RFID indoor positioning method and device based on space-time information, medium and equipment

Abstract

Embodiments of the present application provide a RFID indoor positioning method and device based on space-time information, medium and equipment. The method comprises: determining a data missing position in an original RSSI data vector acquired by an RFID reader; inputting the original RSSI data vector and the data missing position into a pre-trained generator to obtain a to-be-identified RSSI data vector; performing space-time feature extraction on the to-be-identified RSSI data vector to obtain a corresponding space-time feature vector; performing a position information embedding operation on the space-time feature vector according to position information of the RFID reader when acquiring the original RSSI data vector to obtain a target feature vector; and performing prediction on the target feature vector based on an attention mechanism to obtain a positioning prediction result. The technical scheme of the embodiments of the present application can improve the data quality while fully considering the space-time correlation of the RSSI data, thereby improving the accuracy of the indoor positioning result.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of indoor positioning technology, in particular to an RFID indoor positioning method and device based on space-time information, a medium and equipment. BACKGROUND

[0002] Traditional indoor positioning technologies, such as GPS systems, can provide accurate positioning services in outdoor environments, but in indoor environments, due to multipath effects, signal fading and non-line-of-sight challenges, the positioning accuracy is greatly reduced. In the current technical solutions, deep learning models are integrated into indoor positioning algorithms to improve the accuracy and efficiency of positioning results by mining spatial or temporal information in the data. However, the above method often focuses on improving the model, while ignoring the key influence of data quality. In addition, although RSSI data contains rich space-time features, people often only focus on one aspect and fail to fully exploit the space-time correlation. Therefore, how to improve data quality while fully considering the space-time correlation of RSSI data to improve the accuracy of indoor positioning results has become a technical problem to be solved. SUMMARY

[0003] Embodiments of the present application provide an RFID indoor positioning method and device based on space-time information, a medium and equipment, which can at least improve data quality while fully considering the space-time correlation of RSSI data, thereby improving the accuracy of indoor positioning results.

[0004] Other characteristics and advantages of the present application will become apparent from the following detailed description, or will be learned by practice of the present application.

[0005] According to an aspect of an embodiment of the present application, an RFID indoor positioning method based on space-time information is provided, comprising:

[0006] determining a data missing position in the original RSSI data vector based on the original RSSI data vector obtained by the RFID reader;

[0007] inputting the original RSSI data vector and the data missing position into a pre-trained generator to make the generator fill the data missing position in the original RSSI data vector, and obtaining a to-be-identified RSSI data vector;

[0008] extracting space-time features from the to-be-identified RSSI data vector to obtain a corresponding space-time feature vector;

[0009] According to position information of the RFID reader when the original RSSI data vector is acquired, a position information embedding operation is performed on the space-time feature vector to obtain a target feature vector.

[0010] Based on an attention mechanism, the target feature vector is predicted to obtain a positioning prediction result.

[0011] According to an aspect of an embodiment of the present application, an RFID indoor positioning device based on space-time information is provided, comprising:

[0012] A determination module is configured to determine a data missing position in the original RSSI data vector according to the original RSSI data vector acquired by the RFID reader.

[0013] A filling module is configured to input the original RSSI data vector and the data missing position into a generator which is pre-trained to fill data in the data missing position in the original RSSI data vector to obtain a to-be-recognized RSSI data vector.

[0014] An extraction module is configured to perform space-time feature extraction on the to-be-recognized RSSI data vector to obtain a corresponding space-time feature vector.

[0015] An embedding module is configured to perform a position information embedding operation on the space-time feature vector according to position information of the RFID reader when the original RSSI data vector is acquired to obtain a target feature vector.

[0016] A processing module is configured to perform prediction on the target feature vector based on an attention mechanism to obtain a positioning prediction result.

[0017] According to an aspect of an embodiment of the present application, a computer readable medium having a computer program stored thereon is provided, and the computer program is executed by a processor to implement the RFID indoor positioning method based on space-time information as described in the above embodiments.

[0018] According to an aspect of an embodiment of the present application, an electronic device is provided, comprising: one or more processors; a storage device configured to store one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the RFID indoor positioning method based on space-time information as described in the above embodiments.

