Since only tag t 3 preselected time slot 2, reader settings/>For time slot 3 and time slot 4, no class 1 tag preselects the time slot, so will/>And/>Set to 0. Thus, the subvector/>Is "0100". Also similarly, the present embodiment obtains the remaining sub-notification vectors/> Finally, reader construction/>
In the slot rearrangement step, the reader broadcasts a Query command with < f 1,r1,VN,r2 >. After receiving the Query command, the system tag performs hash calculation to obtain an index and checks the value of the index bit in V N. For a class 1 tag, the tag t 1 first computes the hash valueTag t 1 check/>The first bit of (a) has a bit value of 0, which means that a collision slot is preselected by the tag t 1. Thus, it recalculates a new hash value,/>Wherein f 2 represents/>The number of "0" bits in (i.e., f 2 =3).
Since f 1-f2 +j' =2, the tag t 1 is rearranged to the 2 nd slot. Similarly, tag t 2 rearranges it to the 3 rd slot. Tag t 3 passes through the hashExamination/>The index bit value is detected as 1 for the 2 nd bit in (a). Thus, t 3 is calculated at/>Since the elements preceding the index bit are all "0", t 3 rearranges them to the 1 st slot. Similarly, the class 2 tag t 4、t5、t6 is rearranged into time slot 1, time slot 2, time slot 4, respectively; the class 3 tag t 7、t8、t9 is rearranged to slot 1, slot 4, slot 3. When/>When the label is '0000', all labels belonging to the category 4 recalculate the new hash value,/>The tags t 10、t11、t12、t13 are rearranged to slots No. 1, no. 3, no. 4, and No. 4, respectively.
Finally, the mapping between the candidate tags and their rearranged slots is as in the tag verification step shown in fig. 3 (c), in the step of the reader identifying multiple categories of lost tags simultaneously. The signal received by the reader in slot 1 is "OXXX", thus acknowledging that tag t 3 is lost and that tag t 4,t7,t10 already exists. Similarly, when the aggregate signal for slot 2 is received as "XX00", the reader verifies that { t 1,t5 } already exists. In time slot 3, since the received aggregate signal is "X00X", the reader verifies that { t 2,t11 } is present, verifying that t 9 is lost. In time slot 4, the aggregate signal received by the reader is "0010", indicating that tag t 8 is present and tag { t 7,t12,t13 } is absent. Since all candidate tags have been validated, the process terminates PaMI.
3) Parameter optimization
Category queuing: the parameter v i length is optimized to minimize the average time overhead required for queuing a class by optimizing the size of v i. At any ith wheel. Let K i be the number of categories to be queued. In the ith round, the reader only broadcasts one v i -bit queuing vector. Thus, the time overhead is calculated as:
Where t λ denotes a time period of broadcasting the frame initialization parameter, and t id denotes a time required for broadcasting the 96-bit ID.
The probability that a slot is mapped by only one class is denoted by P sc, i.e
The expected number of categories that the ith round can queue is denoted by N v, calculated as:
Thus, the average time overhead for each category to complete the queuing process is calculated as:
to minimize T and i, we let us let
/>
And obtain
Solving the above equation to obtain the productMinimum optimal/>I.e.
Finally, roundingAnd used for execution.
Parallel identification phase: optimizing the size of the parameter f 1 frame length optimizes the size of f 1 to minimize the time overhead of verifying a candidate tag. In the ith round, the total time overhead is first calculated, denoted as T round, and the expected number of tags that can be verified, denoted as N round, resulting in η=n round/Tround.
The parallel recognition phase comprises three steps. The notification vector construction step is performed only at the reader side without any transmission overhead. In the slot reordering step, the reader broadcasts a vector V N of total f 1 ×k bits. In the tag verification step, the reader performs a time frame containing f 1 slots. Thus, the total time overhead is calculated as:
Where t kb is the time overhead for transmitting k bits of data in a slot, and t λ is the time it takes for the reader to broadcast the frame initialization parameters. To obtain N round, the expected number of tags that each class k can verify is first calculated, expressed as Finally obtain N round
For each tag of class k, the verification is performed in three cases: (1) its preselected slot bits are single slots; (2) its rearranged slot bits are single slot; (3) The rearranged slot bits are signal collision slots, but the received dataIs 0, where/>Slot bit index rearranged for the tag.
