CN104159295A - Node positioning method based on filtering algorithm in wireless sensor network - Google Patents

Abstract

The invention discloses a node positioning method in a wireless sensor network. The node positioning method comprises the following steps: establishing a rectangular coordinate system to acquire position coordinates of each anchor node in the wireless sensor network; establishing a corresponding anchor node information list for the node to be positioned through a distance vector exchange agreement; inquiring a self anchor node information list by the node to be positioned; when the number of the anchor nodes is more than or equal to 3 in a 1 hop range, carrying out RSSI (Received Signal Strength Indicator) ranging by using the node to be positioned to acquire a ranging sample between the node to be positioned and each anchor node; combining an anti-pulse interference average valve filtering with voting average federated filtering algorithm, performing correction processing on the ranging sample to acquire accurate ranging information between the node to be positioned and the anchor node, and utilizing a least square method to determine the final position of the node to be positioned. The node positioning method can realize accurate positioning of the node to be positioned.

Description

Node positioning method based on filtering algorithm in a kind of wireless sensor network

Technical field

The invention belongs to wireless sensor network technology field, a kind of WSN of the range finding model based on filtering algorithm (Wireless Sensor Nerwork, wireless sensor network) node positioning method particularly,

Background technology

Wireless sensor network forms by being deployed in microsensor nodes a large amount of in monitored area, the network system of the self-organizing of the multi-hop forming by communication, its objective is perception collaboratively, perceived object information in acquisition and processing network's coverage area, and send to observer.Wireless sensor network set sensor technology, MEMS (micro electro mechanical system) (MEMs) technology, embedded computing technique, wireless communication technology and distributed information processing are in one, mutual with the external world by transducer, data acquisition, process the functions such as communication and management.

In many cases, the node in wireless sensor network need to be known the physical location of self.For great majority application, do not know sensing station and the data of perception are nonsensical.Yet in wireless sensor network, for the manual setting position of each node or for its configuration a GPS receiver be impossible.The location of wireless sensor network is mainly divided into two classes: a class is exactly to detecting target localization, and another kind of is location to sensor node itself, and network node is self-align.In the situation that node location information is not all known, the self-align algorithm of node is obviously the prerequisite of target localization algorithm.Therefore,, in majority application, the precision of sensor localization is most important.

Obtain the purposes that sensor node position has following several respects at least: first, the data that node collects must be combined the position in measuring coordinate system, do not have the data of positional information almost there is no value; Secondly, some systemic-functions of sensor network, such as network topology control, the route based on geography information etc., need positional information, the scheduling mechanism on duty that known location can optimized network run duration is in addition had holidays by turns aperiodically with life-saving redundant node in network; Finally, positional information is extremely important to the service application in sensor network, and what is more important, along with the continuous progress of sensor network technique, there will be agreement and the application of more position-based information very naturally.Just for these reasons, the location technology of sensor network be the normal operation of network be substantially the most also most important condition.Under the environment of thunderstorm weather (or other have impulse disturbances), be arranged on the variation of sensor node (or other) the monitoring water level on korneforos and bank, these sensor nodes are easy to be subject to the impact of impulse disturbances very much and cause the inaccurate of location when location, and node positioning method in a kind of sensor network is provided for this reason.

Summary of the invention

For above deficiency of the prior art, the object of the present invention is to provide a kind of high accuracy, low complex degree, the node positioning method based on filtering algorithm in the wireless sensor network of more low-yield expense.Technical scheme of the present invention is as follows: the node positioning method based on filtering algorithm in a kind of wireless sensor network, and it comprises the following steps:

101, at random in the S of region random set n sensor network nodes/, described sensor network nodes comprises p anchor node and (n-p) individual node to be positioned;

102, under the S of region, set up rectangular coordinate system, obtain the positional information of anchor node in wireless sensor network, wherein, anchor node is the known node of self-position in wireless sensor network, node to be positioned obtains the ID of anchor node in network by distance vector exchange agreement, position coordinates and corresponding jumping figure, node to be positioned is set up anchor node information table according to the anchor node information corresponding with it;

