LoRa positioning method based on RSSI and TDoA fusion

Abstract

The invention relates to the technical field of low-power wide area networks, and discloses a LoRa positioning method based on RSSI and TDoA fusion, which comprises the following steps: S1, in a data receiving module, adopting a Gaussian mean filtering technology which combines traditional mean and median filtering and Gaussian filtering; and S2, accurately estimating the outdoor position. The step S1 specifically comprises the following steps: S11: removing abnormal values with obvious excessive fluctuation in the received data; s12, removing the value of the low probability interval through Gaussian filtering, calculating the mean value navg of the remaining n pieces of data after Gaussian filtering, setting every 30 pieces of data in the n pieces of data as a group ni, and calculating the mean value ni avg and the median ni mid of each group; the data receiving module can achieve the following effects: fast weighting processing of received data is realized, the problem of low positioning precision caused by excessive data fluctuation is reduced by adopting threshold optimization, complicated operation is not involved, and the method can be directly applied to a wireless network node with relatively weak computing power.

Description

Technical Field

[0001] The present invention relates to the technical field of low power wide area network, and in particular to a LoRa positioning method based on RSSI and TDoA fusion. Background Art

[0002] In recent years, the emerging low-power wide area network technology has gradually matured. LoRa is a very mature, representative and rapidly developing low-power wide area network technology, adopted and promoted by the Italian French Semiconductor Group Co., Ltd. The LoRa physical layer is based on linear modulation spread spectrum technology. Because it uses variable frequency encoding information that changes linearly with time, it can effectively resist Doppler frequency shift, and because its entire bandwidth is allocated to the broadcast signal (information or data), the modulation technology used in LoRa makes it robust to channel noise. In addition, since the transmission is spread in a pseudo-random manner, the distribution characteristics of the pseudo-random manner show noise distribution characteristics, so the use of this modulation technology greatly improves the security of the LoRa communication process. The channel supports multiple bandwidths and has forward error correction capabilities. The data encryption method uses AES128 encryption. It has a sensitivity of -130dbm and a long-range communication capability with high receiving sensitivity.

[0003] Positioning technologies are mainly divided into two categories: distance-based positioning technology and distance-free positioning technology. Distance-based technology uses distance or angle information to estimate the location of nodes such as Global Positioning System (GPS), Angle of Arrival (AOA), Time of Flight (TOF), Time Difference of Arrival (TDOA), Received Signal Strength Index (RSSI), etc. Distance-free technology only uses the connectivity information between nodes, such as DV-Hop, Approximate Point-In-Triangulation test (APIT) and centroid system.

[0004] There are many unresolved issues in commonly used positioning technologies. For example, some positioning solutions require high-precision sensors to obtain high-precision positioning, which brings additional hardware overhead. Other solutions are sensitive to environmental interference. If the environment is slightly complex, the performance of the positioning algorithm will be greatly reduced. They cannot meet the positioning application requirements in a wide range of low-power scenarios. Therefore, there is an urgent need to develop a low-power positioning technology that is suitable for a wide range of complex environments.

[0005] Deficiencies of existing technologies: Current mainstream outdoor positioning technologies are limited by power consumption or communication distance, making it difficult to obtain high-precision positioning services in outdoor complex environments at a low cost. Summary of the invention

[0006] The object of the present invention is to provide a LoRa positioning method based on RSSI and TDoA fusion, which solves the problems in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: a LoRa positioning method based on RSSI and TDoA fusion, comprising the following steps:

[0008] S1: In the data receiving module, Gaussian mean filtering technology is used, which combines traditional mean filtering, median filtering and Gaussian filtering;

[0009] S2: Accurate estimation of outdoor location.