[0019] According to an aspect of some embodiments of the present application, a computer program product or computer program is provided, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the method of indoor positioning of RFID based on spatio-temporal information provided in the above embodiments.

[0020] In the technical solutions provided by some embodiments of the present application, according to the original RSSI data vector obtained by the RFID reader, the data missing position in the original RSSI data vector is determined, and the original RSSI data vector and the data missing position are input into the generator which has been trained to obtain the data filling of the data missing position in the original RSSI data vector, so as to obtain the to-be-identified RSSI data vector. Thus, by training the generator to fill the RSSI data with data missing, the rationality of the filled data is ensured, and the data quality is improved. Then, the spatio-temporal feature of the to-be-identified RSSI data vector is extracted to obtain the corresponding spatio-temporal feature vector, the spatio-temporal correlation of the RSSI data is fully mined, and then the position information embedding operation is performed on the spatio-temporal feature vector according to the position information of the RFID reader when obtaining the original RSS data vector to obtain the target feature vector. Based on the attention mechanism, the target feature vector is predicted to obtain the positioning prediction result, so as to improve the accuracy of the indoor positioning prediction result.

[0021] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0022] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. In the drawings:

[0023] Figure 1 Fig. 1 shows a flowchart of a method of indoor positioning of RFID based on spatio-temporal information according to an embodiment of the present application;

[0024] Figure 2 Fig. 2 shows a training flowchart of a generative adversarial network according to an embodiment of the present application;

[0025] Figure 3 Fig. 3 shows a network architecture diagram of a convolutional neural network according to an embodiment of the present application;

[0026] Figure 4 A block diagram of a spatio-temporal information based RFID indoor positioning apparatus is shown according to an embodiment of the present application;

[0027] Figure 5 A structural schematic diagram of a computer system of an electronic device suitable for implementing embodiments of the present application is shown. DETAILED DESCRIPTION

[0028] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations can be implemented in any

[0029] Moreover, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the application. One skilled in the relevant art will recognize, however, that the

[0030] The block diagrams in the drawings show only the functional entities and not necessarily the physical separation of the functional entities. That is, the functional entities can be implemented in software, or in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0031] The flow diagrams shown in the drawings are merely examples and not necessarily to be construed as having all content and operations / steps, nor necessarily to be executed in the order described. For example, some operations / steps can be further divided, and some operations / steps can be combined or partially combined, so that the actual execution order can be changed according to actual conditions.

[0032] Figure 1 A flow diagram of a spatio-temporal information based RFID indoor positioning method is shown according to an embodiment of the present application.

[0033] The method can be applied to a terminal device or a server, wherein the terminal device can include, but is not limited to, one or more of a smartphone, a tablet computer, a portable computer, and a desktop computer, and can also be any electronic device with data receiving and data processing functions; the server can be a physical server or a cloud server.

[0034] As Figure 1 shown, the flowchart of the RFID indoor positioning method based on spatio-temporal information includes steps S110-S150, which are described in detail as follows (the method is applied to a terminal device, hereinafter referred to as "terminal"):

[0035] In step S110, the data missing position in the original RSSI data vector is determined according to the original RSSI data vector obtained by the RFID reader.

[0036] In this embodiment, the RFID reader can send multiple queries during movement, and the RFID middleware can save the response RSSI values of the multiple queries in the order of reception and generate a corresponding original RSSI data vector, and in addition, save the position information of the RFID reader when querying, for subsequent processing.

[0037] After obtaining the original RSSI data vector, the terminal can identify the data missing position in the original RSSI data vector, for example, the value of a certain position is 0 or close to 0, which can be determined as a data missing position, etc. In an example, the terminal can generate a vector to record the data missing position in the original RSSI data vector. The vector is consistent in shape with the original RSSI data vector, and the position corresponding to the data missing position in the original RSSI data vector in the vector is set to 0, and the remaining positions are set to 1.

[0038] In step S120, the original RSSI data vector and the data missing position are input into the generator trained in advance, so that the generator fills the data missing position in the original RSSI data vector to obtain a to-be-identified RSSI data vector.