For each class k, candidate tags |c k | are randomly pre-selected in f 1 slots. Let p s be the probability that the preselected slot is a single slot. The probability that the value of p s is equal to the time slot preselected by only one tag can be expressed as:
thus, the expected number of verifiable tags in case (1) is calculated as:
Tags that do not pre-select a single slot will be coordinated by a random seed r 2. Let p' s be the probability of a rearranged slot being a single slot, calculated as:
Where C k|-n1 represents the expected number of pre-selected conflicting slot tags. Thus, (2) the expected number of verifiable tags in the case is calculated as:
The probability that the rearranged slot is the a signal collision slot (a.gtoreq.2) is denoted by p a, calculated as:
for any time slot j, let P r0 denote the received data Probability of 0. The calculation is as follows:
Wherein L k is the number of system labels of class k, and the value of L k can be obtained by referring to the existing estimation technology. Thus, in the case of (3), the expected number of tags can be verified as:
Where z= |c k|-n1. Solving the combined equation to obtain the total number of verifiable labels of each class k as follows:
/>
Finally, the eta value is calculated by combining the equation,
Our goal is to maximize η to optimize PaMI time overhead. By solving forWe have obtained the optimum f 1, thus achieving the best performance of PaMI. Fig. 4 plots the change in η value for two categories, one of which is set to 1000 in number and the ratio of missing tags is 0.5. The x-axis represents the size of f 1, the y-axis represents the size of other classes, the vertical axis represents the η value, and the maximum value of η is 2.338 when the optimal size of f 1 is 1007 when the y value is 1000.
Experimental simulation and analysis
The reference method comprises the following steps: to fully evaluate the performance of model PaMI, paMI is compared to the latest protocol IIP, SFMTI, CLS, CR-MTI.
Simulation setup
The emulation parameters are set in accordance with the EPC C1G2 standard, which defines a transmission rate between the reader and the tag as 40Kbps. Thus, the duration of exchanging 1-bit data is 0.025ms. Any two consecutive transmission intervals are 0.4ms. The time overhead of broadcasting the initialization frame parameter is set to t λ＝2tid. Thus, we get t kb＝(0.4+0.025K)ms,tid＝2.8,tλ =5.6m.
In the simulation experiment environment, the experiment set the default number of categories to 50 and the default missing tag ratio to 0.5. Two different category size distributions were experimentally studied: uniform distribution (i.e., balanced class) and gaussian distribution (i.e., unbalanced class). To reduce randomness, each simulation experiment was run 100 times and the average results of the independent experiments were reported.
Analysis of experimental results
Ablation analysis
As shown in fig. 5, the black bars represent the total time overhead of PaMI. The white bars represent the time overhead of PaMI-NoVector (i.e., paMI without using notification vectors). Experiments have shown that PaMI performs significantly better than PaMI-NoVector, which verifies the validity of the notification vector. Furthermore, we observe that as the K value increases, the time overhead of PaMI decreases, but PaMI-NoVector time overhead decreases first and then remains unchanged. This is because PaMI can verify more tags in one slot as K increases, thereby shortening execution time.
Category number impact analysis
Fig. 6 varies the number of categories from 10 to 90 to study the scalability of each protocol. In this set of simulations, the size of each class was fixed at 1000 and the missing tag ratio was 0.5. Experiments have shown that PaMI increases at a much slower rate than the single class missing tag identification protocol, which increases in execution time linearly with K. When K increases from 10 to 90, paMI execution times increase by only about 5 times, which suggests that PaMI has good scalability for multi-category loss identification scenarios. As can be seen from fig. 4, paMI performance is always optimal. When k=90, the execution time is reduced by a factor of 4.69 compared to the existing best protocol CR-MTI.