103, the anchor node information table corresponding with himself of setting up in querying node step 102 to be positioned, when the anchor node number having within the scope of its 1 jumping is greater than or equal to 3, this node to be positioned starts the range finding of RSSI received signal strength indicator, obtains N range finding sample between node to be positioned and each anchor node;

104, adopt anti-impulse disturbances average value filtering algorithm to eliminate sampled value deviation processing to N range finding sample, eliminate the sampled value deviation causing due to impulse disturbances;

105, adopt ballot filtering algorithm to process remaining range finding sample after step 104 is eliminated sampled value deviation processing, reject the abnormal range finding of part sample;

106, the M finally remaining a range finding sample averaged to filtering and obtain distance measurement value;

107, adopt maximal possibility estimation algorithm obtain the position coordinates of node to be positioned/, complete the node locating of node to be positioned.

Further, the RSSI received signal strength indicator range finding in step 103 is chosen Shadowing model as signal mode, and this Shadowing model formation is:

${\left[\frac{p_r\left(d\right)}{p_r\left(d_0\right)}\right]}_{\mathrm{dB}}=-10\mathrm{\β}\log \left(\frac{d}{d_0}\right)+X_{\mathrm{dB}}$

Here the each transmitted power of anchor node is identical, p _r(d) expression is apart from the received power of the node to be measured of the position of anchor node d, p _r(d ₀) represent apart from anchor node d ₀the node received power of position, X _dBrepresent a Gaussian random variable that average is 0, d ₀represent a reference distance, generally get 1m, d represents that node to be measured is to the distance of anchor node, and β is the path loss factor;

Further, the anti-impulse disturbances average value filtering algorithm in step 104 comprises step: A1, by the N obtaining a range finding sample { d _j(k), 1≤k≤N} carries out sequence from small to large, removes maximum and minimum value wherein, and remaining N-2 range finding sample is designated as to { x _j(i), 1≤i≤N-2}; A/2, then ask for the sample average μ of remaining N-2 the sample of finding range:

$\mathrm{\μ}=\frac{x_j\left(1\right)+x_j\left(2\right)+...+x_j\left(i\right)}{N-2}$

Further, the ballot filtering algorithm concrete steps in step 105 are as follows:

B1, range finding sample remaining after eliminating sampled value deviation processing in step 104 is asked for to standard deviation sigma:

${\mathrm{\σ}}^2=\frac{1}{N-3}\underset{i=1}{\overset{N-2}{\mathrm{\Σ}}}\left[x_j\left(i\right)\right]$

B2, adopt gaussian probability distribution function Voting Model, calculate the ballot probability P (x of each range finding sample _j(k) | μ):

$P\left(x_j\left(k\right)\right|\mathrm{\μ})=\frac{1}{\sqrt{2\mathrm{\π}}*\mathrm{\σ}}{e}^{-{(x_j\left(k\right)-\mathrm{\μ})}^2/2{\mathrm{\σ}}^2}$

B3, threshold k of setting (numerical value 0.5≤K≤0.8 for example), vote to range finding sample, as the ballot probability P (x of range finding sample _j(k) | while μ) being less than threshold k, just abandon this range finding sample, otherwise just retain, the range finding sample number retaining represents with M.

Advantage of the present invention and beneficial effect are as follows:

Anti-impulse disturbances filtering of the present invention and ballot filtering algorithm combine and can reduce the impact of surrounding environment on RSSI range finding, by eliminating abnormal sampled value, realize the precision distance measurement between adjacent node, realized a kind of high accuracy, low complex degree, the WSN node positioning method of more low-yield expense.

Accompanying drawing explanation

Figure 1 shows that preferred embodiment of the present invention localization method schematic flow sheet;

Fig. 2, wireless sensor network structural system;

Fig. 3, RSSI value and euclidean distance between node pair be related to schematic diagram.

Embodiment

The invention will be further elaborated below in conjunction with accompanying drawing, to provide an infinite embodiment.But should be appreciated that, these describe example just, and do not really want to limit the scope of the invention.In addition, in the following description, omitted the description to known configurations and technology, to avoid unnecessarily obscuring concept of the present invention.

