An indoor positioning method based on 5G spatiotemporal big data collaboration

Abstract

The present invention provides an indoor positioning method based on 5G spatiotemporal big data collaboration, which targets indoor positioning without auxiliary positioning equipment and performs indoor positioning and trajectory tracking based on 5G spatiotemporal big data collaboration. The spatiotemporal big data acquired by the mobile terminal, including spatiotemporal framework side data and spatiotemporal change side data, are respectively input into the RBF radial basis function neural network model for training, and multiple hyperparameters of the training results are used as fingerprint features of the location points to build a multidimensional fingerprint database to achieve indoor positioning and trajectory tracking. The present invention covers the design of a 5G spatiotemporal big data collaboration mechanism, the construction of a multidimensional fingerprint library based on 5G spatiotemporal big data collaboration, real-time positioning in indoor environments, and trajectory tracking for indoor environments. It can effectively solve the problem that users find it difficult to obtain accurate positions indoors without auxiliary positioning equipment, and to a certain extent provide technical support for services based on indoor positioning technology.

Description

Technical Field

[0001] The present invention belongs to the field of electrical digital data processing, and specifically relates to an indoor positioning method based on 5G spatiotemporal big data collaboration. Background Art

[0002] With the guidance of development strategies such as digital transformation, intelligent upgrading, and innovative integration, the demand for indoor location services is becoming increasingly urgent, such as logistics equipment / mobile terminal positioning in smart factories, mobile terminal positioning in manufacturing parks, and shopping mall floor guidance. At present, indoor positioning is mainly based on technologies such as UWB, Bluetooth, WiFi fingerprint library, and 4G, and is achieved with the assistance of positioning equipment (RFID, Bluetooth gateway, etc.). On the one hand, due to the high accuracy of the UWB positioning system, the short transmission distance of Bluetooth positioning, the small coverage area of ​​WiFi fingerprint library positioning, and the weak anti-interference ability of 4G positioning, the independent use of the above methods results in poor positioning accuracy. On the other hand, the deployment requirements of auxiliary positioning equipment have high environmental requirements and need to be prepared in advance, and indoor positioning in large spaces, sudden situations, and harsh environments cannot be achieved. Therefore, based on 5G spatiotemporal big data, the present invention proposes an indoor positioning method based on 5G spatiotemporal big data collaboration, combines data collaboration to design a 5G spatiotemporal big data collaboration mechanism, realizes the collaboration of different positioning data, and improves the accuracy of the final positioning result. In addition, the indoor positioning method does not require additional deployment of positioning equipment and can meet the needs of indoor positioning in large spaces, sudden situations, and harsh environments. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention provides an indoor positioning method based on 5G spatiotemporal big data collaboration, which is aimed at indoor positioning without auxiliary positioning equipment, and performs indoor positioning and position tracking based on 5G spatiotemporal big data collaboration. The spatiotemporal big data is obtained, and the spatiotemporal big data is divided according to the spatiotemporal framework side data and the spatiotemporal change side data. The divided data are respectively input into the RBF neural network model for training, and the multiple hyperparameters of the training results are used as the fingerprint features of the location points. A multidimensional fingerprint database is built with these fingerprint features to achieve indoor positioning and tracking. The present invention covers the design of a 5G spatiotemporal big data collaboration mechanism, the construction of a multidimensional fingerprint library based on 5G spatiotemporal big data collaboration, the realization of real-time positioning of indoor environments, and the realization of indoor environment trajectory tracking of mobile terminals. It can effectively solve the problem that users find it difficult to obtain accurate positions indoors without auxiliary positioning equipment, and to a certain extent provides comprehensive technical support for various service functions based on indoor positioning technology.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is:

[0005] An indoor positioning method based on 5G spatiotemporal big data collaboration includes the following steps:

[0006] Step (1) performing positioning and solving on the spatiotemporal framework side data and the spatiotemporal change side data based on the collaborative mechanism of 5G spatiotemporal big data, realizing the coordination of the spatiotemporal reference data and GPS data in the spatiotemporal framework side data and the positioning reference signals, channel sounding reference signals, round-trip time data, and MR signaling data in the Wi-Fi and 5G network in the spatiotemporal change side data;

[0007] Step (2) building a multidimensional fingerprint database based on a collaborative mechanism: through 5G spatiotemporal big data collaboration, the processed data are input into the radial basis function neural network model for training, and multiple hyperparameters of the training results are used as fingerprint features of the location points, and a multidimensional fingerprint database is built with these fingerprint features;