Fig. 7 shows the time overhead for each category PaMI in fig. 6. Experimental results indicate that the time overhead per class PaMI decreases with increasing K. Because the more categories, the more tags PaMI can verify in a slot, the greater the parallelism efficiency and the shorter the execution time. In contrast, the time overhead of identifying each class of the missing tag protocol from class to class is unchanged.
Fig. 8 is a view of the time efficiency of each protocol per increment of 1 class when the number of classes is 1 to 10. Experiments show that when the category number is 1, the method PaMI is worse than SFMTI and CR-MTI protocols. When the number of categories is greater than 2, paMI always performs better than other protocols. Since there is only one class in the system, the bit tracking technique fails. In this case PaMI will degrade to a serial identification protocol, resulting in time inefficiency. And when K >2 PaMI is always better than the existing best protocol CR-MTI for the following reasons: the CR-MTI coordinates collisions of multiple tags based on random hashing, and can identify multiple colliding tags if and only if each tag has a unique new hash value. However, the randomness of the hash reduces the probability of signal collision coordination. To further increase the probability of successful coordination of conflicting signals, the CR-MTI sets the virtual frame size to be much larger than the number of signal conflicting tags, resulting in wasted time. In contrast PaMI allows each tag to respond to the reader with its own unique thermal code, completely avoiding signal interference from different categories. Meanwhile, the length of the single thermal code is equal to the number of label categories, so that unnecessary time expenditure is reduced.
Missing tag proportion impact analysis
As shown in FIG. 8, IIP, SFMTI, CLS, CR-MTI, paMI take 58.5s, 33.5s, 22.9s, 21.9s, 3.77s, respectively, when the lost tag rate is 0.9. PaMI reduces the time by a factor of 5.81 compared with the existing optimal protocol CR-MTI.
Class size impact analysis
Fig. 9 and 10 plot the total execution time of the protocol when the class sizes obey a uniform distribution and a gaussian distribution, respectively. Wherein fig. 9 varies the average size of each category from 200 to 2000 to investigate its effect on all protocol performances. Experiments have shown that the time overhead of all protocols increases with the average class size. The reason is that an increase in category size brings more candidate tags. Thus, all protocols require configuration of larger frame sizes or longer vectors to accommodate candidate tags, increasing time overhead. Comprehensive experiments show that PaMI significantly reduces the execution time required for missing tag identification compared to existing protocols.
Referring to fig. 11, the embodiment of the present invention further provides a system for identifying missing tags in parallel in a multi-category RFID system, including:
A single thermal code generation module 100, configured to queue K unordered tag class identifiers in the RFID system by using a queuing vector V i, and allocate a single thermal code of a class layer to each tag based on a sequential index;
Specifically, the single thermal encoding generation module 100 includes:
A queuing vector construction unit, configured to construct a queuing vector V i on the assumption that K i classes are queued, where in any ith round, the starting condition K 1 =k; the reader passes through the hash function Calculating an index corresponding to the label category, and setting a bit corresponding to the index bit in V i to be 1 when only a single category is mapped to a certain index; if multiple categories map to the same index, the bit position corresponding to the index bit in V i is 0;
the single-heat code generating unit is used for broadcasting inquiry command with < v i,ri,K-Ki,Vi > parameter by the reader, and the system tag calculates the hash function Obtaining an index, and checking a corresponding value in V i through the obtained index; where t CID is the tag class identifier, V i is the length of V i, and r i is the random seed; if the value is 0, continuing to participate in the next iteration; if the value is 1, the number of bits '1' before the index bit in V i is calculated, the number is marked as j, and a single-hot bit string S (K, K-K i +j+1) is generated until all the categories are queued, so that single-hot coding of a category layer is allocated to each tag; where S (K, K) represents a one-hot code of length K, with bit "1" at the kth bit. /(I)
