Edge computing device deployment method for roadside spatiotemporal data collection

Abstract

The application belongs to the field of roads. The purpose of the application is to solve the problems of complex calculation and low matching rate with sensor node position in the existing deployment of edge computing device nodes, obtain all sensor nodes to be covered on the road, divide all sensor nodes into layers by using KD-Tree space, draw a circle with each sensor node as the center according to a preset radius, sort the coverage potential of all circles in each layer from high to low, start searching from the first circle in each layer, set the position of the edge computing device according to whether the first circle intersects with other circles, the coverage potential and the coverage score, and complete the deployment of all edge computing devices according to the query order of the circle and the deployment method. The application is used for deploying edge computing devices.

Description

Technical Field

[0001] This invention belongs to the field of road infrastructure and relates to the deployment of edge computing devices. Background Technology

[0002] With the continuous advancement of intelligent transportation and smart city construction, roadside spatiotemporal data collection technology has become a key link supporting traffic management and urban planning.

[0003] In real-world projects, the deployment of edge computing device nodes is increasingly becoming an important means of improving data processing capabilities. Taking the Shenzhen Nansha Pearl Bay project as an example, this project adopted high-performance Ascend 310 chip edge computing devices to cover sensor nodes distributed along the roadside, thereby achieving real-time data processing and efficient collection, ensuring the accuracy and precision of data acquisition.

[0004] However, existing edge computing device node deployment methods still have significant shortcomings. Real-world projects often employ a coarse-grained grid or random distribution strategy, determining coverage areas through simple distance calculations. This method cannot be specifically optimized based on the specific location and distribution of sensor nodes, thus exposing many problems in practical applications, including but not limited to increased economic costs, impacted data processing timeliness, and increased traffic risks. For example, in Germany, roadside units on highways are deployed according to a uniform standard, yet some nodes have a utilization rate of less than 10% during off-peak hours (80%).

[0005] The root cause of the above problems lies in the fact that the extensive deployment method fails to accurately plan the location of sensor nodes when deploying edge computing device nodes, making it difficult to adapt to the actual needs of complex roadside environments. Therefore, roadside spatiotemporal data collection urgently needs a more adaptive edge computing device deployment method.

[0006] Existing edge computing device node deployment methods using unit circle coverage, such as greedy algorithms and linear programming, can theoretically provide coverage solutions, but have limitations in practical applications. When traditional greedy algorithms process 200 sensor nodes, the matching rate between edge computing devices and sensor nodes is low, with a missed match rate of 12.3% and a false match rate of 8.7%. Therefore, there is an urgent need for an efficient and optimized edge computing device deployment method that can reduce computational complexity while ensuring coverage and adapting to the needs of large-scale computing node scenarios. Summary of the Invention

[0007] The purpose of this invention is to address the problems of complex calculations and low matching rate between existing edge computing device nodes and sensor node locations, and to propose an edge computing device deployment method for roadside spatiotemporal data collection.

[0008] A method for deploying edge computing devices for roadside spatiotemporal data collection, the method comprising the following:

[0009] Step 1: Obtain all sensor nodes to be covered on the road, and use the KD-Tree space to layer all sensor nodes;

[0010] Step 2: Draw a circle with each sensor node as the center, according to the preset radius;

[0011] Step 3: Calculate the total coverage score and the dispersion of each circle in the i-th layer. The total coverage score of each circle is obtained by summing the coverage scores of the multiple arc segments into which it is divided by other circles, or by taking the coverage score of the entire circle that is not divided by other circles as the total coverage score. Based on the total coverage of each circle and the dispersion of the corresponding circle, obtain the coverage potential of each circle. Sort the coverage potential of all circles in the i-th layer from high to low, and the initial value of i is 1.

[0012] Step 4: Find the j-th circle in the i-th layer, with the initial value of j being 1.

[0013] If the j-th circle intersects with other circles, select the circle with the greatest coverage potential from the intersecting circles. On the arc of the circle with the maximum coverage score, randomly select a point to deploy an edge computing device. After deployment, draw a circle with the deployment location of the edge computing device as the center and a preset radius. Delete the sensor nodes on the current layer and other layers covered by this circle.

[0014] If the j-th circle does not intersect with other circles, take any point on the arc of the circle and place an edge computing device. After placement, take the placement position of the edge computing device as the center and the preset radius to draw a circle. Delete the sensor nodes on the current layer and other layers covered by the circle.