[0039] In this embodiment, those skilled in the art can pre-construct a generative adversarial network and train the generator and discriminator in the generative adversarial network, so that the trained generator can fill the data missing position in the original RSSI data vector and ensure the rationality and effectiveness of the filled data.

[0040] In an embodiment, the RFID indoor positioning method based on spatio-temporal information further includes:

[0041] The RSSI training data vector and its corresponding mask vector are obtained, and the value corresponding to the data missing position in the RSSI training data vector in the mask vector is 0, and the remaining positions are 1.

[0042] The RSSI training data vector, the mask vector, and the noise variable are input into the generator to be trained, so that the generator fills in the missing data positions in the RSSI training data vector to obtain the prediction vector.

[0043] The predicted vector is input into the discriminator so that the discriminator can predict the missing data locations in the RSSI training data vector based on the predicted vector;

[0044] The generator and the discriminator are trained to improve the deceptiveness of the prediction vectors generated by the generator to the discriminator.

[0045] In this embodiment, those skilled in the art can pre-build the generator and discriminator and train them. Specifically, as... Figure 2 As shown, this application uses an RFID system as the hardware carrier to collect the RSSI sequence fingerprints of tags. The RFID reader moves along a predetermined trajectory, sending n queries during the process. The RFID middleware saves the RSSI values ​​of the responses to these n queries, as well as the location of the reader at the time of the query, and saves them together with the coordinates of the tag's location as training data in the database.

[0046] For the data vector X = (RSSI1,...,RSSI) in the database n )(Right now Figure 2 The Data Vector shown is denoted as n, where n is the number of queries by the RFID reader. Let there be a mask vector (i.e., Mask Vector) M = (M1, ..., Mn). n M has the same shape as X. For missing data positions in X, the corresponding position in M ​​is set to 0, and the other positions are set to 1.

[0047] Next, the generator G takes X, M, and the Gaussian noise variable Z as input and generates the imputed complete data according to the following formula.

[0048]

[0049] Here, ⊙ represents element-wise multiplication. The vector estimated by the corresponding generator G (G generates an estimated value for each component in the vector, regardless of whether the component is missing); For the corresponding complete vector (i.e., the complete data after imputation), the missing data locations are used... The value at the corresponding position in X is used, and the values ​​at the other positions are taken from the corresponding positions in X.

[0050] In the task of the discriminator D, its goal is not to identify whether the entire vector generated by the generator G is real or fake, but to distinguish which components in the vector are real and which are missing and then filled, i.e., to predict the mask vector M. In order to make the discriminator D learn better, the input prompt vector H can be introduced into the discriminator, the generation of H depends on M and contains information related to M.

[0051] When the discriminator D has difficulty in distinguishing the filled data generated by the generator G, the input prompt vector H can help improve the identification ability of the discriminator D. Therefore, the training goal of the generative adversarial network is to train the discriminator D to maximize the probability of correctly predicting M, and at the same time train the generator G to minimize the probability of the discriminator D predicting M, i.e.,

[0052]

[0053] After the terminal determines the data missing position of the original RSSI data vector, the terminal can input the original RSSI data vector and the data missing position into the generator trained by the above method, so that the generator can fill the data missing position in the original RSSI data vector to obtain a to-be-identified RSSI data vector.

[0054] Please continue to refer to Figure 1 In step S130, the to-be-identified RSSI data vector is subjected to spatio-temporal feature extraction to obtain a corresponding spatio-temporal feature vector.

[0055] In this embodiment, the terminal can perform spatio-temporal feature extraction on the to-be-identified RSSI data vector obtained after the filling of the generator, so as to extract a corresponding spatio-temporal feature vector.

[0056] In an embodiment, step S130 comprises:

[0057] The to-be-identified RSSI data vector is subjected to time feature extraction and space feature extraction respectively to obtain a corresponding time feature vector and a space feature vector;

[0058] The time feature vector and the space feature vector are combined to obtain a spatio-temporal feature vector.

[0059] In this embodiment, the spatio-temporal feature extraction comprises time feature extraction and space feature extraction. The terminal can perform time feature extraction and space feature extraction on the to-be-identified RSSI data vector respectively, and combine the feature vectors obtained by the two extractions to obtain a corresponding spatio-temporal feature vector.

