Radio propagation irregular coefficient detection method

Abstract

The invention discloses a method for detecting a radio propagation irregular coefficient, which comprises the following steps of: enabling measuring equipment to move around a measured signal generating device at a specified distance d through visual ranging, measuring signal intensity G lambda (d) at each angle lambda, and further detecting the radio propagation irregular coefficient k lambda; the influence of Gaussian white noise on a measurement result is reduced in a manner of averaging through multiple measurements; meanwhile, a motion method of the measurement equipment is designed according to the radio parameter irregular coefficient detection requirement, so that the measurement equipment actively rectifies deviation in the motion process, and the detection result is prevented from being influenced by an overlarge motion path error. The method is low in application cost, convenient to implement, high in precision and wide in application scene.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method for detecting a radio propagation irregularity coefficient. Background Art

[0002] With the continuous advancement of wireless communication technology, location detection is crucial for applications such as routing, data collection and analysis, and navigation. Accurate location information enables networks and related applications to achieve high operational efficiency. However, inaccurate location information, especially information with significant deviations, can significantly impact the normal operation of networks and related applications, and can even paralyze some networks. Driven by profit, attackers constantly seek to attack various networks. Attacks targeting location information can occur both during the positioning process and during the use of location information. Typical attack methods include pseudorange attacks and pseudocoordinate attacks. Attacks targeting location information can not only disrupt network operation but also endanger device security. Currently, there is no systematic method for detecting radio propagation irregularities, but traditional radio parameter detection methods can be used to indirectly measure the corresponding parameters. Possible methods for measuring the radio propagation irregularity coefficient using traditional detection methods include two categories. One method uses traditional detection equipment to independently detect radio propagation conditions in all directions and then derive the radio propagation irregularity coefficient. This method requires manual anchoring of measurement points, resulting in low accuracy and efficiency, making it difficult to meet practical application requirements. The other method uses a large-scale antenna array to detect radio propagation conditions in multiple directions at once, thereby inferring relevant parameters. The main problem with this method is that it is difficult to ensure that the distance between each array element and the signal generator is consistent. Although the radio propagation irregularity coefficient is statistically applicable to any distance in the same environment, since the coefficient itself is affected by factors such as the propagation medium, there are still differences in the radio propagation irregularity coefficient at different distances in the same direction from the signal generator. Accumulating such differences will cause large measurement errors. Summary of the Invention

[0003] Purpose of the invention: The purpose of the present invention is to provide a method for detecting a radio propagation irregularity coefficient to help users deduce the accurate omnidirectional antenna coverage area, thereby improving the working efficiency of the network and related applications, and at the same time helping users use the measured radio propagation irregularity coefficient to protect the security of location information.

[0004] Technical solution: A method for detecting a radio propagation irregularity coefficient, wherein the radio propagation irregularity coefficient detection system includes an on-board processing terminal, an application processing terminal, and a user terminal. The user terminal can customize distance, speed, and accuracy, and then transmit the signal to the interface layer of the on-board processing terminal; the on-board processing terminal completes data measurement according to requirements and transmits it to the application processing terminal via the sending layer; the application processing terminal processes the data, then draws it into a signal strength contour map and optimizes it, and finally outputs the data and graphics to the user terminal, and finally outputs the graphics; the on-board processing terminal includes a signal measurement system, a motion system, a visual ranging system, a correction system, and a data processing system; the steps include the following:

[0005] S1, the user terminal sets the vehicle speed, m measurement distances, reference distance d0, and ranging error threshold T d , image similarity threshold T P , acceptable angle error δ, correction interval t α Send to the vehicle processing terminal;

[0006] S2, the signal measurement system calculates a matching measurement frequency based on the distance and vehicle speed;

[0007] S3, the visual ranging system controls the movement and stop of the vehicle according to the set reference distance d0 and the actual distance d in combination with the motion system; the correction system uses the visual ranging system to detect whether it is yawed, and adjusts the steering angle in combination with the motion system to achieve correction; the distance from the target signal generating device is measured in real time using binocular ranging The initial distance d' between the car and the target signal generator, and the forward distance calculated in real time based on the car's speed The correction system includes a radial correction system and an angular correction system;

[0008] S4, the data processing system calculates m groups of irregularity coefficients k according to the reference distance d0, the actual distance d and the measured signal strength λ and theoretical distance d λ ;

[0009] S5, the m groups of theoretical distance d λ And the position information measured by binocular ranging is sent to the application processing end;

[0010] S6, theoretical distance d from the application processing end λ Draw a contour map of each group of signal strength G0(d), and finally transmit the graph and parameters to the user end;

[0011] S7, according to the irregularity coefficient k corresponding to the m group λ Weibull parameter fitting was performed to obtain the scale parameters and shape parameters of each group.