[0008] Step (3), indoor positioning based on spatiotemporal position solution: Based on the acquired spatiotemporal big data of the mobile terminal, the fingerprint features after the spatiotemporal big data position solution are matched with the fingerprint library, and the position point with the maximum Bayesian probability is output as the current mobile terminal positioning result;

[0009] Step (4), trajectory tracking based on the spatiotemporal threshold method: Based on step (3), the position of the mobile terminal in the indoor environment is obtained, and the spatial stay point of the mobile terminal at a certain moment is extracted using the spatiotemporal interaction verification method based on the spatiotemporal framework side data and the spatiotemporal change side data; the stay area is marked based on the indoor topological layout, and the trajectory tracking of the mobile terminal in the indoor environment is achieved by marking the stay area multiple times in a short period of time.

[0010] Furthermore, the step (1) specifically includes:

[0011] Step (1.1) acquires spatiotemporal framework side data and spatiotemporal change side data, wherein the spatiotemporal framework side data includes spatiotemporal reference data and GPS data, and the spatiotemporal change side data includes WiFi, PRS, SRS, RTT, DL-AoD, UL-AOA, MR, and signaling data;

[0012] Step (1.2) constructs an objective function by minimizing the weighted sum of squares of the residuals on the spatiotemporal frame side data, performs positioning calculation on the objective function, and obtains an independent positioning result;

[0013] In step (1.3), the objective function is constructed by minimizing the weighted sum of squares of the residuals of the spatiotemporal change side data and the spatiotemporal frame side data, and the independent positioning results of the spatiotemporal frame side data in step (1.2) are collaboratively positioned and solved with the spatiotemporal change side objective function.

[0014] Furthermore, the step (2) specifically includes:

[0015] Step (2.1) acquires spatiotemporal framework data and spatiotemporal change data from the mobile terminal: collects mobile terminal RSS data information, MR data and signaling data uploaded by the UE at the sample point, and the location coordinates of the sample in the small area;

[0016] Step (2.2) 5G spatiotemporal big data cleaning: noise cleaning and invalid data elimination are performed on the acquired 5G spatiotemporal big data;

[0017] Step (2.3) Building a multi-dimensional fingerprint database based on 5G spatiotemporal big data collaboration: Divide the spatiotemporal big data processed in step (2.2) into spatiotemporal framework data and spatiotemporal change data, and input the divided spatiotemporal big data into radial basis function neural network for training;

[0018] Step (2.4) obtains the trained radial basis function neural network model hyperparameters based on step (2.3), uses the multiple hyperparameters of each location point of the mobile terminal as the fingerprint features of the location point, and builds a multidimensional fingerprint database with these fingerprints.

[0019] Furthermore, the step (3) specifically includes:

[0020] Step (3.1) Obtaining 5G spatiotemporal big data of the mobile terminal: The cell level, RSS data, and TA+AOA data reported by the mobile terminal in the real-time positioning phase are coordinated through the data coordination mechanism of step (1) to form the feature vector of the measurement point, thereby determining the state space;

[0021] Step (3.2) Real-time matching and positioning: Based on the feature vector of the current measurement point formed after data collaboration, the neural network hyperparameters in the fingerprint library are used to perform optimal recognition and matching with the feature vector to achieve accurate positioning of the mobile terminal.

[0022] Furthermore, the step (4) specifically includes:

[0023] Step (4.1) Identification of stay points and movement points based on the spatiotemporal threshold method: Using the spatiotemporal frame data and spatiotemporal change data processed by the collaborative mechanism, a dynamic spatiotemporal threshold algorithm is used to identify stay points and movement points based on the spatial distance between two adjacent positioning results at an appropriate positioning frequency;

[0024] Step (4.2) Marking the stay area based on the user's indoor topology layout: For the identified stay points, the stay area is marked in combination with the pre-collected user indoor topology and layout map, and the positioning reference signal containing spatial location information processed by the collaborative mechanism to determine the stay area scene;

[0025] Step (4.3) is based on the stay area scene determined in step (4.2), and the stay area is determined multiple times in a short period of time, thereby depicting the user's indoor movement trajectory and realizing indoor environment trajectory tracking of the mobile terminal.