The invention discloses node positioning method in a kind of wireless sensor network, comprised following step: set up rectangular coordinate system, obtain the positional information of each anchor node in wireless sensor network; And by distance vector exchange agreement, make the ID of anchor node in node acquisition network to be positioned, position coordinates and corresponding jumping figure, and set up corresponding anchor node information table; Anchor node is the known node of self-position in wireless sensor network; The anchor node information table of the querying node to be positioned self in network, when the anchor node number having within the scope of its 1 jumping is greater than or equal to 3, this node to be positioned starts range finding, obtains the range finding sample between node to be positioned and each anchor node; Design anti-tampering average value filtering algorithm and ballot filtering algorithm is processed range finding sample, and set up RSSI range finding model and obtain the precision ranging information between node to be positioned and anchor node.

Node to be positioned is found range, and the range finding sample obtaining between node to be positioned and each anchor node comprises: the distance of utilizing RSSI technology to measure between node to be positioned self and anchor node for N time, obtains N range finding sample, with { d _j(k), 1≤k≤N} represents, wherein d _j(k) represent the k time range finding sample between this node to be positioned and anchor node j.

The anti-impulse disturbances average filter of described use algorithm is processed and is comprised the steps: range finding sample

First N range finding sample compared, remove maximum and minimum value wherein, remaining N-2 range finding sample is designated as to { x _j(i), 1≤i≤N-2}, calculates the sample average μ of a remaining N-2 data:

$\mathrm{\μ}=\frac{x_j\left(1\right)+x_j\left(2\right)+...+x_j\left(i\right)}{N-2}$

Calculate standard deviation sigma:

${\mathrm{\σ}}^2=\frac{1}{N-3}\underset{i=1}{\overset{N-2}{\mathrm{\Σ}}}\left[x_j\left(i\right)\right]$

Described use ballot filtering algorithm is processed and is comprised the steps: range finding sample

Voting Model adopts gaussian probability distribution function, calculates the ballot probability P (x of each range finding sample by following formula _j(i) | μ):

$P\left(x_j\left(i\right)\right|\mathrm{\μ})=\frac{1}{\sqrt{2\mathrm{\π}}*\mathrm{\σ}}{e}^{-{(x_j\left(i\right)-\mathrm{\μ})}^2/2{\mathrm{\σ}}^2}$

(the present invention advises that threshold k span is 0.5≤K≤0.8 to setting threshold K, and the value of K is larger, is more conducive to suppressing exception range finding sample, if but value is excessive, will delete normal range finding sample), range finding sample is voted, as the ballot probability P (x of range finding sample _j(i) | while μ) being less than threshold k, just abandon this range finding sample, otherwise just retain, the range finding sample number retaining represents with M.

Then the M retaining a range finding sample averaged to filtering according to the following formula, obtain distance measurement value d _j:

$d_j=\frac{1}{M}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{x}_j\left(i\right)$

Wherein: d _jrepresent the distance recording between location node to be asked for and anchor node j.

Node to be positioned obtain successively and its 1 jumping within the scope of range information between anchor node, adopt maximal possibility estimation algorithm to obtain the position coordinates of node to be positioned.The detailed process of maximal possibility estimation is as follows:

If the coordinate of a known n anchor node is respectively (x ₁, y ₁), (x ₂, y ₂) ..., (x _n, y _n), the distance of they and node M to be positioned is respectively d ₁, d ₂..., d _n.The coordinate of supposing node M is (x, y), has following relation:

$\left\{\begin{array}{c}{(x_1-x)}^2+{(y_1-y)}^2=d_1^2\\ {(x_2-x)}^2+{(y_2-y)}^2=d_2^2\\ ...\\ {(x_n-x)}^2+{(y_n-y)}^2=d_n^2\end{array}\right.$

Above formula system of linear equations can be expressed as:

AX＝d,

Wherein each several part is respectively:

$A=2\left\{\begin{array}{cc}{x}_1-x_n& y_1-y_n\\ ...& ...\\ x_{n-1}-x_n& y_{n-1}-y_n\end{array}\right\},$

$d=\left\{\begin{array}{c}{x}_1^2-x_n^2+y_1^2-y_n^2+d_n^2-d_1^2\\ x_{n-1}^2-x_n^2+y_{n-1}^2-y_n^2+d_n^2-d_{n-1}^2\end{array}\right\}$

$X=\left(\begin{array}{c}x\\ y\end{array}\right)$

The coordinate that finally adopts the Minimum Mean Square Error estimation technique can calculate node M to be positioned is: $\hat{X}={\left(A^{T}A\right)}^{-1}{A}^{T}d.$

Fig. 2 is the structural system of wireless sensor network, and the architecture of a typical wireless sensor network comprises distributed wireless sensor node (group), receiver transmitter aggregation node, the Internet or communication satellite and task management node etc.