[0012] Furthermore, in step S2, during the movement, the radial correction system ensures that the target signal generating device is at the horizontal center of the camera image. If the angle between the camera and the vehicle body is less than the threshold value T d , then measure the signal strength and adjust the steering angle of the wheel according to the angle of the camera; if the angle between the camera and the vehicle body is greater than the threshold value T d At this time, there is no need to measure the signal strength, only the visual ranging result is needed. Adjust the steering angle according to the relationship with the set distance; repeat the above operation every time the vehicle moves forward one degree to ensure that the vehicle runs around the target signal generating device at a fixed distance.

[0013] Furthermore, in step S2, during the movement, the angle correction system will collect corresponding angle records, and the car will move in a circle multiple times until there is at least one record for each angle λ′; when there are more than one record results for the same angle, the average value is taken to complete the signal strength G of each angle λ′ at the reference distance. λ′ (d0) measurement.

[0014] Further, in step S3, the irregularity coefficient k is calculated λ and theoretical distance d λ The implementation steps are as follows:

[0015] S31, the trolley moves around the circle with the radius d unchanged to the 1° direction, measures the signal strength G1(d) there, and calculates the irregularity coefficient k1 in the 1° direction by combining the known G1(d0) and loss parameter η;

[0016] S32, let λ = 0 and substitute k1 into the following formula to calculate the theoretical distance d′1. The theoretical distance d′1 is the distance from the point with the signal strength G0(d) in the 1° direction to the target signal generator. This distance is used to draw lines connecting points with the same signal strength in each direction, indicating the actual coverage area of ​​the target signal generator at that signal strength.

[0017]

[0018] Among them, G λ (d) is the received signal strength at distance d in the direction of λ, d0 is the reference distance, η is the path loss, F σ,λ is zero-mean Gaussian white noise with standard deviation σ in the λ direction; k λ is the irregularity coefficient in the λ direction;

[0019] S33, the trolley moves in a circle with the actual distance d as the radius. Every time it moves forward one degree, it repeats steps S31 and S32 to measure and calculate the irregularity coefficient k in each direction. λ and theoretical distance d′ λ, and repeat this cycle until all m groups of data, including the data measured at point d, are measured.

[0020] Furthermore, radial positioning correction is performed in all m groups of signal strength measurements.

[0021] Compared with the prior art, the present invention has the following significant effects:

[0022] 1. The present invention uses an intelligent vehicle to not only quickly detect the multi-directional signal strength of any node in various environments, and then detect and calculate the radio propagation irregularity coefficient, but also further analyze the irregularity based on the drawn signal strength contour map and the fitted Weber parameter. It can adapt to measurement tasks in most environments and can reduce most manual operations and debugging;

[0023] 2. The present invention combines visual ranging and deep learning, and uses visual positioning to automatically complete the measurement of radio propagation parameters in all directions of the omnidirectional antenna at the same distance, and finally calculates the radio propagation irregularity coefficient; it ensures that it can accurately lock the target signal generating device, greatly improving the accuracy of ranging and measurement data; and the present invention has the advantages of low application cost, wide application scenarios, high precision, small error, and easy operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is the principle diagram of binocular ranging;

[0025] Figure 2 This is a flow chart of a binocular ranging system based on neural network training;

[0026] Figure 3 Schematic diagram of the application scenario of the present invention;

[0027] Figure 4 This is a structural diagram of the signal measurement system of the present invention;

[0028] Figure 5 This is a flow chart of the radial deviation correction system of the present invention;

[0029] Figure 6 This is a flow chart of the angle correction system of the present invention;

[0030] Figure 7 This is a schematic diagram of the vehicle-mounted processing terminal architecture of the present invention;

[0031] Figure 8 Schematic diagram of the overall architecture of the method of the present invention;

[0032] Figure 9 This is a contour map of signal intensity detected by the present invention;

[0033] Figure 10 4 is an implementation flow chart of the method of the present invention. DETAILED DESCRIPTION

[0034] The application will be described in further detail below in connection with the accompanying drawings and detailed description.

[0035] The application provides a signal strength intelligent detection device based on visual ranging, which is used for detecting radio propagation irregularities. The device is carried by an intelligent vehicle, and a binocular camera of the device can identify a signal generating device through neural network training and can rotate omnidirectionally on a horizontal plane. Due to the small size and flexibility of the intelligent vehicle, the device can not only realize rapid detection of multi-directional signal strength of any node in various environments and further detect and calculate a radio propagation irregularity coefficient, but also can accurately locate a target signal generating device and perform ranging in combination with the camera trained by the neural network model, thereby ensuring detection accuracy.

[0036] Figure 1 The principle of typical binocular ranging is shown. L and R points represent left and right cameras of a binocular camera, P represents an object observed by the cameras, T is the distance between left and right optical centers, f is a focal length, A and B are mapping points of the object P on left and right imaging planes, respectively. L and X R are the distances from the imaging points of P on the imaging screen to the projection points of the two lens optical axes, respectively, and the difference between the two is denoted as a parallax d; Z P is the vertical distance from the object to the camera, i.e. a depth. From Figure 1 PAB and PLR are similar triangles, and the depth Z P can be calculated in combination with known quantities f, d and T. It can be seen that the binocular camera is similar to the human eye, and the distance of an object can be perceived through a parallax. The farther the object is, the smaller the parallax is; on the contrary, the larger the parallax is. Therefore, distance perception of the binocular system for a target signal generating device is an absolute measurement, rather than an estimation. The measurement of depth by monocular ranging is only an estimation, and the accuracy is low. Therefore, although the binocular ranging is technically complicated, the accuracy is much higher than that of monocular ranging.

[0037] Figure 2A basic ranging system flowchart is presented, comprising a neural network component and a ranging process. The key is selecting an appropriate neural network model to train the camera. The appearance of typical signal generators is used as the training and validation sets, and a test set consisting of more common device appearances is combined to complete neural network training and optimization. This allows the camera on a smart car to accurately identify signal generators using deep learning techniques. After locking onto the target signal generator, the binocular camera captures the image and performs preprocessing, including feature extraction, convolution output, and sampling. Image semantic segmentation is then performed, and the image is edge-processed. The focal length is calculated using contour coordinates and width, and the distance from the camera to the target signal generator is calculated. This is then measured in real time using a loop structure. The present invention builds a visual ranging system based on this approach.

[0038] The present invention provides a method for detecting radio propagation irregularity coefficients, which can be applied in the following scenarios: Figure 3 shown.

[0039] The signal strength measurement system of the present invention is as follows Figure 4 As shown, the smart car is equipped with a signal receiving device. When the processor receives the speed and radius set by the user, it calculates the measurement frequency and feeds it back to the signal receiving device. The signal receiving device begins to detect the strength of the signal emitted by the target signal generating device in real time, and stores it in the on-board processing end waiting for subsequent calculation operations.

[0040] The flow chart of the radial deviation correction system of the present invention is as follows: Figure 5 During the movement, always ensure that the target signal generating device is in the horizontal center of the camera screen. If the difference between the visual ranging result and the set distance is less than the threshold value T d , then measure the signal strength and adjust the steering angle of the wheel according to the angle of the camera (if the angle between the camera and the vehicle's direction of travel is greater than 90°, increase the vehicle's steering angle; otherwise, reduce the steering angle); if the difference between the visual distance measurement result and the set distance is greater than the threshold value T d , at this time, the signal strength is not measured, only the visual ranging result is needed Adjust the steering angle in relation to the set distance ( If the distance is less than the set distance, reduce the steering angle; If the vehicle moves forward one degree, the steering angle is increased. Repeat the above operation every time the vehicle moves forward one degree to ensure that the vehicle moves around the target signal generating device at a fixed distance, thus avoiding errors in the vehicle's route.

[0041] The flow chart of the angle correction system of the present invention is as follows: Figure 6, the corresponding acquisition position (angle) is recorded during the movement, the trolley circularly moves multiple times until at least one record is obtained at each angle (λ'±δ) °, when the record result at the same angle is more than one, the average value is taken, so as to realize the signal intensity G of each angle λ' at the reference distance λ′ (d0) measurement, note: here λ' is used only to distinguish the angle at the distance signal generator d.

[0042] The vehicle-mounted processing end architecture diagram of the application is shown in Figure 7 The main parts are signal measurement system, movement system, visual distance measurement system, deviation correction system and data processing system (mainly used for calculating related data and drawing contour lines at the same signal intensity). The signal measurement system calculates the matching measurement frequency according to the distance and vehicle speed, the visual distance measurement system controls the movement and stop of the trolley according to the set reference distance d0 and the actual distance d, the deviation correction system detects whether the trolley deviates by using the visual distance measurement system, and adjusts the steering angle to realize deviation correction in combination with the movement system, and finally the data processing system calculates the irregularity coefficient k λ and the theoretical distance d λ according to d, d0 and the measured signal intensity.

[0043] The overall architecture diagram of the application is shown in Figure 8 The main parts are vehicle-mounted processing end, application processing end and user end. The user end can customize the distance, speed and accuracy, and then transmit the signal to the interface layer of the vehicle-mounted processing end. The vehicle-mounted processing end measures the data according to the requirements, and then transmits the data to the application processing end through the sending layer. The application processing end processes the data through the software, draws the signal intensity contour line and optimizes the processing, and finally outputs the data and the figure to the user end. The final output figure is shown in Figure 9 .

[0044] The implementation flowchart of the application is shown in Figure 10 Taking the detection of the irregularity coefficient k λ in the direction of λ as an example (wherein λ∈[0,360) and is an integer), the implementation flowchart includes the following steps:

[0045] Step 1, the application calculates the radio propagation irregularity coefficient by using the RSS model, and the formula is as follows:

[0046]

[0047] Wherein, G λ (d) is the received signal intensity at the distance d in the direction of λ, G λ (d0) is the received signal intensity at the distance d0 in the direction of λ, d0 is the reference distance, η is the path loss k λ represents the irregularity coefficient; F σ,λis Gaussian noise with a standard deviation of σ in the λ direction. In the Line-of-Sight (LoS) environment, since σ is small, F σ,λ The value of k is usually small, and an incorrect estimate of random noise may seriously increase k λ Therefore, when calculating k λ The random noise factor is ignored, that is,

[0048]

[0049] Among them, G λ (d0) represents the received signal strength at a distance d0 in the λ direction.

[0050] Step 2: Before the first measurement, the user terminal sets the speed, m groups of measurement distances, reference distance d0 (usually 1 meter), and ranging error threshold T d , image similarity threshold T P , acceptable angle error δ, correction interval t α Send to the vehicle processing end.

[0051] Step 3: The binocular camera rotates the servo and determines that it is facing the target signal generating device through image semantic segmentation, which is set as point A. The binocular ranging is used to determine the distance from itself to A and record it.

[0052] Step 4: The car adjusts the front of the car to align with the direction of the binocular camera and starts moving forward. At the same time, the binocular ranging is used to measure the distance to the target signal generating device in real time. The initial distance d' between the car and the target signal generator and the forward distance calculated in real time based on the car's speed If equation (3) is satisfied, the vehicle stops moving forward. d is the preset threshold value.

[0053]

[0054] Formula (3) means that the difference Δd′ between the average value of the distance from the target signal generating device obtained by binocular ranging measurement and the distance from the target signal generating device calculated using the vehicle speed and d0 is less than the threshold value T d At this time, it can be considered that the car has reached the distance signal generating device d0.

[0055] Step 5: If Δd′ is increasing for n consecutive times, then stop moving forward immediately and move back at half the current speed; if Δd′ is increasing for n consecutive times during the backward movement, then stop moving back immediately and move forward at half the current speed; and so on, until equation (3) is satisfied. d and the current speed v, that is, at 2T dwithin an acceptable error range, n consecutive received signal generator signal upper limit value, to ensure n≥2, and then add 1 to the calculation result, as shown in equation (4):

[0056]

[0057] Where Δt is the target signal generating device preset signal transmission interval.

[0058] Step 6, the current car is located at a reference distance d0 from the target signal generating device (error less than T d ), set as point B, the bottom steering controls the camera to face the target signal generating device, and the car head is adjusted to a position perpendicular to the camera direction (i.e. the tangent direction of the circular arc with A as the center and d0 as the radius), preparing for the next step of circular motion.

[0059] Step 7, calculate the car wheel steering angle, with the current position angle as 0°, the car advances at a preset speed along the circular arc with A as the center and d0 as the radius, and uses the signal receiver to collect the received signal strength G λ (d0), records the signal strength and the corresponding collection position (angle) in the memory, and the angle is calculated using the car advancing time (inertial navigation).

[0060] Step 8, due to the difficulty of ensuring complete constancy of the signal generator transmission interval and the car advancing speed in the actual environment, angle correction is adopted during the motion, and the car needs to rotate along the circular arc multiple times until each angle (λ'±δ)° has at least one record, where δ is the acceptable angle error. When there is more than one record at the same angle, take the average value, so as to realize the measurement of the signal strength G λ′ (d0) at each angle λ' at the reference distance.

[0061] Step 9, while moving in a circle at the reference distance d0, radial correction is adopted, and the steering is adjusted every degree to ensure that the target signal generating device is in the center of the camera screen. If the intelligent car is at a reasonable distance from the target signal generating device at this time, measure the signal strength G λ (d0), and adjust the steering angle according to the angle between the camera and the vehicle body to move to the next degree (if the angle between the camera and the vehicle direction is greater than 90°, increase the vehicle steering angle; otherwise, decrease the steering angle). If the intelligent car is too far or too close to the target signal generating device at this time, do not measure the signal strength G λ (d0), just adjust the steering angle to move to the next degree and continue the correction process of this step.

[0062] Step 10: To reduce the angle recording error caused by inertial navigation error, the binocular camera collects and records the image at startup (i.e., at 0°), and corrects the angle information through real-time image comparison. That is, correction is performed every time the 0° position is passed. The specific method is: compare the pixel parameters of the recorded image with the collected image, and the similarity is greater than T P That is, it is considered to be back to 0°. To ensure the validity of the comparison result, the more unique direction of the picture should be determined as the 0° direction.

[0063] Step 11: The smart car moves from d0 to the first measurement distance d, selects a reference direction as 0°, establishes the corresponding relationship between the new angle and the measured angle λ′ at d0, and moves forward to the distance d along the reference direction, that is, Figure 3 At point C in the image, measure the signal strength G0(d) there, which is used to calculate the theoretical distance d' at the same signal strength in each direction. λ , theoretical distance d′ λ is the distance from the point with signal strength G0(d) in the λ direction to the target signal generator; (Special note: In communication theory, the signal strength of an omnidirectional antenna should be consistent at the same distance in all directions, that is, the signal strength at distance d in all directions should be G0(d). However, due to the influence of production process, voltage, propagation medium, etc., it is actually anisotropic. At present, it is not possible to achieve completely consistent signal strength at the same distance in all directions. Therefore, it is necessary to obtain the irregularity parameters in each direction through measurement to obtain the actual coverage range of the target signal generator)

[0064] Step 12: Replace G in formula (1) with G0(d) λ (d) (i.e., let λ = 0), let k0 = 1 in the reference direction, combined with the known G λ (d0) can be used to calculate the value of the path loss parameter η in the current environment. The loss parameter η can be regarded as a constant in the same environment and can be directly used in subsequent calculations.

[0065] Step 13: The car keeps the radius d unchanged and moves around the circle to point D in the 1° direction (as shown in the figure). Figure 3 As shown), measure the signal strength G1(d) here, combined with the known G λ (d0) and η, the irregularity coefficient k1 in the 1° direction can be calculated.

[0066] Step 14, replace G in formula (1) with G0(d) λ(d), setting λ = 1 for the remaining parameters allows us to calculate the theoretical distance d′1, which is the distance from the point with signal strength G0(d) in the 1° direction to the target signal generator. (Note: The reason we don't perform a rapid measurement here, as we did at d0, is that d is typically greater than, or even significantly greater than, d0 from the signal generator. The aforementioned method, which only calibrates once per revolution, can result in significant cumulative errors. Therefore, we use a slow, on-the-fly calibration approach.)

[0067] Step 15: The car moves in a circle with the actual distance d as the radius, repeating steps 13 and 14 at each degree to measure and calculate the irregularity coefficient k in each direction. λ and theoretical distance d′ λ , that is, the irregular coefficient and theoretical distance corresponding to another set of measured distances, and this cycle is repeated until all m sets of data, including the data measured at point d, are measured.

[0068] Step 16: Radial positioning correction is also used in all m groups of signal strength measurements. After adjusting the servo to ensure that the target signal generator is in the center of the camera image, the signal strength G at that point is measured regardless of whether the distance between the smart car and the target signal generator is reasonable. λ (d) The measurement, combined with the actual binocular distance measurement results at this time Derived irregularity coefficient k λ (Theoretically, k λ It will not change much with the distance), and finally adjust the steering angle according to the angle between the camera and the car body to move to the next degree to continue the correction process in this step.

[0069] Step 17: The theoretical distance d′ corresponding to each of the m groups λ And the position information measured by binocular ranging is sent to the application processing end;

[0070] Step 18: The application processing end calculates the theoretical distance d' corresponding to the m groups. λ Draw a contour map of each group of signal strength G0(d), and finally transmit the graph and parameters to the user end.

[0071] Step 19: According to the irregular coefficient k corresponding to the m group λ Weibull parameter fitting was performed to obtain the scale and shape parameters of each group for irregularity analysis.

[0072] Specifically, visual ranging is combined with deep learning and smart car control systems. The lenses on the smart cars use deep learning technology to achieve image semantic segmentation. The neural network model is optimized through typical device appearance training and interactive recognition is achieved, ensuring the accuracy of device recognition and the direction of travel of the smart car, thereby improving the accuracy of the detected irregularity coefficient.

[0073] In actual operation, attention should be paid to the coordination of vehicle speed and signal receiver measurement frequency. If the speed of the trolley in one degree is faster than the measurement frequency and the time used for data transmission, it will cause the loss of data in some directions. To avoid this problem, the trolley can be made to move 1 second per 1 degree, so that the speed of the trolley matches the signal receiver measurement frequency; or the trolley can be driven by the program to travel one degree with appropriate pause to ensure the time of data measurement and transmission, so as to avoid data loss.

[0074] In addition, the accuracy of the trolley's movement during signal strength measurement directly affects the accuracy of the measurement results. Since the environment of the intelligent vehicle is not constant, considering various internal and external influences such as uneven ground, the intelligent vehicle may deviate during travel. Radial and angular correction can be combined to adjust the correction in two dimensions.

[0075] Because inertial navigation can cause angle recording errors, the binocular camera collects and records the image at 0°, and corrects at every 0° position. By comparing the pixel parameters of the recorded image and the collected image, if the similarity is greater than T P It is considered to return to 0°.

[0076] During the movement, a large amount of received signal strength data needs to be collected to quickly detect the irregularity coefficient of radio propagation. In this process, it is necessary to prevent as much as possible the interference from the inside and outside.

[0077] Different types of customers have different accuracy requirements for graphical location information, so the accuracy requirements for measuring signal strength are naturally different. To meet various needs, the trolley needs to have different movement plans for different accuracy requirements. The measurement accuracy is related to the speed and the number of laps of the intelligent vehicle moving around the target signal generating device. The faster the speed, the less likely it is to be disturbed by sudden noise, and the more laps, the higher the accuracy of the measured signal strength.