[0026] The advantages of the present invention compared with the prior art are:

[0027] Compared with traditional indoor positioning methods, which are highly dependent on auxiliary positioning equipment and have low positioning accuracy, the present invention provides an indoor positioning method based on 5G spatiotemporal big data collaboration. This method targets the limitations of traditional indoor positioning methods, is based on 5G spatiotemporal big data, and relies on the characteristics of 5G network space, such as wide coverage and fast transmission rate, to perform data collaboration and obtain positioning results. Spatiotemporal big data is obtained, and the spatiotemporal big data is divided according to the spatiotemporal framework side data and the spatiotemporal change side data. The divided data are respectively input into the RBF radial basis function neural network model for training, and multiple hyperparameters of the training results are used as fingerprint features of the location points. A multidimensional fingerprint database is built with these fingerprint features to achieve indoor positioning and tracking. It can effectively solve the problem that users cannot obtain accurate positions indoors, and to a certain extent provide comprehensive technical support for various service functions based on indoor positioning technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flowchart of a 5G spatiotemporal big data collaboration mechanism based on the present invention;

[0029] Figure 2 This is a flowchart of an indoor positioning method based on 5G spatiotemporal big data collaboration of the present invention. DETAILED DESCRIPTION

[0030] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0031] like Figure 2 As shown, the indoor positioning method based on 5G spatiotemporal big data collaboration of the present invention specifically includes the following steps:

[0032] Step (1) positioning and solving the spatiotemporal framework side data and the spatiotemporal change side data based on the collaborative mechanism of 5G spatiotemporal big data, realizing the coordination of spatiotemporal reference data and GPS data in the spatiotemporal framework side data and PRS (positioning reference signal), SRS (channel sounding reference signal), RTT (round trip time) data in the WiFi and 5G network in the spatiotemporal change side data, and MR signaling data;

[0033] Step (2) build a multidimensional fingerprint database based on a collaborative mechanism: through the collaboration of 5G spatiotemporal big data, the processed data are input into the RBF (radial basis function) neural network model for training, and the hyperparameters of the training results are used as multiple fingerprint features of the location points, and a multidimensional fingerprint database is built with these fingerprint features;

[0034] Step (3) Indoor positioning based on spatiotemporal position solution: Based on the acquired current spatiotemporal big data of the mobile terminal, the fingerprint features after the spatiotemporal big data position solution are matched with the fingerprint library, and the position point with the maximum Bayesian probability is output as the current mobile terminal positioning result;

[0035] Step (4) Trajectory tracking based on the spatiotemporal threshold method: Based on step (3), the location of the mobile terminal in the indoor environment is obtained. Based on the spatiotemporal framework data and the spatiotemporal change data, the spatiotemporal interaction verification method is used to extract the spatial stay point of the mobile terminal at a certain moment. The stay area is marked based on the indoor topological layout. By marking the stay area multiple times in a short period of time, the trajectory tracking of the mobile terminal in the indoor environment is achieved.

[0036] Specifically, if Figure 1 As shown, the step (1) includes:

[0037] Step (1.1) acquires spatiotemporal framework side data and spatiotemporal change side data, wherein the spatiotemporal framework side data includes spatiotemporal reference data and GPS data, and the spatiotemporal change side data includes WiFi, PRS, SRS, RTT, DL-AoD, UL-AOA, MR, and signaling data;

[0038] Step (1.2) constructs an objective function by minimizing the weighted sum of squares of the residuals on the spatiotemporal frame side data, performs positioning calculation on the objective function, and obtains an independent positioning result;

[0039] In step (1.3), the objective function is constructed by minimizing the weighted sum of squares of the residuals of the spatiotemporal change side data and the spatiotemporal frame side data, and the independent positioning results of the spatiotemporal frame side data in step (1.2) are collaboratively positioned and solved with the spatiotemporal change side objective function.

[0040] Specifically, the step (2) includes:

[0041] Step (2.1) acquires spatiotemporal frame data and spatiotemporal change data from the mobile terminal. This includes collecting the mobile terminal RSS data, the MR data and signaling data uploaded by the UE at the sample point, and the location coordinates of the sample in the small area.

[0042] Step (2.2) 5G spatiotemporal big data cleaning: noise cleaning and invalid data removal are performed on the acquired 5G spatiotemporal big data;

[0043] Step (2.3) is to build a multi-dimensional fingerprint database based on the collaboration of 5G spatiotemporal big data. The spatiotemporal big data processed in step (2.2) is divided into the spatiotemporal framework side data and the spatiotemporal change side data, and the divided spatiotemporal big data are respectively input into the RBF radial basis function neural network for training;

[0044] Step (2.4) obtains the trained radial basis function neural network model hyperparameters based on step (2.3), uses the multiple hyperparameters of each location point of the mobile terminal as multiple fingerprint features of the location point, and builds a multidimensional fingerprint database with these fingerprints.