[0015] Step 5: Determine if j is equal to the total number of sensor nodes in the i-th layer. If not, set j = j + 1 and proceed to step 6. If yes, proceed to step 7.

[0016] Step 6: Determine whether the j-th circle has been deleted. If yes, set j = j + 1 and return to step 6. If no, execute step 4.

[0017] Step 7: Determine if i is equal to the total number of layers. If not, set i = i + 1 and execute step 3. If yes, all sensor nodes are deleted and the deployment of all edge computing devices is completed.

[0018] Preferably, in step 3, the process of obtaining the total coverage score for each circle is as follows:

[0019] If a circle intersects with other circles, then the circle is divided into multiple arc segments by the other circles. The coverage score of each arc segment is calculated based on the angle of each arc segment and the number of circles covering that arc segment. The coverage scores of all arcs on a circle are added together to obtain the total coverage score of the circle.

[0020] If a circle does not intersect with any other circle, its total coverage score is 0.

[0021] Preferably, in step 3, the dispersion is expressed as:

[0022]

[0023] In the formula, R i w represents the dispersion of the i-th circle. k Let θ be the coverage score of the k-th arc segment. end,k Let θ be the ending angle of the k-th arc. start,k Let be the starting angle of the k-th arc segment.

[0024] Preferably, the method further includes step 8: drawing a distribution map using the node locations of all deployed edge computing devices.

[0025] Preferably, the distribution map is drawn using a Python visualization tool.

[0026] Preferably, the preset radii in steps 2 and 4 are equal.

[0027] Preferably, the specific process for obtaining the coverage potential of each circle based on the total coverage of each circle and the dispersion of the corresponding circle is as follows:

[0028] The coverage potential of each circle is obtained by multiplying the total coverage of each circle by the dispersion of the corresponding circle.

[0029] The beneficial effects of this invention are:

[0030] First, the sensor node set is hierarchically divided. Then, the coverage potential of the circle formed by each sensor node is evaluated based on the intersection relationship between the sensor nodes and the arc coverage index. Finally, the optimal placement location of the edge computing device is selected through a hierarchical greedy algorithm, and the coverage visualization results are generated to verify the optimization effect. This invention utilizes each edge computing device to process the collected data of the sensor nodes covered by the edge computing device, and the matching rate between the location of the edge computing device deployed in this invention and the covered sensor nodes is high.

[0031] The core idea of ​​this invention lies in integrating spatial indexing technology, geometric coverage analysis, and statistical methods to construct a hierarchical and adaptive framework for optimizing edge computing device deployment. Specifically, it accelerates node querying and region partitioning through a KD-Tree hierarchical spatial index, uses arc-based coverage potential analysis to locate the optimal candidate coverage direction, and introduces a cyclic integral index from cyclic statistics to quantify coverage dispersion, guiding a greedy algorithm to select the globally optimal solution. This comprehensive strategy not only significantly reduces computational complexity but also ensures the uniformity of coverage distribution through quantitative indicators, overcoming the limitations of local optimization in traditional methods.

[0032] Based on practical needs, this invention summarizes the problem as follows: given an edge computing device with an existing coverage radius, several data-generating sensor nodes, and their location information, the goal is to find the minimum number of computing nodes and place them in selected locations to ultimately cover all sensor nodes. This problem can be reduced to a unit circle coverage problem.

[0033] This invention uses a recursive partitioning and hierarchical management method of KD-Tree hierarchical index to divide sensor nodes into layers, which improves spatial query efficiency and is more suitable for scenarios with a large number of sensor nodes and a large number of computing nodes to be deployed.

[0034] Traditional methods often rely on brute-force calculation of the number of coverable nodes. However, this invention, through arc coverage potential analysis, can not only accurately quantify the coverage direction of nodes based on polar angle segmentation and vector synthesis, but also effectively avoid the problem of traditional methods getting stuck in local optima. Furthermore, it can reduce computational costs and improve deployment efficiency.

[0035] This invention introduces statistical methods and creatively proposes using the coverage vector magnitude as a discreteness index to guide global optimization, ensuring the uniformity of coverage distribution and algorithm convergence.

[0036] This invention provides an efficient and low-cost sensor deployment optimization method for roadside spatiotemporal data acquisition by deeply integrating KD-Tree spatial indexing, circular arc coverage analysis, and cyclic integral index. Attached Figure Description

[0037] Figure 1 A flowchart illustrating the deployment method of edge computing devices for roadside spatiotemporal data collection.