[0060] In an embodiment, a Convolutional Neural Network (CNN) is employed to extract spatial features from the RSSI data vector to be recognized to obtain a corresponding spatial feature vector. In the spatial feature extraction, the RSSI data vector to be recognized is first converted into a two-dimensional image, and then the image is input into the CNN for spatial feature extraction.

[0061] Specifically, as shown in FIG. 2, the CNN includes three convolutional layers, each of which includes the following four parts: Figure 3

[0062] (1) Convolution operation: For the input image, in the first convolutional layer (convlayerl), 16 convolution kernels with a kernel size of 1x1 are used, which increases the non-linear characteristics while keeping the image size unchanged, thus helping feature extraction. Then, in the subsequent two convolutional layers (i.e., convlayer2 and convlayer3), two convolution kernels with a size of 3x3 are used, and the moving step of the three convolution kernels is 1. Therefore, the convolution kernel and the input image can overlap to the greatest extent, so that as many image features as possible can be captured. In an example, in order to ensure that the output and input feature map sizes are the same and avoid losing image edge information, padding filling with a size of 1 can be performed on the input image.

[0063] (2) Activation function: In order to enable the CNN to capture the complex non-linear relationship between the input and output positions of the input image, a Rectified Linear Unit (ReLU) activation function is used, which is defined as follows:

[0064]

[0065] where x is the input of the activation function.

[0066] (3) Pooling operation: used to perform a pooling operation on the feature map after each convolution operation, which subsamples the feature map by retaining useful details and discarding irrelevant information. The size of the Max-Pool window is set to 2x2, and the stride is 1. The Max-Pool function is defined as follows:

[0067]

[0068] where σ ij is the output of the pooling operator, x pq is the element at position (p, q) in the Max-Pool window, which represents the neighborhood of position (i, j).

[0069] ​(4) Dropout operation: due to the small size of the data set, in order to prevent overfitting, the feature map obtained after the max pooling operation can be subjected to a dropout operation. In an example, the dropout rate used by the CONV layer can be set to 25%.

[0070] In an embodiment, the terminal performs time feature extraction on the to-be-identified RSSI data vector by using an LSTM model, so as to obtain a corresponding time feature vector.

[0071] After obtaining the spatial feature vector and the time feature vector corresponding to the to-be-identified RSSI data vector, in order to capture the spatio-temporal correlation, the spatial feature vector and the time feature vector are subjected to a concatenate operation, so as to obtain a corresponding spatio-temporal feature vector.

[0072] Please continue to refer to Figure 1 In step S140, according to the position information of the RFID reader when obtaining the original RSSI data vector, a position information embedding operation is performed on the spatio-temporal feature vector to obtain a target feature vector.

[0073] In this embodiment, the terminal can obtain the position information of the RFID reader when obtaining the original RSSI data vector, and the position information can include the positioning information of the RFID reader at multiple time points. Then, the terminal can embed the position information into the spatio-temporal feature vector, so as to enrich the information amount of the spatio-temporal feature vector, so as to facilitate subsequent positioning of the RFID chip.

[0074] In an embodiment, step S140 includes:

[0075] performing an embedding operation on the spatio-temporal feature vector to obtain a corresponding embedding vector;

[0076] performing a position information embedding operation on the embedding vector according to the position information of the RFID reader when obtaining the original RSSI data vector to obtain a target feature vector.

[0077] In this embodiment, the terminal can perform an embedding operation on the obtained spatio-temporal feature vector to obtain a corresponding embedding vector. It should be understood that the embedding operation is a process of mapping high-dimensional data or objects to a low-dimensional vector space, and the low-dimensional vectors can capture important features of the original data.

[0078] Then, the terminal can perform a position information embedding operation on the spatio-temporal feature vector according to the position information of the RFID reader when the original RSSI data vector is acquired, so as to obtain a target feature vector. In this way, the expression capability of the target feature vector can be enhanced, so that the subsequent model can better capture the spatial distribution and geographical correlation, and the accuracy of the subsequent model in positioning the RFID chip can be improved.