[0045] Specifically, the step (3) includes:

[0046] Step (3.1) obtains the 5G spatiotemporal big data of the mobile terminal. The cell level, RSS data, TA+AOA and other data reported by the mobile terminal in the real-time positioning phase are coordinated through the data coordination mechanism of step (1) to form the feature vector of the measurement point, thereby determining the state space.

[0047] Step (3.2) Real-time matching and positioning: Based on the feature vector of the current measurement point formed after data collaboration, the neural network hyperparameters in the fingerprint library are used to optimally identify and match the feature vector to achieve accurate positioning of the mobile terminal.

[0048] Specifically, the step (4) includes:

[0049] Step (4.1) Identification of stay points and movement points based on the spatiotemporal threshold method. Using the spatiotemporal frame data and spatiotemporal change data processed by the collaborative mechanism, a dynamic spatiotemporal threshold algorithm is used to identify stay points and movement points based on the spatial distance between two adjacent positioning results at an appropriate positioning frequency.

[0050] Step (4.2) Marking the stay area based on the user's indoor topology layout: For the identified stay points, the stay area is marked in combination with the pre-collected user indoor topology and layout map, and the positioning reference signal containing spatial location information processed by the collaborative mechanism to determine the stay area scene;

[0051] Step (4.3) is based on the stay area scene determined in step (4.2), and the stay area is determined multiple times in a short period of time, thereby depicting the user's indoor movement trajectory and realizing indoor environment trajectory tracking of the mobile terminal.

[0052] The present invention's 5G spatiotemporal big data collaborative mechanism specifically includes: independent positioning of spatiotemporal framework data: acquiring spatiotemporal framework data, including spatiotemporal reference data and GPS data. Based on these two types of data, an objective function is constructed by minimizing the weighted sum of squares of residuals. The objective function is then positioned and solved using the Newton iteration method, outputting an independent positioning result. Fusion positioning of spatiotemporal variation data: acquiring spatiotemporal variation data, including WiFi, PRS, SRS, RTT, DL-AoD, UL-AOA, MR, and signaling data, and combining the spatiotemporal framework data to construct an objective function by minimizing the weighted sum of squares of residuals. The initial positioning result of the spatiotemporal framework data is fused with the spatiotemporal variation data to calculate a closed-form positioning solution, and the positioning result is output for accuracy judgment. If the result meets the accuracy requirements, the final positioning result is directly output; if the result does not meet the accuracy requirements, a numerical solution method is used to process the positioning result and the initial position, and the final positioning result is output.

[0053] In summary, the present invention provides an indoor positioning method based on 5G spatiotemporal big data collaboration. This method aims at the situation where the indoor positioning accuracy is not high and relies on auxiliary positioning equipment. Based on 5G spatiotemporal big data, relying on the characteristics of 5G network space such as wide coverage and fast transmission rate, data collaboration is performed to obtain collaborative positioning results. The spatiotemporal big data of the mobile terminal is obtained, and the spatiotemporal big data is divided according to the spatiotemporal framework side data and the spatiotemporal change side data. The divided data are respectively input into the RBF radial basis function neural network model for training, and the multiple hyperparameters of the training results are used as the fingerprint features of the location points. A multidimensional fingerprint database is built with these fingerprint features to achieve indoor positioning and tracking. The present invention covers the design of a collaborative mechanism based on 5G spatiotemporal big data, the construction of a multidimensional fingerprint library based on the collaboration of 5G spatiotemporal big data, the realization of real-time positioning of indoor environments, and the realization of indoor environment trajectory tracking of mobile terminals. It can effectively solve the problem that users cannot obtain accurate positions indoors, and to a certain extent provide comprehensive technical support for various service functions based on indoor positioning technology.

[0054] The contents not described in detail in the specification of the present invention belong to the prior art known to those skilled in the art.