Claims

1. A method for detecting a radio propagation irregularity coefficient, wherein: The radio propagation irregularity coefficient detection system includes an on-board processing end, an application processing end, and a user end. The user end can customize distance, speed, and accuracy, and then transmit the signal to the interface layer of the on-board processing end; the on-board processing end completes data measurement according to requirements and transmits it to the application processing end via the sending layer; the application processing end processes the data, then draws it into a signal strength contour map and optimizes it, and finally outputs the data and graphics to the user end, and finally outputs the graphics; the on-board processing end includes a signal measurement system, a motion system, a visual ranging system, a correction system, and a data processing system, and is characterized by comprising the following steps: S1, the user terminal sets the vehicle speed, m measurement distances, reference distance d0, and ranging error threshold T d , image similarity threshold T P , acceptable angle error δ, correction interval t α Send to the vehicle processing terminal; S2, the signal measurement system calculates a matching measurement frequency based on the distance and vehicle speed; S3, the visual ranging system controls the movement and stop of the vehicle according to the set reference distance d0 and the actual distance d in combination with the motion system; the correction system uses the visual ranging system to detect whether it is yawed, and adjusts the steering angle in combination with the motion system to achieve correction; the distance from the target signal generating device is measured in real time using binocular ranging The initial distance d' between the car and the target signal generator, and the forward distance calculated in real time based on the car's speed The correction system includes a radial correction system and an angular correction system; S4, the data processing system calculates m groups of irregularity coefficients k according to the reference distance d0, the actual distance d and the measured signal strength λ and theoretical distance d λ ; S5, the m groups of theoretical distance d λ And the position information measured by binocular ranging is sent to the application processing end; S6, theoretical distance d from the application processing end λ Draw a contour map of each group of signal strength G0(d), and finally transmit the graph and parameters to the user end; S7, according to the irregularity coefficient k corresponding to the m group λ Weibull parameter fitting was performed to obtain the scale parameters and shape parameters of each group.

2. The method for detecting radio propagation irregularity coefficient according to claim 1, wherein: In step S2, during the movement, the radial correction system ensures that the target signal generating device is at the horizontal center of the camera image. If the angle between the camera and the vehicle body is less than the threshold value T d , then measure the signal strength and adjust the steering angle of the wheel according to the angle of the camera; if the angle between the camera and the vehicle body is greater than the threshold value T d At this time, there is no need to measure the signal strength, only the visual ranging result is needed. Adjust the steering angle according to the relationship with the set distance; repeat the above operation every time the vehicle moves forward one degree to ensure that the vehicle runs around the target signal generating device at a fixed distance.

3. The method for detecting radio propagation irregularity coefficient according to claim 1, wherein: In step S2, during the movement, the angle correction system will collect the corresponding angle records, and the car will move in a circle multiple times until there is at least one record for each angle λ′; when there are more than one record results for the same angle, the average value is taken to complete the signal strength G of each angle λ′ at the reference distance. λ′ (d0) measurement.

4. The method for detecting radio propagation irregularity coefficient according to claim 1, wherein: In step S3, the irregularity coefficient k is calculated λ and theoretical distance d λ The implementation steps are as follows: S31, the trolley moves around the circle with the radius d unchanged to the 1° direction, measures the signal strength G1(d) there, and calculates the irregularity coefficient k1 in the 1° direction by combining the known G1(d0) and loss parameter η; S32, let λ = 0 and substitute k1 into the following formula to calculate the theoretical distance d′1. The theoretical distance d′1 is the distance from the point with the signal strength G0(d) in the 1° direction to the target signal generator. This distance is used to draw lines connecting points with the same signal strength in each direction, indicating the actual coverage area of ​​the target signal generator at that signal strength. Among them, G λ (d) is the received signal strength at distance d in the direction of λ, d0 is the reference distance, η is the path loss, F σ,λ is zero-mean Gaussian white noise with standard deviation σ in the λ direction; k λ is the irregularity coefficient in the λ direction; S33, the trolley moves in a circle with the actual distance d as the radius. Every time it moves forward one degree, it repeats steps S31 and S32 to measure and calculate the irregularity coefficient k in each direction. λ and theoretical distance d′ λ , and repeat this cycle until all m groups of data, including the data measured at point d, are measured.

5. The method for detecting radio propagation irregularity coefficient according to claim 4, characterized in that: Radial positioning correction was performed in all m groups of signal strength measurements.

Images