[0079] In an example, the position information embedding operation on the embedding vector is performed according to the following formula:

[0080]

[0081] wherein pos is the position information of the RFID reader, 2i is an even dimension of the embedding vector, 2i+1 is an odd dimension of the embedding vector, and dmodel is the dimension of the embedding vector.

[0082] Please continue to refer to Figure 1 In step S150, the target feature vector is predicted based on an attention mechanism to obtain a positioning prediction result.

[0083] In this embodiment, the attention mechanism enables the model to pay more attention to information that is more important for the current task, thereby improving the accuracy of the output positioning prediction result.

[0084] Specifically, by multiplying the target feature vector with randomly initialized matrices W q , W k , and W v , three vectors Q, K, and V are obtained, and then the corresponding attention matrix is calculated according to the following formula:

[0085]

[0086] wherein dk is a one-dimensional parameter.

[0087] Finally, the attention matrix is predicted through a fully connected layer to obtain the corresponding positioning prediction result.

[0088] In this way, the positioning prediction result is obtained based on Figure 1In the embodiment shown, according to the original RSSI data vector acquired by the RFID reader, a data missing position in the original RSSI data vector is determined, the original RSSI data vector and the data missing position are input into the generator which has been trained in advance, so that the generator fills data in the data missing position in the original RSSI data vector, to obtain a to-be-identified RSSI data vector. Thus, by filling the RSSI data with data missing through the trained generator, the rationality of the filled data is ensured, and the data quality is improved. Then, spatio-temporal feature extraction is performed on the to-be-identified RSSI data vector, to obtain a corresponding spatio-temporal feature vector, the spatio-temporal correlation of the RSSI data is fully mined, and then, according to the position information of the RFID reader when acquiring the original RSS data vector, a position information embedding operation is performed on the spatio-temporal feature vector, to obtain a target feature vector. Based on the attention mechanism, the target feature vector is predicted, to obtain a positioning prediction result, so as to improve the accuracy of the indoor positioning prediction result.

[0089] The device embodiment of the present application is introduced below, which can be used to execute the RFID indoor positioning method based on spatio-temporal information in the embodiments of the present application. For details not disclosed in the device embodiment of the present application, refer to the embodiments of the RFID indoor positioning method based on spatio-temporal information described above.

[0090] Figure 4 A block diagram of an RFID indoor positioning device based on spatio-temporal information according to one embodiment of the present application is shown.

[0091] Reference Figure 4 As shown, the RFID indoor positioning device based on spatio-temporal information according to one embodiment of the present application comprises:

[0092] A determination module, configured to determine a data missing position in an original RSSI data vector acquired by an RFID reader according to the original RSSI data vector;

[0093] A filling module, configured to input the original RSSI data vector and the data missing position into a generator which has been trained in advance, so that the generator fills data in the data missing position in the original RSSI data vector, to obtain a to-be-identified RSSI data vector;

[0094] An extraction module, configured to perform spatio-temporal feature extraction on the to-be-identified RSSI data vector, to obtain a corresponding spatio-temporal feature vector;

[0095] An embedding module, configured to perform a position information embedding operation on the spatio-temporal feature vector according to position information of the RFID reader when acquiring the original RSSI data vector, to obtain a target feature vector;

[0096] a processing module configured to predict the target feature vector based on an attention mechanism to obtain a positioning prediction result.

[0097] In an embodiment of the present application, the to-be-identified RSSI data vector is subjected to spatiotemporal feature extraction to obtain a corresponding spatiotemporal feature vector, including:

[0098] The to-be-identified RSSI data vector is subjected to time feature extraction and space feature extraction respectively to obtain a corresponding time feature vector and a space feature vector;

[0099] The time feature vector and the space feature vector are combined to obtain a spatiotemporal feature vector.

[0100] In an embodiment of the present application, the spatiotemporal feature vector is subjected to position information embedding operation according to position information of the RFID reader when the original RSSI data vector is acquired to obtain a target feature vector, including:

[0101] The spatiotemporal feature vector is subjected to embedding operation to obtain a corresponding embedding vector;

[0102] The embedding vector is subjected to position information embedding operation according to position information of the RFID reader when the original RSSI data vector is acquired to obtain a target feature vector.