The slot rearrangement module 200 is configured to, for a candidate tab set C k (K e [1,., K ]) with a category index of K, construct a corresponding sub-notification vector based on its preselected slotsFinally get the notification vectorAnd broadcasting a notification vector V N to perform tag reply time slot rearrangement;
Specifically, the slot reordering module 200 includes:
A preselected time slot calculating unit for generating optimal frame size f 1 and random seed r 1 by using hash function Calculating a preselected time slot of the candidate tag set C k;
a notification vector construction unit for constructing sub-notification vectors by the reader For each time slot j, if |a k (j) |=1 or/>Will/>Setting 1, otherwise setting 0, and obtaining/>, through K iterationsWherein a k (j) is a set of tags of category k in slot j, |a k (j) | is the radix of the set, and U k is a set of tags of category k that are verified but in an active state;
A time slot rearrangement unit for broadcasting V N by the reader, and selecting sub-notification vector by the system tag based on the tag class k And check/>Corresponding bit values of (a); if the bit value is 1, then the number of "1" S before the bit is counted, denoted by j, and in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th slot; when the tag responds to the reader, the tag will transition to a silent state; if the bit value is 0, the tag follows/>Recalculating the reply time slot and transmitting its one-time-code in the (f 1-f2 +j') th time slot bit of the tag verification step; where f 2 is vector/>In "0".
The parallel identification module 300 is configured to decode the independent thermal codes in the reply signals of different types of tags from the aggregate signal of each time slot by using the reader, so as to implement parallel identification of multiple types of tags.
In particular, the parallel identification module 300 is in particular configured to assume that the aggregate signal received in the time slot j is represented as Selecting a time slot j to reply after the time slot rearrangement step, wherein the class is a label set of k; for a tag of category k, if/>Or/>And/>The reader judges/>Tag set loss, will/>Removed from candidate tag set C k, added to missing tag set M k; if/>Or X and/>The reader judges/>Tag set already exists, will/>Added to the validated tag set U k; for each time slot j, if allAll verified, the reader sends a confirmation instruction to make the tag in a silent state, and meanwhile the reader will/>And deleting from C k and U k, otherwise, sending a continuing identification instruction so that the tag is kept in an activated state.
In summary, the embodiment of the invention has the following beneficial effects:
(1) The embodiment of the invention can simultaneously identify the lost tags of a plurality of categories, because in the category queuing stage, the tags respond to the reader by customized single thermal coding, and Manchester coding is introduced in tag signal decoding, theoretical premise is provided for completely avoiding signal interference from different categories, so that the lost tags can be identified in parallel.
(2) The embodiment of the invention can improve the recognition rate of the lost tag because the notification vector is used to further coordinate the collision of the tag signals from the same category in the time slot rearrangement stage, thereby improving the tag verification efficiency, simultaneously being capable of quickly confirming the verified tag and avoiding the data redundancy.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Embodiments of the present invention are not described in detail and are well known to those skilled in the art. Finally, it is noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered by the scope of the claims of the present invention.
Claims (8)
1. A method for identifying missing tags in parallel in a multi-category RFID system, comprising:
Queuing the K unordered tag class identifiers in the RFID system by using a queuing vector V i, and distributing one class layer of one-hot codes for each tag based on the sequence index;
for a candidate tab set C k with a category index of K (K e 1., K) constructs a corresponding sub-notification vector based on its preselected time slots Finally get notification vector/>Broadcasting the notification vector V N to perform tag reply time slot rearrangement;
The reader decodes the independent thermal codes in the reply signals of different types of tags from the aggregate signal of each time slot, and realizes the parallel identification of a plurality of types of tags.