[0010] Preferably, step S1 specifically comprises:

[0011] S11: removing some outliers with obviously large fluctuations in the received data;

[0012] S12: remove the values ​​in the low probability interval by Gaussian filtering, calculate the mean navg of the remaining n data after Gaussian filtering, set every 30 of the n data as a group ni, and calculate the average niavg and median nimid of each group;

[0013] S13: Difference between ni avg / ni mid and navg. The absolute value with the smallest difference occupies a higher weight value, and the others occupy a lower weight value.

[0014] S14: Finally, the corresponding weight value is multiplied and the sum is taken as the result value of the group. The data is dynamically accumulated with 30 data packets as a group and 15 data packets as a step.

[0015] Preferably, step S2 specifically comprises:

[0016] S21: After obtaining the RSSI positioning result, the TDoA algorithm is also required to participate in positioning;

[0017] S22: using the weighted centroid of the triangle to estimate the position of the area formed by the three-sided positioning with a certain accuracy;

[0018] S23: Obtaining a hyperbolic equation of the time difference of arrival using the collected ToA data;

[0019] S24 finally uses the RSSI positioning estimate as the initial value, and uses the Taylor series expansion method to narrow the error range to obtain the final positioning result.

[0020] The present invention provides a LoRa positioning method based on RSSI and TDoA fusion. The LoRa positioning method based on RSSI and TDoA fusion has the following beneficial effects:

[0021] 1. In the data receiving module, the present invention can achieve the following effects: realizing fast weighted processing of received data, and using threshold optimization to reduce the problem of low positioning accuracy caused by excessive data fluctuations, and does not involve complex operations, and can also be directly applied to wireless network nodes with weak computing capabilities.

[0022] 2. The present invention can achieve the following effects in terms of outdoor location estimation: using RSSI and TDoA to fuse outdoor location algorithms can not only achieve low power consumption and fast positioning services for embedded devices, but also improve positioning accuracy in outdoor environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A random node distribution measurement diagram of the present invention;

[0024] Figure 2 This is a schematic diagram of average error at different positioning measurement times according to the present invention;

[0025] Figure 3 Schematic diagram of average error intervals of different anchor nodes of the present invention;

[0026] Figure 4 This is a schematic diagram of the anchor node location distribution in a 100m×100m NLOS environment of the present invention;

[0027] Figure 5 Schematic diagram of the distribution of RSSI values ​​of the present invention;

[0028] Figure 6 Schematic diagram of performance comparison of three filtering algorithms of the present invention;

[0029] Figure 7 This is a schematic diagram of the positioning result of the algorithm designed by the present invention;

[0030] Figure 8 This is a schematic diagram for comparing the cumulative distribution of positioning errors of the present invention;

[0031] Fig. 9 This is a schematic diagram of the physical image of the Lora wireless communication SX1280 module of the present invention. DETAILED DESCRIPTION

[0032] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described with reference to the accompanying drawings.

[0033] A preferred embodiment of the LoRa positioning method based on RSSI and TDoA fusion provided by the present invention is as follows: Figures 1 to 9 As shown:

[0034] The LoRa positioning method based on RSSI and TDoA fusion is characterized by comprising the following steps:

[0035] S1: In the data receiving module, Gaussian mean filtering technology is used, which combines traditional mean filtering, median filtering and Gaussian filtering;

[0036] The specific steps are:

[0037] S11: removing some outliers with obviously large fluctuations in the received data;

[0038] S12: remove the values ​​in the low probability interval by Gaussian filtering, calculate the mean navg of the remaining n data after Gaussian filtering, set every 30 of the n data as a group ni, and calculate the average niavg and median nimid of each group;

[0039] S13: Difference between ni avg / ni mid and navg. The absolute value with the smallest difference occupies a higher weight value, and the others occupy a lower weight value.

[0040] S14: Finally, the corresponding weight value is multiplied and the sum is taken as the result value of the group. The data is dynamically accumulated with 30 data packets as a group and 15 data packets as a step.

[0041] S2: Accurate estimation of outdoor location;

[0042] The specific steps are as follows: S21: after obtaining the RSSI positioning result, the TDoA algorithm is also required to participate in the positioning;

[0043] S22: using the weighted centroid of the triangle to estimate the position of the area formed by the three-sided positioning with a certain accuracy;

[0044] S23: Obtaining a hyperbolic equation of the time difference of arrival using the collected ToA data;

[0045] S24 finally uses the RSSI positioning estimate as the initial value, and uses the Taylor series expansion method to narrow the error range to obtain the final positioning result.

[0046] in, Figures 1 to 3 To simulate NLOS environment data to verify the technical effect, Figures 4 to 8 Table 1 and Table 1 show the technical effects verified by actual complex NLOS environment data.

[0047] Figure 1 For the positioning simulation NLOS environment of this design (assuming an obstacle-free 100m×100m outdoor area), the parameters A=-28, η=3 are selected, and the initial simulation environment parameters are 30 anchor nodes and 30 groups of calculations.

[0048] Figure 2The simulation results show the influence of increasing the step size of the node filter window on the average positioning error. In the 100m×100m positioning area, the total number of anchor nodes deployed is 30, and the step size of the filter window of the collected positioning data gradually increases.

[0049] Depend on Figure 3 It can be seen that as the step size of the filter window increases, the average positioning error first decreases and then tends to stabilize. Among them, the accuracy of the three-sided ranging algorithm based on RSSI is the lowest, because this algorithm does not consider the impact of the environment and directly uses the arrival signal strength to estimate the distance between the anchor node and the node to be measured, thereby obtaining the position. This design uses the random initial value TDoA Taylor expansion algorithm and the fusion algorithm to locate the node to be measured according to the arrival time difference, and the positioning accuracy is better than the traditional three-sided ranging algorithm based on RSSI. In the 100m×100m positioning area, all nodes are randomly deployed, and the number of anchor nodes is gradually increased under the same other conditions. From the experimental results, it can be seen that with the increase in the number of deployed anchor nodes, the positioning accuracy of the three different algorithms is improving. This is because as the number of anchor nodes increases, the closer the anchor nodes are selected in the anchor node selection strategy, the greater the probability of forming an approximately equilateral triangle. In this way, the signal is less attenuated by the environment, the available information is richer, and the calculated positioning coordinates are more accurate.

[0050] Figure 4 The actual 100m x 100m NLOS environment is shown. According to the on-site survey and measurement, the anchor position is first set in the southern area of ​​the campus. Four anchor nodes are arranged 1m above the ground. Including anchor points 1 and 2 on the campus playground, anchor point 3 near the central path, anchor point 4 in the garden, and another node to be randomly tested. The standard track and field is 105 meters long and 68 meters wide. Due to the presence of fences, railings, mobile signal base stations, multipath signals and coherent signals in the environment, it can be approximately regarded as a nonlinear line of sight environment, which is conducive to verifying the effectiveness of the filtering algorithm. The device to be positioned is placed at a height of 1m, 10m, 15m, 20m, 30m, 35m, 40m, 45m, and 50m away from the anchor node. Each node to be tested sends positioning data at the same time, and 30 groups of data received by the anchor node are selected to observe the arrival signal strength at each location.

[0051] The distribution of RSSI is as follows Figure 5 As shown, Figure 5The top line is the maximum RSSI value, the downward line is the median of the RSSI value, is the statistical RSSI value after Gaussian mean filtering, is the mean of RSSI, and is the minimum value of RSSI. The analysis results show that in the first 10m, there is almost no difference between the mean, median and RSSI statistical values ​​after Gaussian mean filter filtering. In the first 10m, the range of RSSI values ​​is about 10dBm. In the range of 15 to 30m, the range of RSSI values ​​is about 20 to 25dBm. In the range of 30 to 40m, the range of RSSI values ​​received exceeds 35dBm. As the distance gradually increases, the range of RSSI values ​​received at the same distance point becomes larger and larger, and the minimum RSSI value fluctuates more and more seriously as the distance gradually increases. Therefore, as the distance increases, the RSSI value obtained is more affected by reflection and refraction. The RSSI value is filtered by mean filtering and Gaussian mean filtering respectively. By comparison, it is found that the attenuation of the signal strength value obtained by the Gaussian mean filter designed in this paper is smoother, and it also contains the effective information of the mean and median, which truly reflects the relationship between the received signal strength and the increase in the distance between the two.

[0052] Figure 6 The filtering effects of three different filtering algorithms, mean filtering, median filtering and Gaussian mean filtering, are shown for 100 sets of RSSI data. At the ideal signal value of -49.7dbm at 20m, the filtering values ​​of mean filtering and Gaussian mean filtering algorithms fluctuate around the ideal value. When the number of collection groups is 50-60, the floating range of the Gaussian mean filtering algorithm is about 0.5-1.5dbm, while the floating range of the mean filtering algorithm is about 2dbm. Before collecting 10 sets of data, the filtering range of the Gaussian mean filtering algorithm is about 3dbm, while the filtering range of the mean filtering algorithm is close to 6dbm. Compared with the traditional mean filtering algorithm, the Gaussian mean filtering algorithm has a smaller degree of fluctuation.

[0053] Figure 8 The error cumulative distribution comparison diagram of the three positioning algorithms in the actual NLOS environment is shown.

[0054] Table 1 shows the positioning error comparison of three different positioning methods under the same data correction method. The positioning results show that the three-sided positioning method based on RSSI has the largest error, with an average error of 7.15m in the presence of obstacles; the fusion algorithm designed in this paper has the best accuracy, with an average error of only 3.29m. The positioning accuracy is 53.99% higher than that of the RSSI three-sided positioning algorithm and 8.10% higher than that of the Taylor expansion algorithm with random initial values.

[0055] Table 1

[0056] Comparison of positioning errors

[0057]

[0058] The above description is only an illustrative specific implementation mode of the present invention and is not intended to limit the scope of the present invention. Any equivalent changes and modifications made by any technician in the field without departing from the concept and principle of the present invention should fall within the scope of protection of the present invention. It should also be noted that the various components of the present invention are not limited to the above-mentioned overall application. The various technical features described in the specification of the present invention can be selected one by one or multiple ones can be selected and used in combination according to actual needs. Therefore, the present invention should naturally cover other combinations and specific applications related to the invention point of this case.

Claims

1. The LoRa positioning method based on RSSI and TDoA fusion is characterized by: The following steps are involved: S1: In the data receiving module, Gaussian mean filtering technology is used, which combines traditional mean filtering, median filtering and Gaussian filtering; S2: Accurate estimation of outdoor location.

2. The LoRa positioning method based on RSSI and TDoA fusion according to claim 1, characterized in that: The specific steps of S1 are: S11: removing some outliers with obviously large fluctuations in the received data; S12: Remove the values ​​in the low probability interval by Gaussian filtering, calculate the mean navg of the remaining n data after Gaussian filtering, set every 30 of the n data as a group ni, and calculate the mean niavg and median ni mid of each group; S13: Difference between ni avg / ni mid and navg. The absolute value with the smallest difference occupies a higher weight value, and the others occupy a lower weight value. S14: Finally, the corresponding weight value is multiplied and the sum is taken as the result value of the group. The data is dynamically accumulated with 30 data packets as a group and 15 data packets as a step.

3. The LoRa positioning method based on RSSI and TDoA fusion according to claim 1, characterized in that: The specific steps of S2 are: S21: After obtaining the RSSI positioning result, the TDoA algorithm is also required to participate in positioning; S22: using the weighted centroid of the triangle to estimate the position of the area formed by the three-sided positioning with a certain accuracy; S23: Obtaining a hyperbolic equation of the time difference of arrival using the collected ToA data; S24 finally uses the RSSI positioning estimate as the initial value, and uses the Taylor series expansion method to narrow the error range to obtain the final positioning result.

Images