[0103] In an embodiment of the present application, the processing module is further configured to:

[0104] acquire an RSSI training data vector and a corresponding mask vector thereof, wherein a value corresponding to a data missing position in the RSSI training data vector in the mask vector is 0, and other positions are 1;

[0105] input the RSSI training data vector, the mask vector and a noise variable into a to-be-trained generator to enable the generator to fill the data missing position in the RSSI training data vector to obtain a prediction vector;

[0106] input the prediction vector into a discriminator to enable the discriminator to predict the data missing position in the RSSI training data vector according to the prediction vector;

[0107] train the generator and the discriminator to improve the cheating of the prediction vector generated by the generator to the discriminator.

[0108] Figure 5 A structural schematic diagram of a computer system of an electronic device suitable for implementing embodiments of the present application is shown.

[0109] It should be noted that Figure 5 The computer system of the electronic device shown is merely one example, and should not bring any limitation to the function and usage range of the embodiments of the present application.

[0110] As Figure 5 shown, the computer system includes a central processing unit (CPU) 501 which can perform various appropriate actions and processes in accordance with a program stored in a read-only memory (ROM) 502 or a program loaded from a storage section 508 into a random access memory (RAM) 503, such as performing the methods described in the above embodiments. In the RAM 503, various programs and data required for the operation of the system are also stored. The CPU 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0111] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as necessary. A removable recording medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 510 as necessary, so that a computer program read out therefrom is installed in the storage section 508 as necessary.

[0112] In particular, according to the embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, the embodiments of the present application include a computer program product including a computer program carried on a computer-readable medium, the computer program containing a computer program for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication section 509, and / or installed from the removable recording medium 511. When the computer program is executed by the central processing unit (CPU) 501, various functions defined in the system of the present application are performed.

[0113] It should be noted that the computer-readable medium in the embodiments of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In this application, the computer-readable signal medium can include a data signal carrying computer-readable computer programs in a baseband or as a part of a carrier wave. Such a propagated data signal can take on various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium that can transmit, propagate or transport programs for use by or in connection with an instruction execution system, device or apparatus. The computer programs contained in the computer-readable medium can be transmitted by any suitable medium, including but not limited to wireless, wired, or the like, or any suitable combination thereof.

[0114] The flowcharts and block diagrams in the drawings illustrate the possible implementation architectures, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In the flowcharts or block diagrams, each block can represent a module, a program segment or a part of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different orders than that shown in the drawings. For example, two blocks that are shown in succession can actually be executed substantially in parallel, and sometimes in reverse order, depending on the involved functions. It should also be noted that each block in the block diagrams or flowcharts, and the combination of blocks in the block diagrams or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0115] The units described in the embodiments of the present application can be implemented by software, or by hardware, or by a combination of software and hardware. The units described can also be located in a single processor. In some cases, the names of the units do not limit the units themselves.

[0116] As another aspect, the present application also provides a computer readable medium, which can be included in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to implement the method described in the above embodiments.

[0117] It should be noted that although several modules or units for performing actions are mentioned in the above detailed description, the division into the modules or units is not mandatory. In fact, according to the embodiments of the present application, features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, features and functions of one module or unit described above can be further divided into a plurality of modules or units.

[0118] From the above description of the embodiments, those skilled in the art will readily appreciate that the example embodiments described herein can be implemented by software and / or by hardware coupled with software. Accordingly, the technical solutions of the embodiments of the present application can be embodied in the form of a software product. The software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, or the like) or on a network, and includes a number of instructions for causing a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to perform the methods according to the embodiments of the present application.

[0119] Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the embodiments disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such

[0120] It should be understood that the present application is not limited to the precise construction that has been described above and illustrated in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the present application. The scope of the present application is limited only by the appended claims.