2. The method for concurrently identifying missing tags in a multi-category RFID system of claim 1 wherein said queuing K unordered tag category identifiers in the RFID system by using a queuing vector V i assigns each tag a category layer of one-hot codes based on a sequential index, comprising:
Assuming that K i classes are queued, in any ith round, starting conditions K 1 =k, constructing the queuing vector V i;
the reader passes through a hash function Calculating an index corresponding to the label category, and setting a bit corresponding to the index bit in V i to be 1 when only a single category is mapped to a certain index; if multiple categories map to the same index, the bit position corresponding to the index bit in V i is 0;
The reader broadcasts a query command with a parameter < v i,ri,K-Ki,Vi >, and the system tag calculates the hash function Obtaining an index, and checking a corresponding value in V i through the obtained index; where t CID is the tag class identifier, V i is the length of V i, and r i is the random seed; if the value is 0, continuing to participate in the next iteration; if the value is 1, the number of bits '1' before the index bit in V i is calculated, the number is marked as j, and a single-hot bit string S (K, K-K i +j+1) is generated until all the categories are queued, so that single-hot coding of a category layer is allocated to each tag; where S (K, K) represents a one-hot code of length K, with bit "1" at the kth bit.
3. The method for concurrently identifying missing tags in a multi-category RFID system of claim 1, the candidate tab set C k for category index K (K e 1, constructing a corresponding sub-notification vector based on its preselected time slotsFinally get notification vector/>And broadcasting the notification vector V N to perform tag reply time slot rearrangement, including:
The reader generates an optimal frame size f 1 and a random seed r 1 using a hash function Calculating a preselected time slot of the candidate tag set C k;
The reader constructs a sub-notification vector For each time slot j, if |a k (j) |=1 or/>Will beSetting 1, otherwise setting 0, and obtaining/>, through K iterationsWherein a k (j) is a set of tags of category k in slot j, |a k (j) | is the radix of the set, and U k is a set of tags of category k that are verified but in an active state;
the reader broadcasts V N and the system tag selects a sub-notification vector based on its tag class k And check/>Corresponding bit values of (a);
If the bit value is 1, then the number of "1" S before the bit is calculated, denoted by j, and in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th time slot; when the tag responds to the reader, the tag will transition to a silent state;
if the bit value is 0, the tag follows Recalculating the reply time slot and transmitting its one-time-code in the (f 1-f2 +j') th time slot bit of the tag verification step; where f 2 is vector/>In "0".
4. The method for parallel identification of missing tags in a multi-category RFID system of claim 1, wherein the reader decodes the one-time thermal codes in the reply signals of different category tags from the aggregate signal of each time slot, enabling parallel identification of a plurality of category tags, comprising:
Assume that the aggregate signal received in slot j is represented as Selecting a time slot j to reply after the time slot rearrangement step, wherein the class is a label set of k;
For a label of category k, if
Or/>And/>The reader determines/>Tag set loss, will/>Removed from candidate tag set C k, added to missing tag set M k;
Or X and/> The reader determines/>Tag set already exists, will/>Added to the validated tag set U k;
For each time slot j, if all All verified, the reader sends a confirmation instruction to make the tag in a silent state, and meanwhile the reader will/>And deleting from C k and U k, otherwise, sending a continuing identification instruction so that the tag is kept in an activated state.
5. A system for identifying missing tags in parallel in a multi-category RFID system, comprising:
the independent heat code generation module is used for queuing K unordered tag class identifiers in the RFID system by using a queuing vector V i, and distributing independent heat codes of a class layer to each tag based on a sequence index;
the time slot rearrangement module is used for constructing a corresponding sub-notification vector based on a preselected time slot of a candidate tag set C k (j epsilon [1,.. The K ]) with the category index of K Finally get notification vector/>Broadcasting the notification vector V N to perform tag reply time slot rearrangement;
and the parallel identification module is used for decoding the independent thermal codes in the reply signals of the different types of tags from the aggregate signal of each time slot by the reader so as to realize the parallel identification of a plurality of types of tags.