[0055] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. An indoor positioning method based on 5G spatiotemporal big data collaboration, characterized in that: The steps include: Step (1), a collaborative mechanism based on 5G spatiotemporal big data: by performing positioning and solving the 5G spatiotemporal framework side data and the 5G spatiotemporal change side data, the spatiotemporal reference data and GPS data in the spatiotemporal framework side data are coordinated with the positioning reference signals, channel sounding reference signals, round-trip time data, and MR signaling data in the WiFi and 5G network in the spatiotemporal change side data, including: Step (1.1) acquiring spatiotemporal framework side data and spatiotemporal change side data, wherein the spatiotemporal framework side data includes spatiotemporal reference data and GPS data, and the spatiotemporal change side data includes WiFi, PRS, SRS, RTT, DL-AoD, UL-AOA, MR, and signaling data; Step (1.2) constructs an objective function by minimizing the weighted sum of squares of the residuals on the spatiotemporal frame side data, performs positioning calculation on the objective function, and obtains an independent positioning result; In step (1.3), the objective function is constructed by minimizing the weighted sum of squares of the residuals of the spatiotemporal change side data and the spatiotemporal frame side data. The independent positioning results of the spatiotemporal frame side data in step (1.2) are collaboratively positioned and solved with the spatiotemporal change side objective function. Step (2) Building a multi-dimensional fingerprint database based on a collaborative mechanism: Based on the 5G spatiotemporal big data collaborative mechanism, the processed data are input into the radial basis function neural network model for training, and multiple hyperparameters of the training results are used as fingerprint features of the location points. These fingerprint features are used to build a multi-dimensional fingerprint database, including: Step (2.1) Obtaining spatiotemporal framework data and spatiotemporal change data from mobile terminals: Collecting mobile terminal RSS data information, MR data and signaling data uploaded by the UE at the sample point, and the location coordinates of the sample in the small area; Step (2.2) 5G spatiotemporal big data cleaning: noise cleaning and invalid data removal are performed on the acquired 5G spatiotemporal big data; Step (2.3) Building a multi-dimensional fingerprint database based on 5G spatiotemporal big data collaboration: The spatiotemporal big data processed in step (2.2) is divided into spatiotemporal framework side data and spatiotemporal change side data, and the divided spatiotemporal big data are respectively input into the radial basis function neural network for training; Step (2.4) obtains the hyperparameters of the trained radial basis function neural network model based on step (2.3), uses the multiple hyperparameters of each location point of the mobile terminal as multiple fingerprint features of the location point, and builds a multidimensional fingerprint database with these fingerprints; Step (3), indoor positioning based on spatiotemporal positioning solution: Based on the current spatiotemporal big data of the mobile terminal, the fingerprint features after spatiotemporal big data positioning solution are matched with the fingerprint library, and the location point with the maximum Bayesian probability is output as the current mobile terminal positioning result; Step (4), trajectory tracking based on the spatiotemporal threshold method: Based on step (3), the position of the mobile terminal in the indoor environment is obtained, and the spatial stay point of the mobile terminal at a certain moment is extracted using the spatiotemporal interaction verification method based on the spatiotemporal framework side data and the spatiotemporal change side data; the stay area is marked based on the indoor topological layout, and the trajectory tracking of the mobile terminal in the indoor environment is achieved by marking the stay area multiple times in a short period of time.

2. The indoor positioning method based on 5G spatiotemporal big data collaboration according to claim 1 is characterized in that: The step (3) specifically includes: Step (3.1) Obtaining 5G spatiotemporal big data of the mobile terminal: The cell level, RSS data, and TA+AOA data reported by the mobile terminal in the real-time positioning phase are coordinated through the data coordination mechanism of step (1) to form the feature vector of the current measurement point, thereby determining the state space; Step (3.2) Real-time matching and positioning: Based on the feature vector of the current measurement point formed after data collaboration, the neural network hyperparameters in the fingerprint library are used to perform optimal recognition and matching with the feature vector to achieve accurate positioning of the mobile terminal.

3. The indoor positioning method based on 5G spatiotemporal big data collaboration according to claim 2 is characterized in that: The step (4) specifically includes: Step (4.1) Identification of stay points and movement points based on the spatiotemporal threshold method: Using the spatiotemporal frame data and spatiotemporal change data processed by the collaborative mechanism, a dynamic spatiotemporal threshold algorithm is used to identify stay points and movement points based on the spatial distance between two adjacent positioning results at an appropriate positioning frequency; Step (4.2) Marking the stay area based on the user's indoor topology layout: For the identified stay points, the stay area is marked in combination with the pre-collected user indoor topology and layout map, and the positioning reference signal containing spatial location information processed by the collaborative mechanism to determine the stay area scenario; Step (4.3) is based on the stay area scene determined in step (4.2), and the stay area is determined multiple times in a short period of time, thereby depicting the user's indoor movement trajectory and realizing indoor environment trajectory tracking of the mobile terminal.

Images