Claims

1. A method for RFID indoor positioning based on spatio-temporal information, characterized in that, The method comprises the following steps: determining a data missing position in the original RSSI data vector acquired by the RFID reader; inputting the original RSSI data vector and the data missing position into a pre-trained generator to make the generator fill the data missing position in the original RSSI data vector, and obtaining a to-be-identified RSSI data vector; extracting a space-time feature of the to-be-identified RSSI data vector to obtain a corresponding space-time feature vector; embedding position information of the RFID reader when acquiring the original RSSI data vector into the space-time feature vector to obtain a target feature vector; predicting the target feature vector based on an attention mechanism to obtain a positioning prediction result; wherein the embedding position information of the RFID reader when acquiring the original RSSI data vector into the space-time feature vector to obtain the target feature vector comprises: performing an embedding operation on the space-time feature vector to obtain a corresponding embedding vector; embedding position information of the RFID reader when acquiring the original RSSI data vector into the embedding vector to obtain the target feature vector; wherein the embedding position information of the RFID reader when acquiring the original RSSI data vector into the embedding vector to obtain the target feature vector is performed according to the following formula: wherein pos is the position information of the RFID reader, 2i is the even dimension of the embedding vector, 2i+1 is the odd dimension of the embedding vector, d model is the dimension of the embedding vector.

2. The method of claim 1, wherein, extracting a space-time feature of the to-be-identified RSSI data vector to obtain a corresponding space-time feature vector comprises: respectively extracting a time feature and a space feature of the to-be-identified RSSI data vector to obtain a corresponding time feature vector and a space feature vector; combining the time feature vector and the space feature vector to obtain a space-time feature vector.

3. The method according to any one of claims 1-2, characterized in that, The method further comprises: acquiring an RSSI training data vector and a corresponding mask vector, wherein a value corresponding to a data missing position in the RSSI training data vector in the mask vector is 0, and other positions are 1; inputting the RSSI training data vector, the mask vector, and a noise variable into a to-be-trained generator to make the generator fill the data missing position in the RSSI training data vector, and obtaining a prediction vector; inputting the prediction vector into a discriminator to make the discriminator predict the data missing position in the RSSI training data vector according to the prediction vector; training the generator and the discriminator to improve the cheating of the prediction vector generated by the generator to the discriminator.

4. A spatio-temporal information based RFID indoor positioning apparatus, characterized by The method comprises the following steps: a determining module configured to determine a data missing position in an original RSSI data vector acquired by an RFID reader; a filling module configured to input the original RSSI data vector and the data missing position into a pre-trained generator to make the generator fill the data missing position in the original RSSI data vector, and obtain a to-be-identified RSSI data vector; An extraction module is configured to perform spatio-temporal feature extraction on the to-be-identified RSSI data vector to obtain a corresponding spatio-temporal feature vector. An embedding module is configured to perform position information embedding on the spatio-temporal feature vector according to position information of the RFID reader when the original RSSI data vector is acquired to obtain a target feature vector. A processing module is configured to perform prediction on the target feature vector based on an attention mechanism to obtain a positioning prediction result. The position information embedding on the spatio-temporal feature vector according to the position information of the RFID reader when the original RSSI data vector is acquired to obtain the target feature vector includes: performing embedding on the spatio-temporal feature vector to obtain a corresponding embedding vector; performing position information embedding on the embedding vector according to the position information of the RFID reader when the original RSSI data vector is acquired to obtain the target feature vector. The position information embedding on the embedding vector is performed according to the following formula: wherein pos is the position information of the RFID reader, 2i is the even dimension of the embedding vector, 2i+1 is the odd dimension of the embedding vector, d model is the dimension of the embedding vector.

5. The apparatus of claim 4, wherein, The spatio-temporal feature extraction on the to-be-identified RSSI data vector to obtain the corresponding spatio-temporal feature vector includes: respectively performing time feature extraction and space feature extraction on the to-be-identified RSSI data vector to obtain a corresponding time feature vector and a space feature vector; jointly processing the time feature vector and the space feature vector to obtain the spatio-temporal feature vector.

6. A computer readable medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the RFID indoor positioning method based on spatio-temporal information according to any one of claims 1 to 3.

7. An electronic device, comprising: It includes: one or more processors; a storage device configured to store one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the RFID indoor positioning method based on spatio-temporal information according to any one of claims 1 to 3.

Images