6. The system for parallel identification of missing tags in a multi-category RFID system of claim 5, wherein the single thermal code generation module includes:
A queuing vector construction unit, configured to construct the queuing vector V i on the assumption that K i classes are queued, and in any ith round, the starting condition K 1 =k; the reader passes through a hash function Calculating an index corresponding to the label category, and setting a bit corresponding to the index bit in V i to be 1 when only a single category is mapped to a certain index; if multiple categories map to the same index, the bit position corresponding to the index bit in V i is 0;
A single thermal code generating unit for broadcasting inquiry command with < v i,ri,K-Ki,Vi > parameter by the reader, and calculating the hash function by the system tag Obtaining an index, and checking a corresponding value in V i through the obtained index; where t CID is the tag class identifier, V i is the length of V i, and r i is the random seed; if the value is 0, continuing to participate in the next iteration; if the value is 1, the number of bits '1' before the index bit in V i is calculated, the number is marked as j, and a single-hot bit string S (K, K-K i +j+1) is generated until all the categories are queued, so that single-hot coding of a category layer is allocated to each tag; where S (K, K) represents a one-hot code of length K, with bit "1" at the kth bit.
7. The system for concurrently identifying missing tags in a multi-category RFID system of claim 5 wherein the slot rearrangement module includes:
A preselected time slot calculating unit for generating an optimal frame size f 1 and a random seed r 1 by the reader using a hash function Calculating a preselected time slot of the candidate tag set C k;
A notification vector construction unit for constructing sub-notification vectors by the reader For each time slot j, if |a k (j) |=1 or/>Will/>Setting 1, otherwise setting 0, and obtaining/>, through K iterationsWherein a k (j) is a set of tags of category j in slot h, |a k (j) | is the radix of the set, and U k is a set of tags of category k that are verified but in an active state;
A time slot rearrangement unit for broadcasting V N by the reader, and selecting sub-notification vector by the system tag based on the tag class k And check/>Corresponding bit values of (a); if the bit value is 1, then the number of "1" S before the bit is calculated, denoted by j, and in the subsequent tag verification step the tag will respond to its one-hot code S (K, K) in the (j+1) th time slot; when the tag responds to the reader, the tag will transition to a silent state; if the bit value is 0, the tag follows/>Recalculating the reply time slot and transmitting its one-time-code in the (f 1-f2+j′) th time slot bit of the tag verification step; where f 2 is vector/>In "0".
8. The system for parallel identification of missing tags in a multi-category RFID system as recited in claim 5 wherein said parallel identification module is specifically configured to assume that the aggregate signal received in slot j is represented asSelecting a time slot j to reply after the time slot rearrangement step, wherein the class is a label set of k; for a label of category k, ifOr/>And/>The reader determines/>Tag set loss, will/>Removed from candidate tag set C k, added to missing tag set M k; if/>Or X and/>The reader determinesTag set already exists, will/>Added to the validated tag set U k; for each time slot j, if allAll are verified, the reader sends a confirmation instruction to make the tag be in a silent state, and at the same timeAnd deleting from C k and U k, otherwise, sending a continuing identification instruction so that the tag is kept in an activated state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410100948.8A CN117938210A (en) | 2024-01-23 | 2024-01-23 | Method and system for identifying lost tags in parallel in multi-category RFID system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410100948.8A CN117938210A (en) | 2024-01-23 | 2024-01-23 | Method and system for identifying lost tags in parallel in multi-category RFID system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117938210A true CN117938210A (en) | 2024-04-26 |
Family
ID=90751805