According to wireless signal attenuation model, simulate the relation of RSSI value and distance, as Fig. 3, from figure, can obviously find out, along with the increase of distance between node, RSSI value reduces gradually.

These embodiment are interpreted as only for the present invention is described, is not used in and limits the scope of the invention above.After having read the content of record of the present invention, technical staff can make various changes or modifications the present invention, and these equivalences change and modification falls into the inventive method claim limited range equally.

Claims (4)

1. the node positioning method based on filtering algorithm in wireless sensor network, is characterized in that comprising the following steps:

101, at random in the S of region random set n sensor network nodes/, described sensor network nodes comprises p anchor node and (n-p) individual node to be positioned;

102, under the S of region, set up rectangular coordinate system, obtain the positional information of anchor node in wireless sensor network, wherein, anchor node is the known node of self-position in wireless sensor network, node to be positioned obtains the ID of anchor node in network by distance vector exchange agreement, position coordinates and corresponding jumping figure, node to be positioned is set up anchor node information table according to the anchor node information corresponding with it;

103, the anchor node information table corresponding with himself of setting up in querying node step 102 to be positioned, when the anchor node number having within the scope of its 1 jumping is greater than or equal to 3, this node to be positioned starts the range finding of RSSI received signal strength indicator, obtains N range finding sample between node to be positioned and each anchor node;

104, adopt anti-impulse disturbances average value filtering algorithm to eliminate sampled value deviation processing to N range finding sample, eliminate the sampled value deviation causing due to impulse disturbances;

105, adopt ballot filtering algorithm to process remaining range finding sample after step 104 is eliminated sampled value deviation processing, reject the abnormal range finding of part sample;

106, the M finally remaining a range finding sample averaged to filtering and obtain distance measurement value;

107, adopt maximal possibility estimation algorithm obtain the position coordinates of node to be positioned/, complete the node locating of node to be positioned.

2. the node positioning method based on filtering algorithm in wireless sensor network according to claim 1, it is characterized in that: the RSSI received signal strength indicator range finding in step 103 is chosen Shadowing model as signal mode, and this Shadowing model formation is:

Here the each transmitted power of anchor node is identical, p _r(d) expression is apart from the received power of the node to be measured of the position of anchor node d, p _r(d ₀) represent apart from anchor node d ₀the node received power of position, X _dBrepresent a Gaussian random variable that average is 0, d ₀represent a reference distance, generally get 1m, d represents that node to be measured is to the distance of anchor node, and β is the path loss factor.

3. the node positioning method based on filtering algorithm in wireless sensor network according to claim 1, is characterized in that: the anti-impulse disturbances average value filtering algorithm in step 104 comprises step: A1, by the N obtaining a range finding sample { d _j(k), 1≤k≤N} carries out sequence from small to large, removes maximum and minimum value wherein, and remaining N-2 range finding sample is designated as to { x _j(i), 1≤i≤N-2}; A/2, then ask for the sample average μ of remaining N-2 the sample of finding range:

4. the node positioning method based on filtering algorithm in wireless sensor network according to claim 1, is characterized in that: the ballot filtering algorithm concrete steps in step 105 are as follows:

B1, range finding sample remaining after eliminating sampled value deviation processing in step 104 is asked for to standard deviation sigma:

B2, adopt gaussian probability distribution function Voting Model, calculate the ballot probability P (x of each range finding sample _j(k) | μ):

B3, a threshold k of setting, 0.5≤K≤0.8, votes to range finding sample, as the ballot probability P (x of range finding sample _j(k) | while μ) being less than threshold k, just abandon this range finding sample, otherwise just retain, the range finding sample number retaining represents with M.