CN110662291A - Three-dimensional DV-Hop positioning method based on Hop count weighting and Hop moment optimization - Google Patents

Abstract

The invention relates to a three-dimensional DV-Hop positioning method based on hop number weighting and hop moment optimization, and belongs to the field of wireless sensor network node positioning. The DV-Hop positioning method belongs to an algorithm that does not require ranging, and can greatly reduce the positioning cost in the process of node positioning. The present invention improves the calculation method of the distance value between nodes in the DV-Hop positioning method, and weights the number of hops between nodes. Calculate and optimize the jump moment between nodes with minimum mean square error and unbiased estimation, reduce the distance calculation error between nodes, and make node positioning more accurate. The algorithm can improve the positioning accuracy and expand the application scope while reducing the positioning cost.

Description

A 3D DV-Hop Localization Method Based on Hop Number Weighting and Hop Moment Optimization

technical field

The invention relates to a three-dimensional DV-Hop positioning method based on hop number weighting and hop moment optimization, and belongs to the field of wireless sensor network node positioning.

Background technique

In the fields of earthquake monitoring, forest fire prevention and control, and monitoring of livestock habits, it is not only necessary to extract various monitoring data, but also to locate target nodes in real time after analyzing the data, so as to realize the significance of monitoring. Although the traditional DV-Hop localization method belongs to the node localization method without ranging, the localization cost is low, but there is a large error in the process of calculating the minimum hop value between nodes and the average hop moment of the anchor node, resulting in the loss of target nodes. The positioning error is large, which makes it impossible to accurately locate the earthquake source, forest fire location, livestock activity range and trajectory in the actual application process, and this algorithm is only used to locate nodes in a two-dimensional plane. It is impossible to use this positioning method to locate target nodes in these terrains combined with plateau basins and other three-dimensional spatial terrains. Therefore, it is necessary to improve the positioning method to reduce the node positioning error, and at the same time extend it to three-dimensional application spaces such as mountainous terrain to ensure positioning. The method can be adapted to practical application scenarios. The invention comes from the Yunnan Provincial Technology Innovation Talent Project (2019HB113) and the National Natural Science Foundation of China (61562051).

SUMMARY OF THE INVENTION

The invention provides a three-dimensional DV-Hop positioning method based on hop number weighting and hop moment optimization, which is used for reducing the positioning error of wireless sensor network nodes and broadening the application field of positioning.

The technical scheme adopted in the present invention is: a three-dimensional DV-Hop positioning method based on hop number weighting and hop moment optimization, and the specific steps are as follows:

Step1. Randomly drop 1000 wireless sensor network nodes in a three-dimensional space of 100×100×100, including known anchor nodes and unknown nodes. First, from each anchor node to itself as the center of the sphere, with R as the communication radius of each node All nodes _adjacent to the _{sphere within} the _{sphere broadcast} data _packets containing their own location information. The identification number ID _i , the identification number Id _i of the neighbor node receiving the anchor node beacon message (empty at this time), the anchor node's own coordinates (x _i , y _i , z _i ) and other unknown nodes and the anchor node The distance dis _i (the field of the anchor node itself is 0), the neighbor node continues to use itself as the center of the sphere after receiving the data packet, and broadcasts the data packet to the neighbor nodes within the same communication range;

In the broadcast process, in order to prevent the data from being broadcast endlessly, the algorithm adopts the controllable flooding method, that is, when a node receives a data packet with a duplicate ID number, it compares the newly calculated distance with the anchor node to the original one in the table. Compared with the saved distance information, if the new distance is less than the original distance, the new distance is used to replace the distance in the original table, and the new data packet is re-broadcast; otherwise, the new data packet is discarded and no longer forwarded. This strategy can not only ensure that the distance information stored by each unknown node is the shortest path to the anchor node, but also make the algorithm end;

Step2, each network node participating in the broadcast communication process establishes a routing vector table, and only retains the minimum hop value from other nodes. The data packet, the minimum hop value h _ij between the nodes participating in the communication process can be found through the routing vector table of each node;

Step3. Correct the minimum hop value h _ij between nodes calculated in Step 2. Since the nodes are within the communication range of a sphere with a radius of R, the distance between each anchor node and its distance is different. If the hop value between nodes is calculated by 1 hop It will cause the calculation deviation of the actual hop value. According to the difference in the RSSI value of another anchor node within its communication range sent and received by the anchor node, the hop number weight h' is constructed, and the hop value between the two anchor nodes is corrected to reduce Error, the number of hops is corrected by introducing the RSSI positioning model. The calculation formula of the RSSI value is as follows:

In the above formula, d ₀ is the reference distance, d is the distance between the node to be measured and the signal sending node, P _r (d) is the signal indication strength of the node at the distance d, and P _r (d ₀ ) is the RSSI receiving node. Receive the signal strength of the node at the reference distance d ₀ , d ₀ =1m, n _p is the path loss index related to the surrounding environment and distance, and takes a value within a certain range. In order to reflect the authenticity of the test results, Gaussian white noise X _{σ is} added, and its intensity value can be directly read in the calculation. The specific correction method is as follows:

The signal strength value received by the anchor node from the neighbor node is denoted as l, otherwise it is denoted as l', and the hop weight is the ratio of the two, namely:

The signal strength value received by the anchor node from the neighbor node is denoted as l, otherwise it is denoted as l', then the hop weight is the ratio of the two, namely:

The corrected minimum hop value between nodes is denoted as h′ _ij , h′ _ij = h _ij ×h′, and the corrected minimum hop value h′ _ij is stored in the node routing vector table, and is carried out to the next neighbor node. broadcast for subsequent use in the calculation of the average hop distance;

Step 4. According to the corrected minimum hop value h′ _ij between each anchor node obtained in Step 3, the known straight-line distance value is calculated according to the coordinates between the anchor node and other anchor nodes, and each anchor is calculated by the arithmetic mean method. The average jump moment value c _i of the nodes is optimized by the minimum mean square error and unbiased estimation to obtain the final average jump moment c′ _i of each anchor _node ;

In the traditional 3D DV-Hop positioning method, when calculating the average hop distance of each anchor node, firstly obtain the straight-line distance d _ij between it and another anchor node, and then use the straight-line distance and the corrected minimum hop value h′ between nodes _ij performs a division operation to obtain the average hop distance c _i of the anchor node, as shown in the following formula:

in In the above formula (x _i , y _i , z _i ), (x _j , y _j , z _j ) are the known parameter values of anchor nodes i and j on the x-axis, y-axis, and z-axis, respectively. According to the formula ( 1) Calculate the minimum mean square error of the node's average jump moment value:

f _∈ =∑(d _ij -c _i h′ _ij ) ²

According to the unbiased estimation, in order to minimize the minimum mean square error value, take the derivative of it and set it to 0, then we have:

2∑(d _ij -c _i h' _ij )h' _ij =0

The average jump moment of anchor node i is calculated as:

Since there are more than one anchor nodes included in the randomly placed network nodes in the three-dimensional space, the average hop distance value of the anchor nodes obtained from the above formula is more, and the network structure is not static, but dynamic and irregular. It is necessary to perform an averaging process on the average hop distance value of many anchor nodes obtained from the above formula to reduce the error caused by randomly selecting one of the anchor nodes' average hop distance value into the calculation, namely:

m is the number of anchor nodes that communicate with the anchor node to be found. In order to present the network topology near the unknown node and reduce the calculation error, the number of anchor nodes m should be a reasonable value to comprehensively reduce the positioning error. Combined with the unknown node, it is The idea of coordinate positioning in the three-dimensional space involving the x-axis, y-axis, and z-axis, so as to determine m=3 is appropriate, and the final calculation formula of the average jump distance of anchor node i is:

Step5. According to the revised minimum hop value h′ _ij between nodes obtained in Step 3, and the average jump moment value c′ _i of each anchor node obtained in Step 4, the two can be multiplied to obtain the unknown node to be sought. The straight-line distance d _i between D and its nearest anchor node i, namely:

d _i =h′ _ij ×c′ _i

Then, a distance calculation equation system containing unknown node coordinate parameters x, y, and z is constructed from the known distance values between the above unknown nodes and their surrounding anchor nodes:

In the above formula (x ₁ , y ₁ , z ₁ ), (x ₂ , y ₂ , z ₂ ), (x ₃ , y ₃ , z ₃ ), …, (x _n , y _n , z _n ), ( x, y, z), are the known and unknown parameter values of the n anchor nodes and the unknown node D on the coordinate axes x, y, z, respectively, d ₁ , d ₂ , d ₃ , ..., d _n are the corrected values The distance value between the unknown node and the known node of ;

The first to n-1 equations and the last equation are respectively calculated by the difference value: AX=v The coordinate value calculation matrix of the unknown node D in the three-dimensional space is transformed by the matrix solution method:

X=(A ^T A) ^-1 A ^T b

Among them, A is the coordinate difference parameter matrix, b is the coordinate square difference parameter matrix, and A ^T is the transpose matrix of the coordinate difference parameter matrix;

Step6. Calculate the average positioning accuracy value of all unknown nodes from the estimated coordinate values (x, y, z) of the unknown nodes calculated in Step 5 and the initial coordinate values (x′ _i , y′ _i , z′ _i ) of all unknown nodes accuracy to prove that this method can greatly reduce the positioning error of unknown nodes:

In the formula, x′ _i , y′ _i , and z′ _i are the initial known coordinate values of unknown node i on the x-axis, y-axis, and z-axis, respectively, n is the number of unknown nodes, and R is the communication radius of unknown nodes.

The beneficial effects of the invention are: the invention improves the calculation method of the distance value between nodes in the DV-Hop positioning method, performs weighted calculation on the number of hops between nodes, and performs minimum mean square error and unbiased estimation optimization on the hop moment between nodes. processing, reducing the distance calculation error between nodes, so that the node positioning is more accurate, and the method is applied from the two-dimensional space to the three-dimensional space, so that the algorithm can reduce the positioning cost, while also improving the positioning accuracy and expanding the scope of application .

Description of drawings

Fig. 1 is the flow chart of the present invention;

Fig. 2 is the relation diagram of anchor node ratio and unknown node average positioning accuracy value of the present invention;

FIG. 3 is a broken line statistical graph of the average positioning accuracy of the present invention changing with the total number of nodes.

Fig. 4 is an unknown node average positioning accuracy value trend chart of the present invention;

FIG. 5 is a distribution diagram of wireless sensor network nodes in a cube three-dimensional space according to an embodiment of the present invention.

Detailed ways

The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

Embodiment 1: As shown in Figure 1, a kind of 3D DV-Hop positioning method based on hop number weighting and hop moment optimization, the method steps are as follows:

Step1. Randomly drop 1000 wireless sensor network nodes in a three-dimensional space of 100×100×100, including known anchor nodes and unknown nodes. First, from each anchor node to itself as the center of the sphere, with R as the communication radius of each node All nodes _adjacent to the _{sphere within} the _{sphere broadcast} data _packets containing their own location information. The identification number ID _i , the identification number Id _i of the neighbor node receiving the anchor node beacon message (empty at this time), the anchor node's own coordinates (x _i , y _i , z _i ) and other unknown nodes and the anchor node The distance dis _i (the field of the anchor node itself is 0), the neighbor node continues to use itself as the center of the sphere after receiving the data packet, and broadcasts the data packet to the neighbor nodes within the same communication range;

In the broadcast process, in order to prevent the data from being broadcast endlessly, the algorithm adopts the controllable flooding method, that is, when a node receives a data packet with a duplicate ID number, it compares the newly calculated distance with the anchor node to the original one in the table. Compared with the saved distance information, if the new distance is less than the original distance, the new distance is used to replace the distance in the original table, and the new data packet is re-broadcast; otherwise, the new data packet is discarded and no longer forwarded. This strategy can not only ensure that the distance information stored by each unknown node is the shortest path to the anchor node, but also make the algorithm end;

Step2, each network node participating in the broadcast communication process establishes a routing vector table, and only retains the minimum hop value from other nodes. The data packet, the minimum hop value h _ij between the nodes participating in the communication process can be found through the routing vector table of each node;

Step3. Correct the minimum hop value h _ij between nodes calculated in Step 2. Since the nodes are within the communication range of a sphere with a radius of R, the distance between each anchor node and its distance is different. If the hop value between nodes is calculated by 1 hop It will cause the calculation deviation of the actual hop value. According to the difference in the RSSI value of another anchor node within its communication range sent and received by the anchor node, the hop number weight h' is constructed, and the hop value between the two anchor nodes is corrected to reduce Error, the number of hops is corrected by introducing the RSSI positioning model. The calculation formula of the RSSI value is as follows:

In the above formula, d ₀ is the reference distance, d is the distance between the node to be measured and the signal sending node, P _r (d) is the signal indication strength of the node at the distance d, and P _r (d ₀ ) is the RSSI receiving node. Receive the signal strength of the node at the reference distance d ₀ , d ₀ =1m, n _p is the path loss index related to the surrounding environment and distance, and takes a value within a certain range. In order to reflect the authenticity of the test results, Gaussian white noise X _{σ is} added, and its intensity value can be directly read in the calculation. The specific correction method is as follows:

The signal strength value received by the anchor node from the neighbor node is denoted as l, otherwise it is denoted as l', and the hop weight is the ratio of the two, namely:

The corrected minimum hop value between nodes is recorded as h _ij , h′ _ij = h _ij ×h′, and the corrected minimum hop value h′ _ij is stored in the node routing vector table, and is carried out to the next neighbor node. broadcast for subsequent use in the calculation of the average hop distance;

Step 4. According to the corrected minimum hop value h′ _ij between each anchor node obtained in Step 3, the known straight-line distance value is calculated according to the coordinates between the anchor node and other anchor nodes, and each anchor is calculated by the arithmetic mean method. The average jump moment value c _i of the nodes is optimized by the minimum mean square error and unbiased estimation to obtain the final average jump moment c′ _i of each anchor _node ;

In the traditional 3D DV-Hop positioning method, when calculating the average hop distance of each anchor node, firstly obtain the straight-line distance d _ij between it and another anchor node, and then use the straight-line distance and the corrected minimum hop value h′ between nodes _ij performs a division operation to obtain the average hop distance c _i of the anchor node, as shown in the following formula:

in

In the above formula (x _i , y _i , z _i ), (x _j , y _j , z _j ) are the known parameter values of anchor nodes i and j on the x-axis, y-axis, and z-axis, respectively. According to the formula ( 1) Calculate the minimum mean square error of the node's average jump moment value:

f _∈ =Σ(d _ij -c _i h′ _ij ) ²

According to the unbiased estimation, in order to minimize the minimum mean square error value, take the derivative of it and set it to 0, then we have:

2∑(d _ij -c _i h' _ij )h' _ij =0

The average jump moment of anchor node i is calculated as:

Since there are more than one anchor nodes included in the randomly placed network nodes in the three-dimensional space, the average hop distance value of the anchor nodes obtained from the above formula is more, and the network structure is not static, but dynamic and irregular. It is necessary to perform an averaging process on the average hop distance value of many anchor nodes obtained from the above formula to reduce the error caused by randomly selecting one of the anchor nodes' average hop distance value into the calculation, namely:

m is the number of anchor nodes that communicate with the anchor node to be found. In order to present the network topology near the unknown node and reduce the calculation error, the number of anchor nodes m should be a reasonable value to comprehensively reduce the positioning error, and the unknown node is The idea of coordinate positioning in the three-dimensional space involving the x-axis, y-axis, and z-axis, so as to determine m=3 is appropriate, and the final calculation formula of the average jump distance of anchor node i is:

Step5. According to the revised minimum hop value h′ _ij between nodes obtained in Step 3, and the average jump moment value c′ _i of each anchor node obtained in Step 4, the two can be multiplied to obtain the unknown node to be sought. The straight-line distance d _i between D and its nearest anchor node i, namely:

d _i =h′ _ij ×c′ _i

Then, a distance calculation equation system containing unknown node coordinate parameters x, y, and z is constructed from the known distance values between the above unknown nodes and their surrounding anchor nodes:

In the above formula (x ₁ , y ₁ , z ₁ ), (x ₂ , y ₂ , z ₂ ), (x ₃ , y ₃ , z ₃ ), …, (x _n , y _n , z _n ), ( x, y, z), are the known and unknown parameter values of n anchor nodes and unknown node D on the coordinate axes x, y, z, respectively, d ₁ , d ₂ , d ₃ , ..., d _n is Corrected distance between unknown nodes and known nodes;

The 1st to n-1 equations are respectively calculated with the difference of the last equation: AX=b The coordinate value calculation matrix of the unknown node D in the three-dimensional space is transformed by the matrix solution method:

X=(A ^T A) ^-1 A ^T b

Among them, A is the coordinate difference parameter matrix, b is the coordinate square difference parameter matrix, and A ^T is the transpose matrix of the coordinate difference parameter matrix;

Step6. Calculate the average positioning accuracy value of all unknown nodes from the estimated coordinate values (x, y, z) of the unknown nodes calculated in Step 5 and the initial coordinate values (x′ _i , y′ _i , z′ _i ) of all unknown nodes accuracy to prove that this method can greatly reduce the positioning error of unknown nodes:

In the formula, x′ _i , y′ _i , and z′ _i are the initial known coordinate values of unknown node i on the x-axis, y-axis, and z-axis, respectively, n is the number of unknown nodes, and R is the communication radius of unknown nodes.

The working principle of the present invention is as follows: first, each network node broadcasts data packets, and establishes a routing vector table, which records the minimum hop value between nodes, and applies the hop weight value to the hops between adjacent nodes within the communication radius of each node. The value is weighted and corrected, the corrected jump value is used to calculate the average jump moment value of the node, and the minimum mean square error and unbiased estimation are used to optimize the average jump moment value, and finally the average jump moment of the anchor node is calculated by the mean operation. The final average jump moment is obtained by processing the average jump moment, and the product operation is performed with the average jump moment and the corrected minimum jump value between the nodes to calculate the straight-line distance between the unknown node and the anchor node, and then the distance equation system between the unknown node and each anchor node is constructed. , and finally use the matrix transformation to calculate the estimated coordinate value of the unknown node in the three-dimensional space.

Further, the steps in the application are described by the following examples:

In order to verify that the improved 3D-DVHop localization method is better than the pre-improved 3D-DVHop localization method, a conclusion is drawn by comparing the change trend of the average localization accuracy value of the two algorithms for unknown nodes under different parameters. .

In the three-dimensional space of a cube with a side length of 100m, 1000 wireless sensor network nodes are randomly placed, as shown in Figure 5.

As shown in Figure 2, when the node communication radius R is set to 60m and the total number of nodes is set to 1000, under the conditions of different anchor node ratios, the algorithm program is set to run 100 times in a loop. The traditional 3D-DVHop algorithm is improved with the present invention. The trend of the average positioning accuracy of unknown nodes of the 3D-DVHop algorithm changes with the increase of the proportion of anchor nodes. It can be concluded that:

When the proportion of anchor nodes is 15%, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.5326, and the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.6287. Compared with the latter, the former algorithm has The accuracy value drops by 0.0961; when the anchor node ratio is 30%, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.4434, and the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.5960, the former is higher than the latter. When the anchor node ratio is 45%, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.4153, and the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.5412, the former is higher than the latter. The precision value is 0.1259 lower. It can be seen that in the two algorithms, the average positioning accuracy value of unknown nodes is inversely proportional to the ratio of anchor nodes, and the improved 3D-DVHop algorithm of the present invention has a downward trend in the average positioning accuracy value of unknown nodes compared with the traditional 3D-DVHop algorithm. DVHop algorithm is more obvious;

As shown in Figure 3, when R is 60m and the ratio of anchor nodes is 30%, the algorithm program is set to run 100 times in a cycle under the condition of different total number of nodes, the difference between the traditional 3D-DVHop algorithm and the improved 3D-DVHop algorithm of the present invention When the average positioning accuracy of unknown nodes varies with the total number of nodes, it can be concluded that:

When the total number of nodes is 400, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.5089, and the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.6028. The former is 0.0939 lower than the latter; When the total number is 700, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.4834, the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.5984, and the former is 0.1150 lower than the latter's accuracy value; when the total number of nodes is 1000, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.4613, and the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.5896, and the former is 0.1283 lower than the latter. From this analysis, it can be seen that the average positioning accuracy of the unknown nodes of the traditional 3D-DVHop algorithm basically does not change with the increase of the total number of nodes, and the average positioning accuracy of the unknown nodes of the improved 3D-DVHop algorithm of the present invention decreases with the increase of the total number of nodes. But its downward trend is not obvious; in general, the improved 3D-DVHop algorithm of the present invention locates unknown nodes more accurately than the traditional 3D-DVHop algorithm;

As shown in Figure 4, when the ratio of anchor nodes to the total number of nodes is set to 30% and the total number of nodes is set to 1000, the algorithm program is set to run 100 times in a loop under different node communication radius conditions. The traditional 3D-DVHop The algorithm and the improved 3D-DVHop algorithm of the present invention show that the average positioning accuracy of the unknown node changes with the change of the node communication radius R. It can be concluded that:

When the node communication radius R is 40m, the average positioning accuracy of the unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.6789, the average positioning accuracy of the unknown nodes of the traditional 3D-DVHop algorithm is 0.8923, the former is 0.2134 lower than the latter; when When the node communication radius R is 70m, the average positioning accuracy of the unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.4834, and the average positioning accuracy of the unknown nodes of the traditional 3D-DVHop algorithm is 0.6213. The former is 0.1379 lower than the latter; when When the node communication radius R is 100m, the average positioning accuracy of unknown nodes of the improved 3D-DVHop algorithm of the present invention is 0.3213, and the average positioning accuracy of unknown nodes of the traditional 3D-DVHop algorithm is 0.4857, the former is 0.1644 lower than the latter. It can be seen from the above experimental data that in the two algorithms, the average positioning accuracy value of unknown nodes is inversely proportional to the communication radius R, and the improved 3D-DVHop algorithm of the present invention locates unknown nodes more accurately than the traditional 3D-DVHop algorithm.

Embodiment 2: Embodiment 1 is the overall algorithm of the present invention. The algorithm can be applied to a variety of practical situations to solve practical technical problems. This embodiment takes the prevention and control of forest fires as an example. The application is specified.

Step1. Construct a three-dimensional coordinate system in the three-dimensional space of the original forest with a fixed range, and randomly place 1000 wireless sensor network nodes in it, including known anchor nodes and unknown nodes, and use the known anchor nodes in the forest to locate the unknown nodes. ①If the fire position is located at the anchor node, the position information can be directly obtained; ②If the fire is located at the unknown node, use this positioning method to locate it and obtain the position information; ③If the fire position is not at the deployed node, select The nearest anchor node or unknown node is located, and the position information is obtained in time; if the above 2 and 3 situations occur, this positioning method needs to be used to locate the forest fire in time. First, each anchor node in the forest is a ball to itself In the sphere with R as the communication radius of each node, all nodes in the adjacent forest broadcast data packets containing their own location information, the data packet format is {ID _i , Id _i , x _i , y _i , zi _i , dis _i }, which contains the identification number ID i of the anchor node, the identification number Id _i of the neighbor node receiving the anchor node beacon message ₍ empty at this time), the coordinates of the anchor node itself (x _i , y _{i )} , z _i ) and the distance between other unknown nodes and the anchor node dis _i (the field of the anchor node itself is 0), the neighbor node continues to use itself as the center of the sphere after receiving the data packet, and broadcasts data to the neighbor nodes within the same communication range Bag;

In the broadcast process, in order to prevent the data from being broadcast endlessly, the algorithm adopts the controllable flooding method, that is, when a node receives a data packet with a duplicate ID number, it compares the newly calculated distance with the anchor node to the original one in the table. Compared with the saved distance information, if the new distance is less than the original distance, the new distance is used to replace the distance in the original table, and the new data packet is re-broadcast; otherwise, the new data packet is discarded and no longer forwarded. This strategy can not only ensure that the distance information stored by each unknown node is the shortest path to the anchor node, but also make the algorithm end;

Step 2. Each network node in the forest participating in the broadcast communication process establishes a routing vector table, and only retains the data packets with the smallest hop value from other nodes. Through the routing vector table of each node, the minimum hop value h _ij between the nodes participating in the communication process can be found. ;

Step3. Correct the minimum hop value h _ij between nodes in the forest calculated in Step 2. Since the nodes are within the communication range of a sphere with R as the radius, the distances between the anchor nodes are different. If the hop values between nodes are all 1 Hop counting will cause the calculation deviation of the actual hop value. According to the difference between the RSSI value sent and received by the anchor node and the RSSI value of another anchor node within its communication range, the hop weight value h' is constructed, and the hop value between the two anchor nodes is corrected. , reduce the error, and correct the number of hops by introducing the RSSI positioning model. The calculation formula of the RSSI value is as follows:

In the above formula, d ₀ is the reference distance, d is the distance between the node to be measured and the signal sending node, P _r (d) is the signal indication strength of the node at the distance d, and P _r (d ₀ ) is the RSSI receiving node. Receive the signal strength of the node at the reference distance d ₀ , d ₀ =1m, n _p is the path loss index related to the surrounding environment and distance, and takes a value within a certain range. Considering the objective natural factors existing in the forest space, in order to reflect the authenticity of the positioning results, Gaussian white noise X _σ is added to simulate the real natural environment, and its intensity value can be directly read in the calculation. The specific correction method is as follows:

The signal strength value received by the anchor node in the forest from the neighbor node is denoted as l, otherwise it is denoted as l', then the hop weight is the ratio of the two, namely:

The corrected minimum hop value between nodes is recorded as h _ij , h′ _ij = h _ij ×h′, and the corrected minimum hop value h′ _ij between each node in the forest is stored in the node routing vector table, and the next Neighbor nodes broadcast for use in subsequent calculation of average hop distance;

Step 4. According to the minimum hop value h′ _ij between the nodes in the revised forest that has been obtained in Step 3, the known straight-line distance value is calculated according to the coordinates between the anchor node and other anchor nodes, and each anchor node is calculated by the arithmetic mean method. The average jumping moment value c _i , the average jumping moment value c _i of each anchor node is optimized through the minimum mean square error and unbiased estimation, and the final average jumping moment c′i of each anchor node is obtained;

In the traditional 3D DV-Hop positioning method, when calculating the average hop distance of each anchor node, firstly obtain the straight-line distance d _ij between it and another anchor node, and then use the straight-line distance and the corrected minimum hop value h′ between nodes _ij performs a division operation to obtain the average hop distance c _i of the anchor node, as shown in the following formula:

in

In the above formula (x _i , y _i , z _i ), (x _j , y _j , z _j ) are the known parameter values of anchor nodes i and j on the x-axis, y-axis, and z-axis, respectively. According to the formula ( 1) Calculate the minimum mean square error of the node's average jump moment value:

f _∈ =∑(d _ij -c _i h′ _ij ) ²

According to the unbiased estimation, in order to minimize the minimum mean square error value, take the derivative of it and set it to 0, then we have:

2∑(d _ij -c _i h′ _ij )h′ _ij =0

The average jump moment of anchor node i is calculated as:

Since the network nodes randomly placed in the fixed forest space include more than one anchor node, the average hop distance value of the anchor nodes obtained by the above formula is more, and the network structure in the space is not static, but dynamic and not static. Therefore, it is necessary to perform an averaging process on the average hop distance value of many anchor nodes obtained from the above formula, so as to reduce the error caused by randomly selecting the average hop distance value of one of the anchor nodes into the calculation, namely:

m is the number of anchor nodes that communicate with the anchor node to be found. In order to present the network topology near the unknown node and reduce the calculation error, the number of anchor nodes m should be a reasonable value to comprehensively reduce the positioning error, and the unknown node is The idea of coordinate positioning in the three-dimensional space involving the x-axis, y-axis, and z-axis, so as to determine m=3 is appropriate, and the final calculation formula of the average jump distance of anchor node i is:

Step5. According to the minimum hop value h′ _ij between nodes in the revised forest obtained in Step 3, and the average jump moment value c′ _i of each anchor node obtained in Step 4, the two can be multiplied to obtain the Find the straight-line distance d _i between the unknown node D and its nearest anchor node i, namely:

d _i =h′ _ij ×c′ _i

Then, a distance calculation equation system containing unknown node coordinate parameters x, y, and z is constructed from the known distance values between the above unknown nodes and their surrounding anchor nodes:

In the above formula (x ₁ , y ₁ , z ₁ ), (x ₂ , y ₂ , z ₂ ), (x ₃ , y ₃ , z ₃ ), …, (x _n , y _n , z _n ), ( x, y, z), are the known and unknown parameter values, d ₁ , d ₂ , d ₃ ,... d _n is the distance value between the corrected unknown node and the known node;

The 1st to n-1 equations are respectively calculated by difference with the last equation: AX=b The coordinate value calculation matrix of the unknown node D in the three-dimensional forest space is transformed by the matrix solution method:

X=(A ^T A) ^-1 A ^T b

Among them, A is the coordinate difference parameter matrix, b is the coordinate square difference parameter matrix, and A ^T is the transpose matrix of the coordinate difference parameter matrix;

Step6. The estimated coordinate values (x, y, z) of the unknown nodes in the three-dimensional forest space calculated in Step5 and the initial coordinate values (x′ _i , y′ _i , z′ _i ) of all unknown nodes in the three-dimensional forest space ) calculates the average positioning accuracy value of all unknown nodes to prove that this method can greatly reduce the positioning error of unknown nodes:

In the formula, x′ _i , y′ _i , and z′ _i are the initial known coordinate values of the unknown node i on the x-axis, y-axis, and z-axis of the three-dimensional forest space, n is the number of unknown nodes, and R is the unknown node communication radius.

Conclusion: The algorithm of the present invention is applied to the prevention and control of forest fires. First, the algorithm itself does not need to calculate the straight-line distance between nodes, which greatly reduces the positioning cost. Second, the algorithm corrects the number of hops and the hop moment between nodes. and optimization, reduce the positioning error of the fire location, ensure the timeliness and accuracy of rescue, and improve the quality of forest fire prevention and control. Finally, the practical application space of the algorithm is increased from two-dimensional to three-dimensional, so that the algorithm is not limited. In the two-dimensional space plane, its application scene perfectly fits the complex three-dimensional forest terrain.

The specific embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned embodiments, and can also be made within the scope of knowledge possessed by those of ordinary skill in the art without departing from the purpose of the present invention. Various changes.

Claims (1)

1. A three-dimensional DV-Hop positioning method based on Hop count weighting and Hop moment optimization is characterized in that: the method comprises the following specific steps:

step1, randomly launching 1000 wireless sensor network nodes in a 100 x 100 three-dimensional space, wherein the wireless sensor network nodes comprise known anchor nodes and unknown nodes, firstly, each anchor node broadcasts a data packet containing self-position information to all nodes adjacent to the anchor node within a sphere range taking the anchor node as a sphere center and taking R as a communication radius of each node, and the format of the data packet is { ID_i，Id_i，x_i，y_i，z_i，dis_iContains the ID of the anchor node_iIdentification number Id of neighbor node receiving the Anchor node Beacon message_iAt this time Id_iNull, anchor node self coordinates (x)_i，y_i，z_i) And distances dis of other unknown nodes from the anchor node_iThe field of the anchor node is 0, and the neighbor node continues to broadcast the data packet to the neighbor nodes in the same communication range by taking the neighbor node as the sphere center after receiving the data packet;

in the broadcasting process, the algorithm adopts a controllable flooding method, namely when a node receives a data packet with repeated ID number, the newly calculated distance between the anchor node and the distance information stored in the table is compared, if the new distance is less than the original distance, the new distance is used for replacing the distance in the original table, and the new data packet is broadcasted again; otherwise, discarding the new data packet and not forwarding;

step2, each network node participating in the broadcast communication process establishes a route vector table, only the data packet with the minimum hop value away from other nodes is reserved, and the minimum hop value h between the nodes participating in the communication process can be found through each node route vector table_ij；

Step3, and the minimum hop count h between the nodes calculated in Step2_ijCorrecting, wherein the distance between each anchor node and each anchor node is different in a sphere communication range taking R as a radius, if hop values among the nodes are all calculated by 1 hop, the calculation deviation of an actual hop value can be caused, according to the difference of RSSI values of another anchor node in the communication range of the anchor nodes, the RSSI values among two anchor nodes are corrected, the error is reduced, the RSSI values are corrected by introducing an RSSI positioning model, and the calculation formula of the RSSI values is as follows:

in the above formula, d₀D is the distance between the node to be measured and the signal transmitting node, P_r(d) Indicating strength for receiving signals from nodes at distance d, P_r(d₀) Receiving a reference distance d for an RSSI receiving node₀Signal indication strength at node, d₀＝1m，n_pAdding Gaussian white noise X to the path loss index related to the surrounding environment and distance_σThe intensity value can be directly read in the calculation, and the specific correction method is as follows:

recording the strength value of the signal indication received by the anchor node from the neighbor node as l, otherwise recording the strength value as l', and taking the hop count weight as the ratio of the two, namely:

the corrected minimum inter-node hop count value is recorded as h'_ij，h′_ij＝h_ijX h ', and calculating the corrected minimum inter-node hop count value h'_ijStoring the route vector table in the node route vector table, and broadcasting the route vector table to the next neighbor node for use in the calculation of the subsequent average hop distance;

step4, minimum hop count value h 'corrected by each anchor node obtained in Step 3'_ijCombining the anchor nodes with other anchor nodes according to a coordinate systemCalculating the known linear distance value, and calculating the average jump moment value c of each anchor node by an arithmetic mean method_iAverage jump moment value c of each anchor node through minimum mean square error and unbiased estimation_iOptimizing to obtain final average hop moment c 'of each anchor node'_i；

In the traditional three-dimensional DV-Hop positioning method, when the average Hop distance of each anchor node is calculated, the linear distance d between each anchor node and another anchor node is firstly calculated_ijAnd using the linear distance and the corrected minimum inter-node hop count value h'_ijDividing to obtain the average jump distance c of the anchor node_iAnd solving the minimum mean square error according to a calculation formula for calculating the average hop moment value of the node:

f_∈＝∑(d_ij-c_ih′_ij)²

according to the unbiased estimation, to minimize the minimum mean square error value, derivative is taken and made 0, then:

2∑(d_ij-c_ih′_ij)h′_ij＝0

and calculating to obtain the average hop moment of the anchor node i as follows:

an averaging process is carried out on the average hop distance values of the anchor nodes obtained in the formula so as to reduce errors caused by the fact that one of the average hop distance values of the anchor nodes is randomly selected and is brought into calculation, namely:

m is the number of anchor nodes communicating with the anchor node to be solved, and the thought that the unknown node is used for carrying out coordinate positioning in a three-dimensional space relating to an x axis, a y axis and a z axis is combined, so that m is determined to be 3, and the final calculation formula of the average jump distance of the anchor node i is as follows:

step5, and the corrected minimum inter-node hop count value h 'obtained in Step 3'_ijAnd the average hop moment value c 'of each anchor node obtained at Step 4'_iThe two can be multiplied to obtain the linear distance D between the unknown node D to be solved and the anchor node i which is close to the unknown node D_iNamely:

d_i＝h′_ij×c′_i

and then constructing a distance calculation equation set containing coordinate parameters x, y and z of the unknown nodes according to the known distance values between the unknown nodes and the surrounding anchor nodes:

in the above formula (x)₁，y₁，z₁)，(x₂，y₂，z₂)，(x₃，y₃，z₃)，…，(x_n，y_n，z_n) (x, y, z) known and unknown parameter values for n anchor nodes and unknown nodes D on the coordinate axes x, y, z, respectively, D₁，d₂，d₃，…，d_nThe corrected distance value between the unknown node and the known node is obtained;

the 1 st to n-1 st equations are respectively calculated by difference with the last equation: and (3) converting AX (x) b into a coordinate value calculation matrix of the unknown node D in the three-dimensional space by a matrix solving method:

X＝(A^TA)^-1A^Tb

wherein A is a coordinate difference parameter matrix, b is a coordinate squared difference parameter matrix, A^TA transpose matrix which is a coordinate difference parameter matrix;

step6, and the estimated coordinate values (x, y, z) of the unknown nodes calculated in Step5 and the initial coordinate values (x'_i、y′_i、z′_i) Calculating the average positioning accuracy values accurve of all unknown nodes to prove that the method can greatly reduce the positioning error of the unknown nodes:

in formula (II), x'_i、y′_i、z′_iThe initial known coordinate values of the unknown node i on the x axis, the y axis and the z axis are respectively, n is the number of the unknown nodes, and R is the communication radius of the unknown nodes.

Images