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410100948.8A Pending CN117938210A (en) | 2024-01-23 | 2024-01-23 | Method and system for identifying lost tags in parallel in multi-category RFID system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117938210A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070078664A (en) * | 2006-01-28 | 2007-08-01 | 변종섭 | Method and system of searching for lost article using rfid tag and wireless system |
| CN103366143A (en) * | 2013-07-03 | 2013-10-23 | 清华大学 | Methods for quickly identifying lost tag in large-scale radio-frequency identification system |
| CN112966534A (en) * | 2020-11-23 | 2021-06-15 | 扬州大学 | Multi-classification-based lost label detection method in RFID system |
| CN116451714A (en) * | 2022-01-06 | 2023-07-18 | 山东省天然气管道有限责任公司 | Multi-packet loss label detection method for expected single time slot |
| CN117113804A (en) * | 2023-07-06 | 2023-11-24 | 南京航空航天大学 | RFID lost tag identification method based on tree splitting |
| US20230385569A1 (en) * | 2021-11-08 | 2023-11-30 | Nanjing University Of Information Science And Technology | Multi-tag concurrent identification method and system for query tree based on feature groups |
-
2024
- 2024-01-23 CN CN202410100948.8A patent/CN117938210A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070078664A (en) * | 2006-01-28 | 2007-08-01 | 변종섭 | Method and system of searching for lost article using rfid tag and wireless system |
| CN103366143A (en) * | 2013-07-03 | 2013-10-23 | 清华大学 | Methods for quickly identifying lost tag in large-scale radio-frequency identification system |
| CN112966534A (en) * | 2020-11-23 | 2021-06-15 | 扬州大学 | Multi-classification-based lost label detection method in RFID system |
| US20230385569A1 (en) * | 2021-11-08 | 2023-11-30 | Nanjing University Of Information Science And Technology | Multi-tag concurrent identification method and system for query tree based on feature groups |
| CN116451714A (en) * | 2022-01-06 | 2023-07-18 | 山东省天然气管道有限责任公司 | Multi-packet loss label detection method for expected single time slot |
| CN117113804A (en) * | 2023-07-06 | 2023-11-24 | 南京航空航天大学 | RFID lost tag identification method based on tree splitting |
Non-Patent Citations (1)
| Title |
|---|
| 梁雪萍;马存庆;梁颖升;: "一种快速检测RFID丢失标签的方法", 计算机工程, no. 01, 15 January 2016 (2016-01-15) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Efficient information collection protocols for sensor-augmented RFID networks | |
| CN110807339B (en) | Lost label detection method of RFID system | |
| Liu et al. | Efficient polling-based information collection in RFID systems | |
| CN110378157B (en) | Multi-seed-based large-scale RFID system probabilistic clone attack identification method | |
| CN109446857A (en) | Label anti-collision method and system based on quick dynamic frame slot | |
| Su et al. | An efficient missing tag identification approach in RFID collisions | |
| CN103020569A (en) | Radio-frequency identification multi-label anti-collision method | |
| US20080122583A1 (en) | System and method for RFID tag communication | |
| CN107895130B (en) | Self-adaptive multi-branch tree anti-collision method based on collision tree | |
| He et al. | A fast RFID tag identification algorithm based on counter and stack | |
| Samsami et al. | Novel RFID anti-collision algorithm based on the Monte–Carlo query tree search | |
| CN117938210A (en) | Method and system for identifying lost tags in parallel in multi-category RFID system | |
| Zhu et al. | Dynamic grouping in RFID systems | |
| CN103679096A (en) | Optimized RFID anti-collision method | |
| Shakiba et al. | Fitted dynamic framed slotted ALOHA anti-collision algorithm in RFID systems | |
| CN117113804A (en) | RFID lost tag identification method based on tree splitting | |
| Wang et al. | Missing tag identification in open RFID systems | |
| Pupunwiwat et al. | Unified Q-ary tree for RFID tag anti-collision resolution | |
| Wang et al. | Time-efficient missing tag identification in an open RFID system | |
| Kalache et al. | Performances comparison of RFID anti-collision algorithms | |
| Kang et al. | RCSMA: Receiver-based carrier sense multiple access in UHF RFID systems | |
| Chu et al. | Efficient and robust missing key tag identification for large-scale RFID systems | |
| Guo et al. | Efficient collision-slot utilization for missing tags identification in RFID system | |
| KR100921862B1 (en) | Method For Generating RFID Tag ID For Collision Avoidance And RFID Tag Having ID For Collision Avoidance | |
| CN111860009B (en) | Enhanced binary grouping RFID anti-collision method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |