A wireless sensor network node localization method using fuzzy decision making

Abstract

The invention discloses a wireless sensor network node positioning method using fuzzy decision-making, comprising the following steps: quantizing at least one factor affecting channel conditions to obtain a quantization function; determining the number m of anchor nodes that an unknown node can connect to; and calculating the score of each anchor node in combination with the quantization function; calculating a parameter A according to RSSI values ​​received by two anchor nodes at different distances; selecting S anchor nodes with the highest scores among the m anchor nodes, and calculating a first distance between the S anchor nodes, and calculating S first channel attenuation parameters n1 according to the first distance and the parameter A; normalizing the scores of the S anchor nodes, and calculating a second channel attenuation parameter n2 in combination with the S first channel attenuation parameters n1; using the RSSI value of the anchor node received at the node to be positioned, and according to the parameter A and the second channel attenuation parameter n2, calculating a second distance between the node to be positioned and the anchor node, and then determining the coordinates of the node to be positioned by using a traditional three-sided positioning method to complete the positioning.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless sensor networks, and more specifically to a wireless sensor network node positioning method using fuzzy decision making. Background Art

[0002] Wireless sensor networks (WSNs) are known as one of the 21 most influential technologies in this century. They belong to the cross-disciplinary frontier field of computer, communication and mathematics. It is a self-organizing data processing network formed by multiple sensor nodes with certain functions arranged in the monitoring area through data collection, data transmission and other methods. Through the dense deployment of sensor nodes in the network, data collection quantification, data aggregation and data transmission operations can be realized. As a modern intelligent network, wireless sensor networks have been widely used in military, Internet of Things, environmental monitoring, modern transportation (such as shared bicycles in the past two years), convenient express delivery, modern medical care, industry, space exploration and other fields, and have broad development prospects. In the application of a large number of wireless sensor networks, if the collected data does not have corresponding location information, the data will lose its value. Therefore, positioning technology is one of the hot spots in the research field of wireless sensor networks.

[0003] In wireless sensor networks, positioning algorithms can be divided into two stages: top-level distance estimation and bottom-level position estimation. The top-level distance estimation method is the main difference between ranging and non-ranging positioning algorithms. Depending on whether it is necessary to directly measure the distance between transceivers through hardware means, the positioning algorithms of wireless sensor networks can be divided into two types: range-free and range-based. The non-ranging positioning algorithm uses the connectivity of the network to estimate the distance between nodes. According to the different ways of estimating the distance, it can be mainly divided into positioning algorithms based on geometric constraints, hop-progress based, and machine learning based. Most of the ranging-based positioning algorithms require the installation of specific hardware devices between the transceivers, and use these hardware to estimate the distance between the transceivers through the physical characteristics of the received signals.

[0004] The non-range positioning technology estimates the distance between nodes by the number of data packet hops or geometric constraints between nodes. Its positioning hardware cost and power consumption are low and easy to implement. However, the positioning accuracy is poor and it is suitable for open outdoor environments. Compared with the non-range positioning technology, the distance information between the unknown node and the anchor node in the ranging-based positioning algorithm is obtained by measuring the physical quantities such as signal strength, angle, and signal arrival time between nodes. The distance information is closer to the actual value, so the positioning accuracy is more accurate, but the required hardware cost is higher.

[0005] The following are some existing representative ranging and positioning methods, such as Figure 1 As shown, it is a node positioning method based on ranging;

[0006] In the positioning algorithm based on ranging, AOA (Angle of Arrival), that is, the arrival angle positioning algorithm, its basic process is as follows: calculate the angle α1 between the line between anchor node 1 and the node to be located and the reference direction, and make a ray L1; similarly, calculate the angle α2 between the line between anchor node 2 and the node to be located and the reference direction, and make a ray L2. The intersection of L1 and L2 is the position of the node to be located. AOA positioning determines the position by the intersection of two straight lines, and there cannot be multiple intersections, thus avoiding the ambiguity of positioning. However, in order to measure the incident angle of electromagnetic waves, the receiver must be equipped with a highly directional antenna array, and the equipment is highly complex.

[0007] TOA (Time Of Arrive), that is, the arrival time positioning algorithm, its basic process is as follows: the anchor node sends a data packet with a sending timestamp T1 to the node to be located. After receiving this data packet, the node to be located returns an ACK data packet. The anchor node receives the ACK data packet of the node to be located and records the current time T2. (where c is the speed of light) to find the distance between nodes and then determine the coordinates of unknown nodes. In order to locate once, the TOA algorithm requires two communications between each anchor node and the node to be located. The advantage of this positioning method is its convenience. It only requires several anchor nodes placed in different positions and one node to be located. Its disadvantages are also obvious: first, the positioning speed is low; second, since the quality of each communication cannot be guaranteed, and a pair of anchor nodes and nodes to be located cannot perform self-calibration, the accuracy will be affected.

[0008] TDOA (Time Different Of Arrive), that is, the ranging method of the signal arrival time difference algorithm is to calculate the distance difference between the node to be located and the two anchor nodes through the time difference between two different anchor nodes receiving the data packets sent by the node to be located. Using four anchor nodes, three sets of such distance differences can be obtained, and then the coordinates of the node to be located can be determined. Compared with the TOA positioning algorithm, the advantage of TDOA is that the number of communications for one positioning is significantly reduced. Secondly, because the distance is measured by time difference rather than absolute time, its accuracy is also higher than TOA. The disadvantage is that the clocks of each anchor node in the TDOA system must be strictly synchronized. Since this positioning is essentially dependent on the speed of light, an inherent clock error of 1ns can cause an inherent distance error of 30cm. It is relatively expensive to build a precise synchronization system with a relatively large node spacing.

[0009] RSSI (receive signal strength indicator), that is, the received signal strength indicator algorithm is based on the principle that the signal power attenuates with the propagation distance when radio waves or sound waves are transmitted in the medium. According to the transmission power of the known signal of the beacon node and the signal power received by the node, the distance between the nodes can be calculated through the attenuation model between the signal and the distance RSSI = A-10n·log d (A can be regarded as the power of the received signal when the signal is transmitted 1m away, n is the environmental attenuation factor reflecting the channel conditions, and d is the distance between the nodes). Compared with the TOA and TDOA ranging methods, the RSSI positioning algorithm has low requirements on the time system and is not affected by factors such as transmission delay and antenna delay. Without additional hardware, the strength of the received wireless signal can be used to infer the distance between the transmitting and receiving nodes. Therefore, the hardware cost, software cost and time cost are relatively low; at the same time, RSSI is part of the Bluetooth protocol stack, the hardware design of the transmitter is simple, the power consumption is low, and the RSSI signal is easy to obtain. The disadvantage of the RSSI positioning algorithm is that the positioning accuracy is greatly affected by environmental factors. Its ranging accuracy is affected by many factors such as obstacles, weather, and non-ranging signals. For example, if there are obstacles (especially metal obstacles) between the sender and the receiver, the energy of the electromagnetic wave will be attenuated. At the same time, multipath fading in the wireless channel will also cause unstable signal energy, resulting in large distance calculation errors, which in turn affects positioning accuracy.

[0010] The physical quantity used by the ranging-based positioning algorithm is RSSI, the attenuation model between the signal and the distance

[0011] RSSI = A-10n·log d (1)

[0012] It is an empirical formula, in which the parameters A and n are empirical parameters that reflect the node transmission power and channel conditions respectively. When the empirical parameters cannot reflect the actual channel conditions well, the distance data error calculated according to the formula will increase, thus affecting the positioning accuracy. The positioning error of the traditional RSSI positioning method mainly comes from the selection of the same empirical parameter n for distance calculation in a large area when calculating the distance between nodes, without considering the impact of the following factors on channel conditions.

[0013] (1) Distance between node and wall

[0014] The closer the distance between the node and the wall, the more serious the multipath effect will be on the received signal. The multipath effect refers to the fact that after the electromagnetic wave propagates through different paths, each component field arrives at the receiving end at different times, and the superposition of each other according to their respective phases causes interference, which distorts the original signal or generates errors. For example, if the electromagnetic wave propagates along two different paths, and the lengths of the two paths differ by exactly half a wavelength, then the two signals will cancel each other out when they arrive at the end (the peaks and troughs coincide). Figure 2 For example, propagation path 1 refers to the signal transmitted by node 1 directly reaching node 2 without reflection; propagation path 2 refers to the electromagnetic wave transmitted by node 1 reaching node 2 after encountering the wall and reflecting during propagation. During the propagation of wireless signals, the signal strength will decrease with the increase of propagation distance. When node 2 is close to the wall, the distance between propagation path 1 and propagation path 2 is not much different, so the strength of the two signals received by node 2 is almost the same. If the lengths of the two paths are exactly different by (2k+1)λ, k=0,1,2,… (λ is the wavelength of the electromagnetic wave signal), then the two signals are exactly offset and the signal received by node 2 is weakened; if the lengths of the two paths are exactly different by (2k+2)λ, k=0,1,2,…, then the two signals are exactly superimposed and the signal received by node 2 is enhanced. When node 2 is far away from the wall, the length of path 2 is larger than that of path 1. At this time, the signal strength received by node 2 is mainly determined by the component of path 1. At this time, the multipath effect has no obvious effect on the signal strength received by the node. The traditional RSSI positioning algorithm does not consider the impact of multipath effect when the anchor node is close to the wall. The selected empirical parameter n value cannot well reflect the channel conditions when the anchor node is close to the wall. Therefore, it will cause large errors in distance calculation and affect the positioning accuracy.

[0015] (2) Obstacles

[0016] When electromagnetic waves encounter obstacles (especially metal obstacles) during propagation, they need to use the external electric and magnetic fields to induce the electric and magnetic fields inside the medium. The energy will be attenuated. Figure 3 For example, when other conditions are the same, the actual channel conditions are well fitted, resulting in errors in distance calculation and affecting positioning accuracy.

[0017] (3) Distance between nodes

[0018] from Figure 4 It can be seen that the characteristic curve of the actual measured RSSI value is not smooth. Overall, as the distance between nodes increases, the RSSI value gradually decays, but when the distance between two nodes is far, the RSSI value fluctuates greatly. The empirical formula cannot fit the relationship between the RSSI value and the distance between nodes well. The traditional RSSI positioning method does not consider the impact of the distance between nodes on the ranging accuracy. When the distance between two nodes is large, the distance calculated by the RSSI value will have a large deviation from the actual value, thus affecting the positioning accuracy.

[0019] (4) Node density

[0020] The frequencies of the signals transmitted by nodes of the same model are all in the same frequency band. When the signal frequencies of two nodes are adjacent or multiple frequencies, mutual interference may occur. The denser the nodes, the more significant the interference. When the interference is serious, the nodes cannot communicate normally or the communication distance is very short. The traditional RSSI positioning algorithm does not consider the impact of node density on ranging accuracy. When the density of nodes near a node is high, the distance calculated by the RSSI value will have a large deviation from the actual value, thus affecting the positioning accuracy.

[0021] In summary, multiple factors in the actual environment will affect the channel conditions. The traditional RSSI positioning algorithm does not take into account that the channel conditions near the locations of different nodes are different. The same n value is selected in the empirical formula to fit the RSSI-d relationship, resulting in a large deviation between the distance value calculated based on the RSSI value of some anchor nodes at the node to be located and the true value, which affects the positioning accuracy. Summary of the invention

[0022] In order to solve the problems of the above-mentioned deficiencies and defects of the traditional RSSI positioning algorithm, the present invention provides a wireless sensor network node positioning method using fuzzy decision-making, which can effectively improve the positioning accuracy.

[0023] In order to achieve the above-mentioned purpose of the present invention, the technical scheme adopted is as follows:

[0024] A wireless sensor network node positioning method using fuzzy decision making, the method comprising the following steps:

[0025] quantizing at least one factor affecting the channel condition to obtain a corresponding quantization function;

[0026] Determine the number m of anchor nodes that an unknown node can connect to; and calculate the score of each anchor node based on the quantization function of the factors affecting the channel condition of each anchor node;

[0027] The parameter A is calculated based on the RSSI values ​​received by two anchor nodes at different distances, where parameter A represents the node transmission power;

[0028] Selecting S anchor nodes with the highest scores from the m anchor nodes, and calculating the first distances between the S anchor nodes, substituting the first distances and parameter A into the attenuation model between the signal and the distance, to obtain S first channel attenuation parameters n1, where 0<S≤m;

[0029] Normalizing the scores of the S anchor nodes, and combining the S first channel attenuation parameters n1 to calculate the second channel attenuation parameter n2;

[0030] Using the RSSI value of the anchor node received at the node to be located, and substituting the parameter A and the second channel attenuation parameter n2 into the attenuation model between the signal and the distance, the second distance between the node to be located and the anchor node is calculated;

[0031] According to the second distance, the coordinates of the node to be located are determined using the traditional three-sided positioning method to complete the positioning.

[0032] Preferably, the factors affecting the channel conditions include one or more of the distance between the node and the wall, obstacles near the node, the distance between the nodes, and the density of the nodes.

[0033] Furthermore, the quantization function of the distance between the node and the wall is as follows:

[0034]

[0035] Where (x, y) represents the coordinates of the anchor node; x m Indicates the maximum horizontal length of the positioning area, y m Indicates the maximum vertical length of the positioning area.

[0036] Furthermore, the quantization function of the obstacles near the node is as follows:

[0037]

[0038] In the formula, k represents the number of neighboring anchor nodes that the anchor node to be evaluated can connect to; n i represents the channel attenuation parameter n between the anchor node to be evaluated and the i-th adjacent anchor node; μ(n i ) represents the average of k values ​​of n.

[0039] Furthermore, the quantization function of the distance between the nodes is as follows:

[0040]

[0041] In the formula, RSSI(d x ) is the RSSI value received by the node to be located from the anchor node; RSSI(d m ) is the RSSI value when the distance between the two nodes is 10m; RSSI(d0) is the RSSI value when the distance between the two nodes is 1m.

[0042] Furthermore, the quantization function of the node density is as follows:

[0043]

[0044] Preferably, a weight matrix is ​​set according to the importance of each factor affecting the channel condition; and a score of each anchor node is calculated in combination with a quantization function of the factors affecting the channel condition, as follows:

[0045] p i =BR i T

[0046] In the formula, B represents the weight matrix, R i represents the channel condition factor matrix affecting the i-th anchor node, 0<i≤m, m represents the number of anchor nodes that the unknown node can connect to; p i represents the score of the i-th anchor node.

[0047] Preferably, select S=3 anchor nodes with the highest scores, denoted as a, b, c, and corresponding scores as p a , p b , p c ;

[0048] The calculation formula of the second channel attenuation parameter n2 is as follows:

[0049]

[0050] In the formula, g a , g b , g c Respectively by p a , p b , p c After normalization, g a +g b +g c =1;

[0051] n ab Indicates the first distance d between anchor node a and anchor node b abGet the first channel attenuation parameter; n ac represents the first distance d between anchor node a and anchor node c ac Get the first channel attenuation parameter; n bc represents the first distance d between anchor node b and anchor node c bc A first channel attenuation parameter is obtained.

[0052] A computer system comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of a wireless sensor network node positioning method utilizing fuzzy decision making are implemented.

[0053] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a wireless sensor network node positioning method using fuzzy decision making.

[0054] The beneficial effects of the present invention are as follows:

[0055] The present invention utilizes the idea of ​​fuzzy decision-making in fuzzy mathematics to quantify at least one factor affecting channel conditions; scores the anchor nodes that a node to be located can connect to, selects the top S anchor nodes in the score, calculates an n value that is relatively more suitable for the actual situation according to the RSSI values ​​of the top S anchor nodes in the score, and then uses the n value in the empirical formula RSSI=A-10n·log d to perform distance calculation to improve the ranging accuracy, and then uses the three-sided positioning method to calculate the position coordinates of the node to be located, so as to realize accurate positioning of the node to be located. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 There are some existing node positioning methods based on ranging.

[0057] Figure 2 This is a schematic diagram of the impact of the distance between the node and the wall.

[0058] Figure 3 This is a schematic diagram of the impact of obstacles.

[0059] Figure 4 This is a schematic diagram of the impact of the distance between nodes.

[0060] Figure 5 The present invention is a flowchart of a wireless sensor network node positioning method using fuzzy decision making.

[0061] Figure 6 It is a schematic diagram of Example 1 for determining obstacles based on neighboring anchor nodes.

[0062] Figure 7 It is a schematic diagram of determining node coordinates using the three-side positioning method in Example 1.

[0063] Figure 8 It is a schematic diagram of simulation node distribution and simulated positioning in Example 1.

[0064] Fig. 9 Schematic diagram of the positioning error of Example 1 as the signal-to-noise ratio changes.

[0065] Fig.10 It is a schematic diagram of the variation of positioning error with communication radius in Example 1.

[0066] Fig.11 It is a schematic diagram of the node arrangement of Example 1.

[0067] Fig.12 It is a schematic diagram of the physical positioning result of Example 1. DETAILED DESCRIPTION

[0068] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0069] Example 1

[0070] The physical quantity used in the wireless sensor network node positioning method using fuzzy decision-making proposed in this embodiment is RSSI, and the attenuation model between the signal and the distance is:

[0071] RSSI = A-10n·log d (1)

[0072] is an empirical formula, where parameter A represents the node transmission power and parameter n represents the empirical parameter of the channel condition. When the empirical parameter cannot reflect the actual channel condition well, the distance data error calculated according to formula (1) will increase, thus affecting the positioning accuracy. The positioning error of the traditional RSSI positioning method mainly comes from the selection of the same empirical parameter n for distance calculation in a large area when calculating the distance between nodes, without considering the influence of factors affecting the channel condition on the channel condition.

[0073] Based on this, this embodiment proposes a wireless sensor network node positioning method using fuzzy decision making, such as Figure 5 As shown, the method comprises the following:

[0074] quantizing at least one factor affecting the channel condition to obtain a corresponding quantization function;

[0075] Determine the number m of anchor nodes that an unknown node can connect to; and calculate the score of each anchor node based on the quantization function of the factors affecting the channel condition of each anchor node;

[0076] The parameter A is calculated based on the RSSI values ​​received by two anchor nodes at different distances, where parameter A represents the node transmission power;

[0077] Selecting S anchor nodes with the highest scores from the m anchor nodes, and calculating the first distances between the S anchor nodes, substituting the first distances and parameter A into the attenuation model between the signal and the distance, to obtain S first channel attenuation parameters n1, where 0<S≤m;

[0078] Normalizing the scores of the S anchor nodes, and combining the S first channel attenuation parameters n1 to calculate the second channel attenuation parameter n2;

[0079] Using the RSSI value of the anchor node received at the node to be located, and substituting the parameter A and the second channel attenuation parameter n2 into the attenuation model between the signal and the distance, the second distance between the node to be located and the anchor node is calculated;

[0080] According to the second distance, the coordinates of the node to be located are determined using the traditional three-sided positioning method to complete the positioning.

[0081] In a specific embodiment, the factors affecting the channel condition include one or more of the distance between the node and the wall, obstacles near the node, the distance between the nodes, and the density of the nodes. Only one or more combinations of the distance between the node and the wall, obstacles near the node, the distance between the nodes, and the density of the nodes can be considered, and any combination belongs to the protection scope of the embodiment. This embodiment is described by combining the distance between the node and the wall, obstacles near the node, the distance between the nodes, and the density of the nodes as factors affecting the channel condition.

[0082] In a specific embodiment, according to the background technology, the closer the node is to the wall, the stronger the influence of the multipath effect on the electromagnetic wave propagation is, and the greater the fluctuation of the received RSSI value is. At this time, the RSSI value cannot fit the empirical formula (1) curve well, and the error between the calculated node distance and the actual value is relatively large. In this embodiment, it is believed that the farther the anchor node is from the wall, the better it is. According to the coordinates of the anchor node, the quantitative function of the distance between the node and the wall is given as follows:

[0083]

[0084] Where (x, y) represents the coordinates of the anchor node; x m Indicates the maximum horizontal length of the positioning area, y m Indicates the maximum vertical length of the positioning area.

[0085] In a specific embodiment, when there are obstacles between nodes, the signal strength received by the node will be lower than that in the absence of obstacles. The RSSI value at this time cannot fit the empirical formula (1) curve well. The distance calculated using the RSSI value of the anchor node with obstacles within the straight-line distance will have a large error compared to the actual value. In this embodiment, it is believed that under the same other conditions, the more "open" (fewer obstacles) the nearby environment is, the "better" the anchor node is. The only information that can be obtained by the positioning method of this embodiment is the coordinates of the anchor node and the RSSI value of the anchor node signal. Therefore, to evaluate the "goodness" of an anchor node in terms of "whether there are obstacles nearby", it is necessary to indirectly judge through the information of its neighboring anchor nodes. Figure 6 As shown in the figure, the position coordinates of all anchor nodes are known. It is assumed that the anchor node to be evaluated can receive signals from the five neighboring anchor nodes. According to the RSSI value of the neighboring anchor node 1 received by the anchor node to be evaluated, the channel attenuation parameter n1 on this path can be obtained in combination with the empirical formula (1). Similarly, n2, n3, n4, and n5 can be obtained. Under the same conditions, if these five values ​​are not much different, it can be considered that the environment near the anchor node to be evaluated has not changed much. If the five values ​​are very different, it is very likely that there are obstacles near the anchor node to be evaluated, resulting in inconsistent channel environments around the anchor node to be evaluated. The channel attenuation parameter n value obtained based on the RSSI value of the neighboring anchor node of the anchor node to be evaluated gives the quantization function of the obstacles near the node as follows:

[0086]

[0087] In the formula, k represents the number of neighboring anchor nodes that the anchor node to be evaluated can connect to; n i represents the channel attenuation parameter n between the anchor node to be evaluated and the i-th adjacent anchor node; μ(n i ) represents the average of k values ​​of n.

[0088] In a specific embodiment, Figure 4 It can be seen that the characteristic curve of the actually measured RSSI value is not smooth. Overall, as the distance between nodes increases, the RSSI value gradually decays, but when the distance between two nodes is far, the RSSI value fluctuates greatly, and the empirical formula (1) cannot fit the relationship between the RSSI value and the distance between nodes well. The distance calculated by the RSSI value will have a large deviation from the actual value. This embodiment believes that under the same other conditions, the closer the anchor node is to the node to be located, the "better". As the distance between nodes increases, the RSSI value gradually decays. Therefore, based on the received anchor node RSSI value, the quantization function of the distance between nodes is given as follows:

[0089]

[0090] In the formula, RSSI(d x ) is the RSSI value received by the node to be located from the anchor node; RSSI(dm) is the RSSI value when the distance between the two nodes is 10m; RSSI(d0) is the RSSI value when the distance between the two nodes is 1m.

[0091] In a specific embodiment, the influence of node density on ranging accuracy has been described in the background technology. Under the premise of other conditions being the same, it is considered that the smaller the number of other anchor nodes near the anchor node, the better. Therefore, according to the number of other anchor nodes that can be connected at the anchor node, the quantitative function of node density is given as follows:

[0092]

[0093] In a specific embodiment, the number of anchor nodes that the unknown node can connect to is determined. In this embodiment, it is assumed that the number of anchor nodes that an unknown node can connect to is m. For the i-th anchor node, four factors affecting the channel condition are calculated, and recorded as follows:

[0094] R i =[f 1i , f 2i , f 3i , f 4i ] (6)

[0095] According to the importance of each factor affecting the channel condition, the weight matrix is ​​set as follows:

[0096] B=[0.2, 0.3, 0.4, 0.1] (7)

[0097] Combined with the quantization function of the factors affecting the channel condition, the score p of the i-th anchor node is calculated. i =BR i T , and similarly calculate the scores of m anchor nodes p1, p2, ..., p m , written as:

[0098] G=[p1,p2,…,p m ] (8)

[0099] In the formula, B represents the weight matrix, R i represents the channel condition factor matrix affecting the i-th anchor node, 0<i≤m, m represents the number of anchor nodes that the unknown node can connect to; p i represents the score of the i-th anchor node.

[0100] In a specific embodiment, in formula (1), parameter A and parameter n are both empirical parameters, which need to be measured in the actual environment and calculated to fit the actual environment. Among them, parameter A reflects the signal transmission power, which is related to the performance of the beacon node itself, so a preliminary experiment is performed to calculate parameter A:

[0101]

[0102] Wherein, RSSI(d1) is the RSSI value received at the distance node d1; RSSI(d2) is the RSSI value received at the distance node d2; d1 and d2 can be any two distances of different sizes less than the node communication radius, for example, d1=1m and d2=2m can be selected.

[0103] In a specific embodiment, the scores of the calculated m anchor nodes are arranged in descending order, and the S=3 anchor nodes with the highest scores are selected, denoted as a, b, and c, and their corresponding scores are p, a , p b , p c Since the coordinates of the anchor nodes are known, the first distance d between the anchor nodes ab, ac, and bc can be calculated. ab , d ac , d bc The S described in this embodiment may also be 4, 5, etc.

[0104] The first distance d will be calculated ab Substituting the parameter A calculated by equation (9) into the attenuation model between the signal and the distance in equation (1) can obtain the first channel attenuation parameter n ab Similarly, the first channel attenuation parameter n can be obtained ac , n bc .

[0105] The score p of the three anchor nodes a , p b , p c After normalization, we can get g a , g b , g c , where g a +g b +g c =1, based on which the second channel attenuation parameter n2 is calculated, and the specific calculation formula is as follows:

[0106]

[0107] Where n ab Indicates the first distance d between anchor node a and anchor node b ab Get the first channel attenuation parameter; n acrepresents the first distance d between anchor node a and anchor node c ac Get the first channel attenuation parameter; n bc represents the first distance d between anchor node b and anchor node c bc A first channel attenuation parameter is obtained.

[0108] In a specific embodiment, after the parameter A and the second channel attenuation parameter n2 are calculated, the RSSI value rss of the anchor node a received by the node to be located can be used. a , combined with formula (1), the distance d between the node to be located and the anchor node a is calculated a , d can be calculated in the same way b and d c .like Figure 7 As shown, the coordinates of the node to be located are determined using the traditional three-sided positioning method, that is, the coordinates of the anchor node a are taken as the center of the circle, d a Draw a circle with radius , and similarly draw circles b and c. The intersection of the three circles is the coordinates of the node to be located. In this way, the node to be located can be accurately located.

[0109] In order to verify the effectiveness of the node positioning method for wireless sensor networks in this embodiment, a simulation experiment was conducted. The computer configuration of the experimental environment: CPU is core (TM) i7-8700K, memory is 16G, operating system is Windows10, and programming is done using Matlab. The specific experimental parameters are set as follows: First, a network containing a total of 100 nodes (including 40 anchor nodes and 60 unknown nodes) is generated in a square area of ​​200m×200m. In the simulation analysis of this embodiment, two variables are considered, namely the communication radius and signal-to-noise ratio of the node. The transmission radius of the node ranges from 20m to 50m, and the signal-to-noise ratio ranges from 1 to 10.

[0110] In order to reflect the impact of the distance between nodes on the RSSI value of the node receiving signal, a distance-related random noise is added to all anchor node signals to simulate the situation where the RSSI value fluctuation increases when the two nodes are far apart:

[0111]

[0112] In the formula, d represents the distance between two nodes, x m Indicates the maximum length of the positioning area.

[0113] In order to reflect the impact of multipath effect on the RSSI value of the node receiving signal, a noise related to the distance from the wall is added to all anchor node signals:

[0114] noise multipath =a d ×rand (12)

[0115] Where d represents the distance between the anchor node and the wall, and a represents a constant coefficient, which is set to 0.8 during simulation.

[0116] In order to reflect the impact of node density on the RSSI value of the node receiving signal, a noise related to the number of connected nodes is added to all anchor node signals:

[0117]

[0118] In order to reflect the impact of obstacles on the RSSI value of the node receiving signal, a random noise is added to the random anchor node signal to simulate the situation where there are obstacles in part of the area to be positioned:

[0119] noise obstacle =a×rand (14)

[0120] In the formula, a is a constant coefficient.

[0121] like Figure 8 It is the distribution of nodes in the simulation experiment. Both anchor nodes and unknown nodes are randomly distributed.

[0122] 1) Comparison of positioning results calculated under the conditions of constant number of anchor nodes, constant node communication radius, and varying signal-to-noise ratio. Fig. 9 ;from Fig. 9 It can be seen that as the signal-to-noise ratio increases, the positioning errors of the three algorithms all show a downward trend. The positioning error of the positioning method described in this embodiment is smaller than that of the traditional algorithm and the TWC positioning algorithm under the condition of a signal-to-noise ratio of 1 to 10.

[0123] 2) Comparison of positioning results calculated under the conditions of constant signal-to-noise ratio, constant number of anchor nodes, and variable node communication radius Fig.10 As shown, from Fig.10It can be seen that under the condition of SNR=8dB, the positioning errors of the three algorithms do not change much with the node communication radius. Because the most basic idea of ​​the three algorithms is to use the RSSI values ​​of three or more anchor nodes to obtain the distance between the node to be located and the anchor node, it is necessary to obtain the RSSI values ​​of at least three anchor nodes to uniquely determine the coordinates of the node to be located. Therefore, when R=20m, the node to be located cannot receive the RSSI information of three or more anchor nodes, so the three methods cannot determine the coordinates of the node to be located. When R=25m, the node to be located receives the RSSI values ​​of three anchor nodes, and the three algorithms can determine the position. When R=25m, the positioning method proposed in this embodiment has a larger error than that after R=30m, because at this time only the RSSI values ​​of three nodes can be received, and one of the nodes has a lower score, but only the information of this node can be used for distance calculation, so the relative positioning error is large. When R continues to increase, the impact on the node positioning accuracy is no longer the main factor, and it is mainly affected by the signal-to-noise ratio. Therefore, the positioning errors of the latter three algorithms are relatively gentle. From Fig.10 It can be seen that under multiple communication radii, the positioning method proposed in this embodiment has advantages in positioning accuracy over the other two algorithms.

[0124] In summary, it can be considered that the simulation experiment proves that the positioning method proposed in this embodiment has advantages in positioning accuracy compared with the prior art.

[0125] In addition to verifying the feasibility of the algorithm and comparing the performance of the positioning method proposed in this embodiment with the existing algorithm in the simulation experiment, the positioning method proposed in this embodiment is also physically tested. A beacon node equipped with an nRF52810 / CC2640R2F Bluetooth chip (Bluetooth version BLE4.2, receiving sensitivity -96dBm) is used to locate the actual node in an actual environment using the positioning method proposed in this embodiment.

[0126] like Fig.11 As shown in the figure, in a laboratory environment of 20m×12m, 7 Bluetooth beacons are arranged relatively evenly as anchor nodes, and mobile phones that can receive the RSSI value of Bluetooth beacons are used as nodes to be located. A simulation coordinate system is established on Matlab. The RSSI value of each node is received by mobile phone software, and the data is processed and calculated using the traditional RSSI positioning method and the positioning method proposed in this embodiment.

[0127] The positioning results of the two methods are as follows Fig.12As shown, it can be seen that the coordinate value of the node to be located calculated by the positioning method proposed in this embodiment is closer to the actual position of the node to be located than the coordinate value obtained by the traditional RSSI positioning method. The positioning error of the traditional RSSI positioning method is 0.7652m, and the positioning error of the positioning method proposed in this embodiment is 0.3538m. Physical inspection can also prove that the positioning method proposed in this embodiment has considerable advantages in positioning accuracy.

[0128] The positioning method proposed in this embodiment selects anchor nodes that participate in positioning, and adopts the RSSI value of the anchor node that better fits the empirical formula during positioning to improve positioning accuracy. The factors affecting the channel conditions are mathematically quantified, which can intuitively reflect the channel environment of the area where the node is located to a certain extent. The calculation method of the empirical parameters in the traditional RSSI positioning algorithm is improved, so that the channel attenuation parameter n can better reflect the channel environment of the area where the node is located.

[0129] Example 2

[0130] A computer system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented:

[0131] quantizing at least one factor affecting the channel condition to obtain a corresponding quantization function;

[0132] Determine the number m of anchor nodes that an unknown node can connect to; and calculate the score of each anchor node based on the quantization function of the factors affecting the channel condition of each anchor node;

[0133] The parameter A is calculated based on the RSSI values ​​received by two anchor nodes at different distances, where parameter A represents the node transmission power;

[0134] Select the S anchor nodes with the highest scores from the m anchor nodes, and calculate the first distances between the S anchor nodes. Substitute the first distances and parameter A into the attenuation model between the signal and the distance to obtain S first channel attenuation parameters n1; where 0 <S≤m;

[0135] Normalizing the scores of the S anchor nodes, and combining the S first channel attenuation parameters n1 to calculate the second channel attenuation parameter n2;

[0136] Using the RSSI value of the anchor node received at the node to be located, and substituting the parameter A and the second channel attenuation parameter n2 into the attenuation model between the signal and the distance, the second distance between the node to be located and the anchor node is calculated;

[0137] According to the second distance, the coordinates of the node to be located are determined using the traditional three-sided positioning method to complete the positioning.

[0138] Example 3

[0139] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0140] quantizing at least one factor affecting the channel condition to obtain a corresponding quantization function;

[0141] Determine the number m of anchor nodes that an unknown node can connect to; and calculate the score of each anchor node based on the quantization function of the factors affecting the channel condition of each anchor node;

[0142] The parameter A is calculated based on the RSSI values ​​received by two anchor nodes at different distances, where parameter A represents the node transmission power;

[0143] Select the S anchor nodes with the highest scores from the m anchor nodes, and calculate the first distances between the S anchor nodes. Substitute the first distances and parameter A into the attenuation model between the signal and the distance to obtain S first channel attenuation parameters n1; where 0 <S≤m;

[0144] Normalizing the scores of the S anchor nodes, and combining the S first channel attenuation parameters n1 to calculate the second channel attenuation parameter n2;

[0145] Using the RSSI value of the anchor node received at the node to be located, and substituting the parameter A and the second channel attenuation parameter n2 into the attenuation model between the signal and the distance, the second distance between the node to be located and the anchor node is calculated;

[0146] According to the second distance, the coordinates of the node to be located are determined using the traditional three-sided positioning method to complete the positioning.

[0147] Obviously, the above embodiments of the present invention are only examples for clearly explaining the present invention, and are not intended to limit the implementation methods of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A wireless sensor network node positioning method using fuzzy decision making, characterized in that: The methods described include the following: quantizing at least one factor affecting the channel condition to obtain a corresponding quantization function; Determine the number m of anchor nodes that the unknown node can connect to; and calculate the score of each anchor node in combination with the corresponding quantization function; The parameter A is calculated based on the RSSI values ​​received by two anchor nodes at different distances, where parameter A represents the node transmission power; Select the S anchor nodes with the highest scores from the m anchor nodes, and calculate the first distances between the S anchor nodes. According to the first distances and parameter A, S first channel attenuation parameters n1 are calculated; where 0 <S≤m; Normalizing the scores of the S anchor nodes, and combining the S first channel attenuation parameters n1 to calculate the second channel attenuation parameter n2; Using the RSSI value of the anchor node received at the node to be located, and according to the parameter A and the second channel attenuation parameter n2, a second distance between the node to be located and the anchor node is calculated; According to the second distance, the coordinates of the node to be located are determined by using the traditional three-sided positioning method to complete the positioning; The calculation process of the second channel attenuation parameter n2 includes: Select the S=3 anchor nodes with the highest scores, denoted as a, b, c, and their corresponding scores are p a ,p b ,p c ; The calculation formula of the second channel attenuation parameter n2 is as follows: In the formula, g a ,g b ,g c Respectively by p a ,p b ,p c After normalization, g a +g b +g c =1; n ab Indicates the first distance d between anchor node a and anchor node b ab Get the first channel attenuation parameter; n ac represents the first distance d between anchor node a and anchor node c ac Get the first channel attenuation parameter; n bc represents the first distance d between anchor node b and anchor node c bc A first channel attenuation parameter is obtained.

2. The wireless sensor network node positioning method using fuzzy decision making according to claim 1 is characterized in that: The factors affecting the channel conditions include one or more of the distance between the node and the wall, obstacles near the node, the distance between the nodes, and the density of the nodes.

3. The wireless sensor network node positioning method using fuzzy decision making according to claim 2 is characterized in that: The quantization function of the distance between the node and the wall is as follows: Where (x, y) represents the coordinates of the anchor node; x m Indicates the maximum horizontal length of the positioning area, y m Indicates the maximum vertical length of the positioning area.

4. The wireless sensor network node positioning method using fuzzy decision making according to claim 2 is characterized in that: The quantization function of the obstacles near the node is as follows: In the formula, k represents the number of neighboring anchor nodes that the anchor node to be evaluated can connect to; n i represents the channel attenuation parameter n between the anchor node to be evaluated and the i-th adjacent anchor node; μ(n i ) represents the average of k values ​​of n.

5. The wireless sensor network node positioning method using fuzzy decision making according to claim 2 is characterized in that: The quantization function of the distance between nodes is as follows: In the formula, RSSI(d x ) is the RSSI value received by the node to be located from the anchor node; RSSI(d m ) is the RSSI value of the two nodes when the distance is 10m; RSSI(d0) is the RSSI value of the two nodes when the distance between them is 1m.

6. The wireless sensor network node positioning method using fuzzy decision making according to claim 2 is characterized in that: The quantification function of the node density is as follows: 。 7. The wireless sensor network node positioning method using fuzzy decision making according to claim 1, characterized in that: According to the importance of each factor affecting the channel condition, a weight matrix is ​​set; combined with the quantization function of the factors affecting the channel condition, the score of each anchor node is calculated as follows: p i =BR i T where B represents the weight matrix, and R i represents the influence channel condition factor matrix of the i-th anchor node, where 0 < i ≤ m, and m represents the number of anchor nodes that the unknown node can communicate with; p i represents the score of the i-th anchor node.

8. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images