Big data-based positioning method and system with privacy protection function

Abstract

The invention belongs to the field of big data application, and provides accurate positioning for users through combination of big data information and satellite positioning signals. Used data mainly comprises regional historical positioning data, user historical positioning data and multi-source satellite positioning information data, and related knowledge of machine learning is applied, so that accurate positioning service is provided for users, the defects of large positioning errors and unstable signals caused by direct application of satellite signals are overcome, and the positioning accuracy is improved. And meanwhile, certain privacy protection can be provided for the user, so that powerful information support is provided for automatic driving, smart traffic, smart city and social management.

Description

Technical Field

[0001] This invention belongs to the field of big data applications. By combining big data information with satellite positioning signals, it provides users with accurate positioning. The data used mainly includes regional historical positioning data, user historical positioning data, and multi-source satellite positioning information data. By applying relevant knowledge of machine learning, it provides users with accurate positioning services, overcoming the shortcomings of large errors and unstable signals when directly using satellite signals for positioning. This provides strong information support for autonomous driving, intelligent transportation, smart cities, and social management. Background Technology

[0002] Location services have been applied to all sectors of the national economy, greatly changing people's production and lives, bringing immense convenience, and creating significant economic and social benefits. Location services play a crucial role in ocean shipping, geological exploration, disaster relief, search and rescue, intelligent transportation, and smart city construction. With the maturity and widespread adoption of intelligent driving and IoT technologies, they will play an increasingly important role in future production and daily life.

[0003] Currently, the mainstream positioning methods are primarily based on satellite positioning systems. There are four major navigation and positioning systems in the world: the US Global Positioning System (GPS), Russia's GLONASS system, the European Union's Galileo satellite navigation system, and China's BeiDou system. These systems all locate positions directly or indirectly through ranging. This method relies on satellite signals, but satellite signals are often affected by factors such as the ionosphere and atmosphere, inevitably leading to some positioning errors. Furthermore, this method does not consider ground conditions or social factors, resulting in less than ideal positioning results.

[0004] The positioning method of this invention fully utilizes regional and user historical data, and employs machine learning and data mining techniques to correct satellite positioning information, thereby improving the scientific validity and rationality of the positioning. This invention overcomes the shortcomings of satellite positioning, which does not consider ground information and relevant information about user habits. Summary of the Invention

[0005] This invention discloses a big data-based positioning method and system with privacy protection features, which utilizes regional historical positioning data and user historical positioning data, combined with satellite positioning information, to provide accurate positioning.

[0006] This invention first divides the area to be located into a grid using a two-step method. The data used in this invention mainly includes three parts: first, historical location data for the area; second, historical location data for the user; and third, satellite positioning data.

[0007] The special meshing method uses a two-step process: first, the area to be located is uniformly divided into large square grids, designated R1, R2, ..., R... m , grid R t The length and width of (t=1,2,…,m) are on the order of kilometers. Let R be the side length of the grid. Then, query the historical location database for this area to obtain the grid R. t In the middle, the number of historical location records is Then put R t Divide evenly into sides of length The small grid is given by h, where h is a real number in the interval [2, 15].

[0008] For all R1, R2, ..., R m After completing the above division, the final mesh division result is obtained, and each secondary mesh is denoted as... The side length of each grid is Each grid contains the following number of historical location records for the region. Each grid in the region is obtained based on historical data of the region. The area positioning score is Where i = 1, 2, ..., n.

[0009] Using the two-step partitioning method described above, the area to be located is divided based on the user's historical location data, i.e., the user's historical appearances in this area, resulting in a new grid partitioning result. The side length of each grid is Each grid contains the following number of historical location records for the region. Based on the user's historical location data in this area, we can obtain the location data for this user in each grid cell p within this area. j u Historical positioning score Where j = 1, 2, ..., n u .

[0010] The positioning accuracy is set according to the user's needs, and the user's area is divided into a uniform grid. If the location requested by the user differs from the actual location by no more than 1 meter, then the location area is divided into a grid with a side length of no more than 1 / 2 meter.

[0011] Users continuously request positioning data from different satellite positioning systems. Currently, there are four major global satellite positioning systems: the US GPS, China's BeiDou, Russia's GLONASS, and Europe's Galileo. Each system has its own advantages and disadvantages. Users obtain positioning data multiple times from different systems, resulting in a satellite positioning dataset S. Represents a grid The number of location data included is calculated based on satellite positioning data for each grid cell in this region. The satellite positioning score is Where k = 1, 2, ..., n s .

[0012] For a user's location request, the user first sends multiple location requests to multiple positioning systems consecutively to obtain the user's satellite positioning dataset. Based on the satellite positioning dataset, the satellite positioning score of each cell in the positioning area is calculated. The area location score for each cell is obtained based on the historical location data of the area. Calculate the user's historical location score based on the current user's historical location data. The overall location score for the l-th cell is calculated as follows:

[0013]

[0014] The cell with the highest comprehensive positioning score is selected as the final positioning result. Here, w1 is the empirical positioning coefficient, w2 is the privacy positioning coefficient, and w3 is the satellite positioning coefficient, with w1 + w2 + w3 = 1. The larger the w1 value, the more likely the positioning result is to be located in a densely populated area. The privacy positioning coefficient w2 is used to adjust the degree of user privacy protection. Its value can be positive or negative. The smaller the value, the stronger the privacy protection, and the less likely it is to expose the user's historical positioning records and lifestyle habits. The larger the value, the weaker the privacy protection, and the more likely it is to locate the user in places they have been before. When its value is large, the positioning can be corrected based on the user's lifestyle habits. For example, if a user frequently goes fishing by a river and usually stays on one side, increasing the w2 value can prevent the user from being located on the other side of the river. The satellite positioning coefficient w3 is used to adjust the role of satellite signals in the positioning process. When the satellite positioning signal is strong, the positioning is relatively accurate, and the w3 value can be set to a larger value. Conversely, when the satellite positioning signal is weak, the w3 value can be lowered.

[0015] The setting of w2 is generally based on the user's privacy protection needs. The values ​​of w1 and w3 are obtained through learning. The learning method is as follows: first, the value of w2 is set according to the user's needs. Then, a portion of data is randomly selected from the historical location data as a sample. The values ​​of w1 and w3 are adjusted using the error backpropagation method.

[0016] Another object of the present invention is to provide a computer-readable storage medium including instructions, which, when executed on a computer and a related system, cause the computer and the related system to perform the implementation method of the big data-based positioning method and system with privacy protection function.

[0017] Another objective of this invention is to provide a big data-based positioning method and system that implements the aforementioned privacy-protecting function to control the system.

[0018] In summary, the advantages and effects of this invention are as follows: This invention does not simply utilize satellites for positioning, but rather combines regional historical positioning big data with the user's personal historical positioning data for comprehensive positioning. On the one hand, this improves positioning accuracy; on the other hand, it allows for adjustments to the positioning based on the user's lifestyle, providing privacy protection while also offering personalized services.

[0019] Historical location big data can be used to correct positioning errors caused by poor satellite positioning signals. This is mainly applicable to areas with large differences in traffic volume, such as fields through which highways pass. Most users should be on the highway. If the satellite signal positioning result is a field by the roadside, it is likely due to an error in satellite positioning. The positioning result can be corrected to the highway by using past positioning data.

[0020] The satellite positioning results can be corrected by using the user's historical positioning data. Generally speaking, the user's living area is relatively fixed, and they have their own living habits in a fixed area. Therefore, the satellite positioning results can be corrected by using the user's historical positioning data. This method can improve the positioning accuracy, but it is also easy to expose the user's privacy and living habits. Therefore, this invention introduces a privacy positioning coefficient w2. When the value of w2 is positive, it is mainly used to improve the positioning accuracy. When the value of w2 is negative, it is mainly used to protect the user's privacy. Attached Figure Description

[0021] Figure 1 System Flowchart Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0023] Example 1:

[0024] The regional historical location data mentioned in this invention refers to all data that can be collected within a specified area, including vehicle location data, pedestrian location data, and location data of buildings or natural objects. All location data that can be collected can be added to the regional historical location database. Within the specified area, the collected location records cannot be evenly distributed throughout the area. There may be more location data in a certain part of the specified area, indicating that users in that part need to make more location requests, and the next location request is more likely to be initiated in that part.

[0025] Example 2:

[0026] The two-step meshing method described in this invention refers to: firstly, the specified area in Example 1 is coarsely divided into a first-level mesh, such as dividing the specified area into square meshes with a length and width of 1 kilometer each. Experimental results show that the granularity of the first mesh division of the specified area has little impact on the overall system performance; then, the number of positioning records contained in each first-level mesh is counted, and then the formula is applied... Calculate the granularity at which each grid needs further subdivision, such as dividing it into grids with side lengths of... A square secondary grid, where the symbols represent the same meanings as described in the invention, is used. The size of the secondary grid is affected by the number of records contained in the first-level grid. If a certain level grid contains more positioning records, then according to the formula... The primary grid will be further divided into smaller secondary grids. This division method can provide more accurate positioning for areas with frequent positioning needs. The size of the secondary grid reflects the likelihood of a positioning request being initiated in that area. If the granularity of the secondary grid is small and the side length is short, it means that, based on historical positioning data, the positioning request is more likely to originate from that area; conversely, it means that positioning requests are rarely initiated within that secondary grid, and the likelihood of a positioning request originating from that secondary grid is smaller. Calculate the user's location score for their region.

[0027] Example 3:

[0028] In this invention, the processing of user historical location data adopts the same method as in Example 3, the difference being that the gridded areas are those where the user has previously initiated location requests. The size of the secondary grid reflects the user's lifestyle habits. If the grid of user historical location data in a certain sub-region is small, it indicates that the user frequently appears in that area, suggesting a higher probability that the user's next location request will be issued in that area. On the one hand, this information can be used to correct satellite positioning information and improve positioning accuracy. On the other hand, this information may expose user lifestyle habits and user information to some extent. Therefore, this invention introduces a privacy positioning coefficient w2. When w2 is positive, it essentially uses the user's lifestyle habits to improve positioning accuracy; when w2 is negative, it actually corrects the positioning result, causing the final positioning result to deviate from the area where the current user frequently appears, thereby avoiding the exposure of the user's lifestyle habits and protecting the user's privacy. Calculate the user in each grid p j u Historical positioning score.

[0029] Example 4:

[0030] This invention utilizes satellite data for positioning, essentially employing a voting mechanism to overcome positioning anomalies caused by signal interference during a single positioning attempt. The satellite positioning data incorporates data from different positioning systems, leveraging the advantages of each system by sending multiple location requests to multiple systems and calculating a satellite positioning score based on the acquired data. Specifically, in this embodiment, a uniform grid division method is used, with the grid size depending on the user's positioning needs. If the user requires high positioning accuracy, a relatively small grid is used; if the accuracy requirement is relatively low, a larger grid is used. The number of positioning records falling into each grid is counted, and... Calculate the satellite positioning score.

[0031] Example 5:

[0032] Based on the formulas in Examples 2, 3, and 4, the regional positioning score, historical positioning score, and satellite positioning integral are calculated respectively, and f is applied. l =w1*f l 1 +w2*f l 2 +w3*f l 3 Calculate the overall score; the w2 value is set according to the user's privacy protection requirements; the values ​​of w1 and w3 are generally obtained through learning. Once the w2 value is determined, it can be learned from a portion of the samples using the error backpropagation algorithm.

[0033] In this invention, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented, in whole or in part, as a computer program product, the computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0034] The above description is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. This invention discloses a big data-based positioning method and system with privacy protection features. It utilizes historical positioning data of a region and historical positioning data of the user, combined with satellite positioning information, to provide accurate positioning. The invention first divides the area to be positioned into grids using a two-step method. The data used in this invention mainly includes three parts: first, historical positioning data of the region; second, historical positioning data of the user; and third, satellite positioning data. Positioning scores are calculated on each of the three datasets, and finally, they are fused to obtain the final positioning result.

2. The big data-based positioning method and system with privacy protection function according to claim 1, wherein the grid division method first divides the area to be located into large square grids, namely R1, R2, ..., R... m , grid R t The length and width of (t=1,2,…,m) are on the order of kilometers. Let R be the side length of the grid; then query the historical location database for that area to obtain the grid R. t In the middle, the number of historical location records is Then put R t Divide evenly into sides of length The small grid is given by h, where h is a real number in the interval [2, 15].

3. A big data-based positioning method and system with privacy protection function according to claims 1 and 2, wherein for all R1, R2, ..., R m After completing the above division, the final mesh division result is obtained, and each secondary mesh is denoted as... The side length of each grid is Each grid contains the following number of historical location records for the region. Each grid in the region is obtained based on historical data of the region. The area positioning score is Where i = 1, 2, ..., n.

4. A big data-based positioning method and system with privacy protection function according to claims 1 and 2, employing the above-described two-step partitioning method, divides the positioning area based on the user's historical positioning data, i.e., the user's historical appearance data in the area, to obtain a new grid partitioning result. The side length of each grid is Each grid contains the following number of historical location records for the region. Based on the user's historical location data in this area, we can obtain the location data for this user in each grid cell p within this area. j u Historical positioning score Where j = 1, 2, ..., n u .

5. A big data-based positioning method and system with privacy protection function as described in claim 1, wherein the positioning accuracy is set according to the user's needs, and the user's location area is divided into a uniform grid. The user obtains multiple positioning data from different positioning systems to obtain a satellite positioning dataset S. Represents a grid The number of location data included is calculated based on satellite positioning data for each grid cell in this region. The satellite positioning score is Where k = 1, 2, ..., n s .

6. A big data-based positioning method and system with privacy protection function according to claims 1, 2, 3, 4 and 5, which calculates the satellite positioning score of each cell in the positioning area based on a satellite positioning dataset. The location score for each cell is obtained based on the historical location data of the area. Calculate the user's historical location score based on the current user's historical location data. The overall location score for the l-th cell is calculated as follows: The cell with the highest overall positioning score is selected as the final positioning result. Here, w1 is the empirical positioning coefficient, w2 is the privacy positioning coefficient, and w3 is the satellite positioning coefficient, with w1 + w2 + w3 = 1.

7. In the big data-based positioning method and system with privacy protection function according to claim 6, the setting of w2 is based on the user's privacy protection needs, and the values ​​of w1 and w3 are obtained through learning. The learning method is as follows: first, the value of w2 is set according to the user's needs, and then, a portion of data is randomly selected from the historical positioning data as a sample, and the values ​​of w1 and w3 are adjusted using the error backpropagation method.

8. A computer-readable storage medium comprising instructions, when executed on a computer, causing the computer to perform a big data-based positioning method and system with privacy protection features as described in claim 1.

9. A control system for implementing the big data-based positioning method and system with privacy protection function as described in claim 1.

Images