CN102438310A - Wireless sensor network positioning method based on trust rating - Google Patents

Abstract

The invention provides a wireless sensor network positioning method based on trust rating and belongs to the technical field of wireless communication. The method comprises the following steps: dividing a network into a plurality of grids serving as a cluster and randomly selecting one beacon node in the cluster to serve as a cluster head; distributing an ID (identity) number for each node and setting the initial GTR (global trust rating) value and LTR (local trust rating) value of each beacon node; respectively measuring the RSSI (received signal strength indicator) value of each beacon node and the RSSI value of each node to be measured and obtaining the distance information of each node to be measured and each beacon node; utilizing the cluster head to carry out positioning calculation on all the nodes to be measured in the cluster and updating the trust rating of each node; and after the current-round positioning is finished, starting next-round positioning, and obtaining the final position coordinates of the nodes to be measured and the trust rating of each beacon node until N-round positioning is finished. According to the invention, in the positioning calculation carried out on the nodes to be measured, the trust rating of each beacon node is referenced, thus the interference brought by faulty beacon nodes and channel noise can be effectively eliminated, and the node positioning accuracy is improved; and simultaneously, abnormal beacon nodes can be detected.

Description

A kind of localization method of the wireless sensor network based on degree of belief

Technical field

What the present invention relates to is the method in a kind of wireless sensor network technology field, specifically is a kind of localization method of the wireless sensor network based on degree of belief.

Background technology

Wireless sensor network (WSNs) is widely used in daily life and the military equipment, and the sensor node location is an importance of WSNs technology.Many application all are based on positional information, like forest fire monitoring, the tracking of wild animation, patient care etc.Sensor node is deployed in the special environment such as field, indoor, underground mine usually, yet in actual environment, beaconing nodes (Beacon) can not provide correct reference position information to node to be positioned (Node) because of fault own.Problems such as wireless channel existence reflection affected by environment, multipath transmisstion, ambient interferences, thereby, the measurement range error between beaconing nodes and the node to be positioned caused.Along with the WSNs application is extensive gradually, to the precision and the higher requirement of antijamming capability proposition of node locating.

Summary of the invention

This paper proposes a kind of Distributed localization method based on the management of sensor node degree of belief, to reach high accuracy and jamproof requirement in the position fixing process.

This method is introduced the global trusting degree (Global TrustRating is hereinafter to be referred as GTR) and local degree of belief (Local Trust Rating is hereinafter to be referred as the LTR) notion of beaconing nodes on cluster structured basis, make up trust degree matrix.Through many wheels refinement location Calculation, detect the reliability of Beacon node and locating information, the degree of belief of suspect node is punished, the degree of belief of safe node is rewarded, and by the trust matrix of new node on bunch first watch.The neighbor table information of node also can be upgraded with trusting the matrix change.This localization method is not only put forward the positioning accuracy of ability high node, can also detect abnormal nodes and containment channel disturbance to location influence.

The present invention realizes through following technical scheme, specifically may further comprise the steps:

Step 1: with the network grid sub-clustering, and a beaconing nodes in the picked at random bunch is as bunch head;

Step 2: in each sensor node distribute unique ID number, and the initial overall situation of Beacon node, local trust degree value and location times N are set;

Step 3: measure the RSSI value of Beacon node and Node node respectively, obtain the estimated distance of Node node and each Beacon node, set up neighbors information table separately;

Step 4: the location Calculation of all Node nodes in bunch head is responsible for bunch: when the number of the adjacent Beacon node of Node node to be measured during more than or equal to 4; According to nearest 4 the Beacon nodes of principles of selected of degree of belief ultimate range node as a reference; Use 4 location algorithm location, judge each Beacon node state; When the adjacent Beacon node of Node node to be measured has only 3, just use the least square method location; When the adjacent Beacon node of Node node to be measured number is less than 3, this Node node to be measured delocalization this time;

Step 5: take turns the degree of belief that the location Calculation result upgrades the Beacon node according to every;

Step 6: continuous repeating step four, five, after N the location Calculation, obtain the N time position location of Node node, this position is exactly the final position information of Node node.

Further, the described cluster-dividing method of step 1 may further comprise the steps:

1) confirms bunch number compute cluster center, and distribute bunch number;

2) calculate the distance at each Beacon node and bunch center, promptly add this bunch near certain cluster center, the nearest Beacon node of each bunch distance center point then becomes a bunch head first;

3) the Node node calculates the average distance of Beacon node in each bunch, adds and nearest one bunch of its average distance.

Further, the initial overall situation of the said beaconing nodes of said step 2, local trust degree value all are 1, and the span of degree of belief is: [0,1].

Further, the described estimated distance of said step 3 is: $d_j=d_0{e}^{\frac{P_j-{\mathrm{<figure-callout\; id="0"\; label="P"\; filenames="HDA0000093760200000011.png"\; state="\{\{state\}\}">P</figure-callout>}}_0}{10n}},$

Wherein, d _jBe the distance of j beaconing nodes and node to be measured, P _jBe j beaconing nodes RSSI value, P ₀Be that distance is d ₀Two sensor nodes between the reception signal strength signal intensity, d ₀Be reference distance, 1≤j≤N, N are the total number of node to be measured.

Further, the described node neighbor information of said step 3 table is: neighbor node ID sign, and corresponding range information.

Further, described 4 location algorithms of said step 4 may further comprise the steps:

1) select four Beacon nodes as location reference point according to the rise of degree of belief from nearest principle;

2) combine three wherein different nodes of these four reference nodes and be divided into four different groups, these four groups are used least-squares calculation;

3) if a certain group of range error is all little than other groups, then judge to remove this organize in that beaconing nodes three nodes be abnormal nodes; Otherwise four reference nodes this time selecting are no abnormal; The result of calculation of this group is the position location.

Further, the described renewal degree of belief of said step 5 is: the degree of belief to unusual beaconing nodes is punished, promptly reduces degree of belief; Degree of belief to normal Beacon node is rewarded, i.e. degree of enhancing trust.

Compared with prior art, the present invention has advantages such as low computation complexity, high accuracy, the accuracy that not only trust checking location in location Calculation, but also can detect abnormal nodes, infer the interference that channel possibly exist.

Description of drawings

Fig. 1 wireless sensor network node layout

Fig. 2 is based on the wireless sensor network locating method flow chart of degree of belief

Embodiment

Below method of the present invention is further described: present embodiment provided detailed execution mode and concrete operating process, but protection scope of the present invention is not limited to following embodiment being to implement under the prerequisite with technical scheme of the present invention.

As shown in Figure 1; The wireless sensor network environment parameter of present embodiment is following: 20 beaconing nodes (to call the Beacon node in the following text), 30 nodes to be measured (to call Node node to be measured in the following text) are deployed in 100 * 100 the zone at random, and their communication radius is respectively 20 and 10.Behind the node deployment, adopt method sub-clustering of the present invention, each node distributes unique sign (ID).And global trusting degree (the GlobalTrust Rating of initialization beaconing nodes; Hereinafter to be referred as GTR), local trust degree (Local Trust Rating; Hereinafter to be referred as LTR) be 1; Global trusting threshold value (Global Trust Threshold is hereinafter to be referred as GTT), local trust threshold value (LocalTrust Threshold is hereinafter to be referred as LTT) are made as 0.75,0.5 respectively.Then, carry out the Node tested point to be measured location and the Beacon node confidence of being correlated with thereof and upgrade, and repeat, finish until the location.

Specifically may further comprise the steps, flow process is as shown in Figure 2:

Step 1: set up rectangular coordinate system, obtain the positional information of each Beacon node in the wireless sensor network, according to method sub-clustering of the present invention, and a beaconing nodes in the picked at random bunch is as bunch head,

Specifically may further comprise the steps:

1) according to the size of network size, the node distributed areas are divided into several grids---bunch, the initial center that the center of grid is bunch, and be that each bunch distributes bunch number;

2) distance at each Beacon node calculating and each bunch center promptly adds this bunch near certain cluster center, and the nearest Beacon node of each bunch distance center point then becomes a bunch head first;

3) the Node node calculates the average distance of Beacon node in each bunch, adds and nearest one bunch of its average distance;

Step 2: for bunch in each sensor node distribute unique ID number, and the initial GTR of Beacon node is set, the LTR value is 1, the span of this degree of belief: [0,1], initialization GTT, LTT are respectively 0.75,0.5;

Step 3: acknowledge(ment) signal intensity (the ReceivedSignal Strength Indicator that measures Beacon node and Node node to be measured respectively; Hereinafter to be referred as RSSI) value; Obtain the estimated distance of Node node to be measured and each Beacon node, set up neighbors information table separately;

Described estimated distance is:

$d_{i,j}=d_0{e}^{\frac{P_{i,j}-{\mathrm{<figure-callout\; id="0"\; label="P"\; filenames="HDA0000093760200000011.png"\; state="\{\{state\}\}">P</figure-callout>}}_0}{10n}},$

Wherein, the Beacon node is identical with the range measurement model of Node node to be measured, d _{I, j}Be the distance of i Node point node to be measured and j Beacon node, P _{I, j}Be the RSSI value that i Node node to be measured receives j Beacon node, P ₀Be that distance is d ₀Two sensor nodes between the RSSI value, d ₀Be reference distance, 1≤j≤N, N are the total number of node to be measured, and n is an attenuation coefficient.

Present embodiment is got d ₀=0.1m, n=2.

Described neighbors information table is: neighbor node ID sign, and corresponding range information.

Step 4: the neighbors information table that all nodes send in bunch first merge bunch, and be responsible for the location Calculation of node to be measured.Choose four Beacon nodes node as a reference, utilize least square method to position;

When the number of the adjacent Beacon node of certain Node node to be measured during,, use 4 location algorithms described below location according to nearest 4 the Beacon nodes of principles of selected of degree of belief ultimate range node as a reference more than or equal to 4; When the adjacent Beacon node of certain Node node to be measured has only 3, just use 3 point location algorithms (least square) location.When the adjacent Beacon node of adjacent certain Node node to be measured number is less than 3, this Node node to be measured delocalization this time.

Select 4 Beacon nodes node as a reference, obtain following equation group:

$\left\{\begin{array}{c}\left\{\begin{array}{c}{(x-x_1)}^2+{(y-y_1)}^2={d_1}^2\\ {(x-x_2)}^2+{(y-y_2)}^2={d_2}^2\\ {(x-x_3)}^2+{(y-y_3)}^2={d_3}^2\end{array}\right.\\ {(x-x_4)}^2+{(y-y_4)}^2={d_4}^2\end{array}\right.$

At first, selecting wherein, first three equation adopts three-point fix algorithm computation Node node coordinate N ₁(x, y), and the error sum f of this node and these 3 Beacon nodes _Error(i):

$f_{\mathrm{error}}\left(i\right)=\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}[{({x^x}_i-x_j)}^2+{({y^y}_i-y_j)}^2-d_j^2]i=\mathrm{1,2,3,4};j=\mathrm{1,2,3}$

Then, calculate Node node and the Beacon (x that is asked ₄, y ₄) range error:

${\mathrm{\&Delta;}}_1={|{({x^x}_1-x_4)}^2+{({y^y}_1-y_4)}^2-d_4^2|}^2$

By that analogy, four kinds of different combinations obtain f _Error(i), Δ _i:

f _error＝[f _error(1)，f _error(2)，f _error(3)，f _error(4)] ^T

Δ＝[Δ ₁，Δ ₂，Δ ₃，Δ ₄] ^T

At last, by 4 Beacon nodes point as a reference, utilize least square method to try to achieve Node node coordinate Node (x ₅, y ₅) and these 4 Beacon nodal distance error F _ErrorBe the Node (x that has tried to achieve from these 5 _i, y _i), i=1 ... select a Node node coordinate the most accurate in 5, a choice function following formula need be set, min (g (i)), i=1,2,3,4 are corresponding Node (x _i, y _i) for this reason the wheel preferred orientation result.

$g\left(i\right)=\frac{{\mathrm{\&Delta;}}_i}{f_{\mathrm{error}}\left(i\right)+1}i=\mathrm{1,2,3,4}$

Work as F _Error≤min (f _Error(i)), i=1,2,3,4 o'clock, 4 Beacon nodes no abnormal or more than one have fault, Node (x, y)=Node (x ₅, y ₅), putting 4 Beacon node flag is 1.

Work as F _Error＞min (f _Error(i)), i=1,2,3,4 o'clock, one of them Beacon node was very big to the position error influence, then gets rid of this f _Error(i) the Beacon node during minimum value, putting this Beacon node flag is 0, remaining 3 Beacon node flag is 2.

For each Beacon node, all have three different positioning results, i.e. flag={0,1,2}, the value of flag is as a parameter estimating the Beacon node confidence.

Step 5: according to every location Calculation result that takes turns, i.e. the flag value of Beacon node is upgraded GTR, the LTR degree of belief of Beacon node;

Described LTR degree of belief is:

$f_{\mathrm{trust}}\left[(n,\mathrm{Beacon})\right|\mathrm{Node}]=\left\{\begin{array}{cc}f\left(0\right)=1& \\ \max \{f(n-1)+\frac{1}{2\mathrm{LocationNum}},1\}& \mathrm{flag}=1\\ \max \{f(n-1)+\frac{1}{\mathrm{LocationNum}},1\}& \mathrm{flag}=2\\ \min \{f(n-1)-\frac{1}{\mathrm{LocationNum}},0\}& \mathrm{flag}=0\end{array}\right.$

In the following formula, n is that current algebraically is located in refinement, and LocationNum is a refinement location algebraically, and f (0)=1 is 1 for the LTR initial value;

Flag=0: participate in 4 Beacon reference nodes of location, this node is judged to suspect node, and LTR reduces by 0.1;

Flag=1: do not find out suspect node, for each reference node LTR increases by 0.05;

Flag=2: find out a suspect node, for other 3 safe node LTR increase by 0.1.

Described GTR is:

$F_{\mathrm{trust}}(n,\mathrm{Beacon})=\frac{\underset{i}{\overset{\mathrm{NbrNodeNum}}{\mathrm{\&Sigma;}}}{f}_{\mathrm{trust}}\left[(n,\mathrm{Beacon})\right|\mathrm{Node}]}{\mathrm{NbrNodeNum}}$

In the following formula, NbrNodeNum is the contiguous Node node number of Beacon node.

Step 6: continuous repeating step four, five, after N the location Calculation, obtain the N time position location of node to be measured, this position is exactly the final position information of node to be measured.

Present embodiment, N=10.

Position error mainly is the interference from unusual beaconing nodes except from measuring the noise in addition.The present invention's node to be measured gives the evaluation of beaconing nodes degree of belief simultaneously in many wheel refinement position fixing process.Unusual beaconing nodes degree of belief drops to below the threshold value, will carry on the back network and abandon, thereby positioning accuracy is improved.The global trusting degree is the evaluations of a plurality of Node nodes to the Beacon node.GTR descends fast and explains that there is fault in Beacon node self, in order not influence other node locating, only exits network.LTR descends fast, and GTR descends and lets alone that very bright Beacon node is unusual to certain Node node performance, and reason possibly be a channel disturbance between node, so this Beacon node is not participated in this Node node locating, but can not exit network.

Being merely the preferred embodiment of invention in sum, is not to be used for limiting practical range of the present invention.Be that all equivalences of doing according to the content of claim of the present invention change and modification, all should be technological category of the present invention.

Claims (7)

1. the localization method based on the wireless sensor network of degree of belief is characterized in that, may further comprise the steps:

Step 1: with the network grid sub-clustering, and a beaconing nodes in the picked at random bunch is as bunch head;

Step 2: in each sensor node distribute unique ID number, and the initial overall situation of Beacon node, local trust degree value and location times N are set;

Step 3: measure the RSSI value of Beacon node and Node node respectively, obtain the estimated distance of Node node and each Beacon node, set up neighbors information table separately;

Step 4: the location Calculation of all Node nodes in bunch head is responsible for bunch: when the number of the adjacent Beacon node of Node node to be measured during more than or equal to 4; According to nearest 4 the Beacon nodes of principles of selected of degree of belief ultimate range node as a reference; Use 4 location algorithm location, judge each Beacon node state; When the adjacent Beacon node of Node node to be measured has only 3, just use the least square method location; When the adjacent Beacon node of Node node to be measured number is less than 3, this Node node to be measured delocalization this time;

Step 5: take turns the degree of belief that the location Calculation result upgrades the Beacon node according to every;

Step 6: continuous repeating step four, five, after N the location Calculation, obtain the N time position location of Node node, this position is exactly the final position information of Node node.

2. the wireless sensor network locating method based on trust management according to claim 1 is characterized in that, the described cluster-dividing method of step 1 may further comprise the steps:

1) confirms bunch number compute cluster center, and distribute bunch number;

2) calculate the distance at each Beacon node and bunch center, promptly add this bunch near certain cluster center, the nearest Beacon node of each bunch distance center point then becomes a bunch head first;

3) the Node node calculates the average distance of Beacon node in each bunch, adds and nearest one bunch of its average distance.

3. the wireless sensor network locating method based on trust management according to claim 1 is characterized in that, the initial overall situation of the said beaconing nodes of step 2, local trust degree value all are 1, and the span of degree of belief is: [0,1].

4. the wireless sensor network locating method based on trust management according to claim 1 is characterized in that, the described estimated distance of step 3 is:

$d_j=d_0{e}^{\frac{P_j-P_0}{10n}},$

Wherein, d _jBe the distance of j beaconing nodes and node to be measured, P _jBe j beaconing nodes RSSI value, P ₀Be that distance is d ₀Two sensor nodes between the reception signal strength signal intensity, d ₀Be reference distance, 1≤j≤N, N are the total number of node to be measured.

5. the wireless sensor network locating method based on trust management according to claim 1 is characterized in that, the described node neighbor information of step 3 table is: neighbor node ID sign, and corresponding range information.

6. the wireless sensor network locating method based on trust management according to claim 1 is characterized in that, described 4 location algorithms of step 4 may further comprise the steps:

1) select four Beacon nodes as location reference point according to the rise of degree of belief from nearest principle;

2) combine three wherein different nodes of these four reference nodes and be divided into four different groups, these four groups are used least-squares calculation;

3) if a certain group of range error is all little than other groups, then judge to remove this organize in that beaconing nodes three nodes be abnormal nodes; Otherwise four reference nodes this time selecting are no abnormal; The result of calculation of this group is the position location.

7. the wireless sensor network locating method based on trust management according to claim 1 is characterized in that, the described renewal degree of belief of step 5 is: the degree of belief to unusual beaconing nodes is punished, promptly reduces degree of belief; Degree of belief to normal Beacon node is rewarded, i.e. degree of enhancing trust.

Images