[0038] Figure 2 This is a comparison diagram of the edge computing device deployment method of this application and existing edge computing device deployment methods. Figure 2 (a) is a layout diagram of edge computing devices using a conventional greedy algorithm. Figure 2 (b) is a layout diagram of the edge computing device of the present invention. Figure 2(c) is a diagram of the edge computing device deployment for a conventional legacy algorithm;

[0039] Figure 3 for Figure 2 (b) shows a magnified view of a portion of the image. Figure 3 (a) is a diagram showing the intersection of multiple circles. Figure 3 (b) is a diagram showing a circle divided into multiple arcs by other circles;

[0040] Figure 4 This is a diagram showing the intersection of two circles formed by two sensor nodes.

[0041] Figure 5 This is a comparison chart of coverage methods. Figure 5 (a) is a layout diagram of edge computing devices using a conventional greedy algorithm. Figure 5 (b) is a layout diagram of the edge computing device of the present invention. Figure 5 (c) is a diagram of the edge computing device deployment for a conventional legacy algorithm;

[0042] Figure 6 This is a comparison chart showing the number of edge computing devices deployed using existing methods and those deployed in this application. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0045] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0046] Example:

[0047] A method for deploying edge computing devices for roadside spatiotemporal data collection, the method comprising the following:

[0048] Step 1: Obtain all sensor nodes to be covered on the road, and use the KD-Tree space to layer all sensor nodes;

[0049] Step 2: Draw a circle with each sensor node as the center, according to the preset radius;

[0050] Step 3: Calculate the total coverage score and the dispersion of each circle in the i-th layer. The total coverage score of each circle is obtained by summing the coverage scores of the multiple arc segments into which it is divided by other circles, or by taking the coverage score of the entire circle that is not divided by other circles as the total coverage score. Based on the total coverage of each circle and the dispersion of the corresponding circle, obtain the coverage potential of each circle. Sort the coverage potential of all circles in the i-th layer from high to low, and the initial value of i is 1.

[0051] Step 4: Find the j-th circle in the i-th layer, with the initial value of j being 1.

[0052] If the j-th circle intersects with other circles, select the circle with the greatest coverage potential from the intersecting circles. On the arc of the circle with the maximum coverage score, randomly select a point to deploy an edge computing device. After deployment, draw a circle with the deployment location of the edge computing device as the center and a preset radius. Delete the sensor nodes on the current layer and other layers covered by this circle.

[0053] If the j-th circle does not intersect with other circles, take any point on the arc of the circle and place an edge computing device. After placement, take the placement position of the edge computing device as the center and the preset radius to draw a circle. Delete the sensor nodes on the current layer and other layers covered by the circle.

[0054] Step 5: Determine if j is equal to the total number of sensor nodes in the i-th layer. If not, set j = j + 1 and proceed to step 6. If yes, proceed to step 7.

[0055] Step 6: Determine whether the j-th circle has been deleted. If yes, set j = j + 1 and return to step 6. If no, execute step 4.

[0056] Step 7: Determine if i is equal to the total number of layers. If not, set i = i + 1 and execute step 3. If yes, all sensor nodes are deleted and the deployment of all edge computing devices is completed.

[0057] Specifically, in step 1, a KD-Tree hierarchical spatial index is used to accelerate node querying and region partitioning. Arc-based coverage potential analysis is used to locate the optimal candidate coverage direction, and a cyclic integral index from cyclic statistics is introduced to quantify coverage dispersion, guiding the greedy algorithm to select the globally optimal solution. This comprehensive strategy not only significantly reduces computational complexity but also ensures the uniformity of coverage distribution through quantitative indices, overcoming the limitations of local optimization in traditional methods.

[0058] After KD-Tree stratifies all sensor nodes, a coordinate graph will display all the stratified sensor nodes. Figure 2The red stars represent each sensor node. Blue circles are drawn with each sensor node as the center and a preset radius. Edge computing devices are first deployed on the sensor nodes with the greatest coverage potential in the first level. If... Figure 3 The circle in the bottom left corner is the circle with the highest coverage potential among the three intersecting circles. Then, find the arc with the highest coverage score from this circle. Figure 3 (b) shows that the circle is divided into four arcs by other intersecting circles, represented by yellow, brown, blue, and red respectively. The coverage score of each arc is calculated. It is found that the brown arc has the highest coverage score because it is inside the three circles. Therefore, a green edge computing device is placed at any point on the colored arc. After placement, a green circle is drawn. If the green circle covers all three red stars, the three stars and the corresponding blue circle are removed. This process is repeated to query the second sensor node in this layer (the sensor node with the second largest coverage potential). This process is repeated until edge computing devices are deployed on all sensor nodes. After deployment, each edge computing device is used to process the data of the covered sensors.

[0059] pass Figure 2 and Figure 5 It can be seen that the coverage quality of this embodiment is significantly improved compared to existing algorithms. Furthermore... Figure 6 The number of computing nodes required by the method proposed in this embodiment was compared with that of existing methods using data quantification. It can be seen that the method is significantly better than the existing methods in most cases. The number of computing nodes in this embodiment is significantly less than the number of computing nodes in existing deployments. Therefore, this embodiment has a faster deployment speed, fewer computing nodes, and better performance.

[0060] Further specifying, in step 3, the process of obtaining the total coverage score for each circle is as follows:

[0061] If a circle intersects with other circles, then the circle is divided into multiple arc segments by the other circles. The coverage score of each arc segment is calculated based on the angle of each arc segment and the number of circles covering that arc segment. The coverage scores of all arcs on a circle are added together to obtain the total coverage score of the circle.

[0062] If a circle does not intersect with any other circle, its total coverage score is 0.

[0063] Further specifying, in step 3, the dispersion is expressed as:

[0064]

[0065] In the formula, R i w represents the dispersion of the i-th circle. k Let θ be the coverage score of the k-th arc segment. end,k Let θ be the ending angle of the k-th arc.start,k Let be the starting angle of the k-th arc segment.

[0066] Further defining the method, step 8 is: drawing a distribution map using the node locations of all deployed edge computing devices.

[0067] To further specify, the distribution map was drawn using a Python visualization tool.

[0068] Furthermore, the preset radii in steps 2 and 4 are equal.

[0069] Further specifying, the specific process for obtaining the coverage potential of each circle based on its total coverage and the corresponding circle's dispersion is as follows:

[0070] The coverage potential of each circle is obtained by multiplying the total coverage of each circle by the dispersion of the corresponding circle.

[0071] Specifically, it is defined as follows: Given a fixed set of geographic coordinates of sensor nodes, P = {p1, p2, ..., p...} n The edge computing device has a coverage radius of r, and each edge computing device is called a computing node. A set of computing nodes C = {c1, c2, ..., c...} needs to be selected. k}, such that: each sensor node p i At least one computing node c i Coverage, i.e., ||p i -c j ||≤r; Minimize the number of sensors used, k; Ensure uniform coverage distribution, avoiding localized dense or sparse areas.

[0072] First, a KD-Tree structure is recursively used to partition the sensor node coordinates. The sensor node set is alternately partitioned along the x-axis and y-axis using the median, generating a multi-level tree structure. By recording the index information of the sensor nodes at each level, a hierarchical dictionary is formed. This dictionary actually reflects the mapping from depth to the sensor node list, allowing the algorithm to prioritize processing the leaf node regions of the KD-Tree, reducing redundant computation. Compared to traditional brute-force search, the KD-Tree reduces the complexity of neighbor lookup from O(n^2) to O(n^2). 2 The value decreases to 0(nlogn).

[0073] Secondly, in determining the coverage potential of each node, this embodiment abandons simple distance measurement and instead uses polar angle segmentation and arc analysis to accurately locate the coverage direction. For the j-th sensor node, KD-Tree is used to quickly query neighboring nodes with a distance less than 2r, and the relative polar angle θ is calculated. ij And generate the covering arc segment. For example, node p i With neighbor p jWhen the spacing is 1.5r, the span of the covered arc is 2arccos(1.5r / 2r) = 120°, and the midpoint angle θ opt This is the optimal coverage direction.

[0074] This embodiment also creatively introduces the result vector from cyclic statistics. By synthesizing the arc segments of all neighbors, a covering vector is generated, with a magnitude R. i The calculation formula is:

[0075]

[0076] The indicator R i Ideal for quantifying the concentration of coverage direction: R i →1 indicates a high concentration of coverage along the direction, making it suitable as a candidate center; R i →0 indicates that the coverage is scattered and redundant deployment should be avoided.

[0077] Finally, this embodiment also addresses the problem that traditional covering methods easily get stuck in the search difficulty when neighboring points are dense, proposing a hierarchical greedy covering method. It innovatively introduces the concept of cyclic integrals from cyclic statistics, using the coverage vector magnitude R... i The core decision-making indicator is used. The specific process is as follows: Starting from the deepest layer of the KD-Tree, uncovered nodes are screened layer by layer and their R is calculated. i Value, prioritize R i The largest sensor node p j As a candidate. According to p j Determine the optimal center position c of the covered arc segment. j =p j +r(cosθ opt sinθ opt This ensures maximum coverage. Upon deployment of each sensor, its covered nodes are immediately queried via the KD-Tree, and the uncovered set is updated. Compared to traditional greedy algorithms, the cyclic integral metric effectively balances coverage density and uniformity, avoiding local optima traps.

[0078] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A method for deploying edge computing devices for roadside spatiotemporal data collection, characterized in that, The method includes the following: Step 1: Obtain all sensor nodes to be covered on the road, and use the KD-Tree space to layer all sensor nodes; Step 2: Draw a circle with each sensor node as the center, according to the preset radius; Step 3: Calculate the total coverage score and the dispersion of each circle in the i-th layer. The total coverage score of each circle is obtained by summing the coverage scores of the multiple arc segments into which it is divided by other circles, or by taking the coverage score of the entire circle that is not divided by other circles as the total coverage score. Based on the total coverage score of each circle and the dispersion of the corresponding circle, obtain the coverage potential of each circle. Sort the coverage potential of all circles in the i-th layer in descending order. The initial value of i is 1. Step 4: Find the j-th circle in the i-th layer, with the initial value of j being 1. If the j-th circle intersects with other circles, select the circle with the greatest coverage potential from the intersecting circles. On the arc with the highest coverage score on this circle, randomly select a point to deploy an edge computing device. After deployment, draw a circle with the deployment location of the edge computing device as the center and a preset radius. Delete the sensor nodes on the current layer and other layers covered by this circle. If the j-th circle does not intersect with other circles, take any point on the arc of the circle and place an edge computing device. After placement, take the placement position of the edge computing device as the center and the preset radius to draw a circle. Delete the sensor nodes on the current layer and other layers covered by the circle. Step 5: Determine if j is equal to the total number of sensor nodes in the i-th layer. If not, set j = j + 1 and proceed to step 6. If yes, proceed to step 7. Step 6: Determine whether the j-th circle has been deleted. If yes, set j = j + 1 and return to step 6. If no, execute step 4. Step 7: Determine if i is equal to the total number of layers. If not, set i = i + 1 and execute step 3. If yes, all sensor nodes are deleted and the deployment of all edge computing devices is completed. The specific process for obtaining the coverage potential of each circle based on its total coverage score and the corresponding circle's dispersion is as follows: The coverage potential of each circle is obtained by multiplying the total coverage score of each circle by the dispersion of the corresponding circle.

2. The method for deploying edge computing devices for roadside spatiotemporal data collection according to claim 1, characterized in that, In step 3, the process of obtaining the total coverage score for each circle is as follows: If a circle intersects with other circles, then the circle is divided into multiple arc segments by the other circles. The coverage score of each arc segment is calculated based on the angle of each arc segment and the number of circles covering that arc segment. The coverage scores of all arcs on a circle are added together to obtain the total coverage score of the circle. If a circle does not intersect with any other circle, its total coverage score is 0.

3. The method for deploying edge computing devices for roadside spatiotemporal data collection according to claim 1, characterized in that, In step 3, the dispersion is divided into the dispersion when a circle intersects with other circles and the dispersion when a circle does not intersect with other circles. The dispersion when a circle does not intersect with other circles is 0, and the dispersion when a circle intersects with other circles is expressed as follows: In the formula, For the first The dispersion of a circle, For the first The score for the coverage of the arc segment. For the first The ending angle of the arc segment. For the first The starting angle of the arc segment.

4. The method for deploying edge computing devices for roadside spatiotemporal data collection according to claim 1, characterized in that, The method also includes step 8: drawing a distribution map using the node locations of all deployed edge computing devices.

5. The method for deploying edge computing devices for roadside spatiotemporal data collection according to claim 4, characterized in that, The distribution map was drawn using a Python visualization tool.

6. The method for deploying edge computing devices for roadside spatiotemporal data collection according to claim 1, characterized in that, The preset radii in steps 2 and 4 are equal.

Images