KR101188840B1 - Location estimation system using direction angle in sensor network and method thereof - Google Patents

Abstract

PURPOSE: An intruder location estimating system and method are provided to precisely detect and trace an intruder by giving a feedback of target location information to a sensor node after estimating the location of the intruder based on a directional angle transmitted from the sensor node. CONSTITUTION: A plurality of sensor nodes(100) calculates vibration signal detecting time difference of other vibration sensor about a reference sensor. The plurality of sensor nodes estimates a directional angle for trespassing location of an intruder using the calculated vibration signal detecting time difference and propagation speed of a vibration signal. A central server estimates the location of the intruder using location information of each sensor node received from the each node through network, arrangement angle information, and the directional information of the intruder.

Description

Intruder position estimation system using direction angle in sensor network and its method {LOCATION ESTIMATION SYSTEM USING DIRECTION ANGLE IN SENSOR NETWORK AND METHOD THEREOF}

The present invention relates to an intruder position estimation system and method for enabling an intruder position estimation using a direction angle in a sensor network composed of nodes having a vibration sensor array.

In general, a system using a wireless sensor network performs an intruder monitoring boundary and position estimation using a passive sensor or a vibration sensor array.

By the way, in the case of using the passive sensor is mainly the detection of sensor node position-based detection or have a very large estimation error, in order to accurately estimate the position of the intruder, it is necessary to use an active sensor or complex system configuration.

In the case of using the vibration sensor array, position estimation is performed by using a time difference of arrival (TDoA) of each vibration sensor (TDoA method). In this case, a position estimation error is seriously increased. .

In addition, in the case of using the vibration sensor array, the propagation speed of the vibration signal which varies depending on time and place has a great influence on the position estimation error, and in order to estimate the position by measuring the propagation speed, at least five sensors for each sensor The array needs to be used.

Accordingly, an object of the present invention relates to an intruder position estimation system and method for precisely detecting an intruder position estimation using a direction angle in a sensor network composed of nodes having a vibration sensor array.

In order to achieve the above object, the intruder position estimation method in the intruder position estimation system according to an embodiment of the present invention, transmitting each of the sensor node, its own location information and placement angle information to the server; Detecting, at each sensor node, a vibration signal generated by an intruder's intrusion through the vibration sensor array; Calculating, at each sensor node, a difference in vibration signal detection time of another vibration sensor with respect to a reference sensor in the vibration sensor array; Estimating a direction angle with respect to the intrusion position of the intruder by using the calculated detection time difference and the propagation speed of the vibration signal in each sensor node and transmitting the same to the server; And estimating, at the server, the final position of the intruder using the direction angle information of the intruder received from each sensor node, the position information and the placement angle information of each sensor node.

The reference sensor is arbitrarily selected from among the vibration sensor arrays, or is selected as the vibration sensor in which the vibration signal arrives first.

The direction angle of the intruder is calculated by the maximum likelihood estimation method or the least square method.

The propagation speed of the vibration signal is estimated at each sensor node based on the intruder position information transmitted from the server.

The propagation speed of the vibration signal is estimated directly at each sensor node using the vibration signal detected by the vibration sensor.

The method for estimating an intruder may include feeding back the estimated intruder location information from the server to each sensor node; And estimating a propagation speed of the vibration signal by using the intruder position information fed back from each sensor node.

The propagation speed of the vibration signal is estimated using a moving average method.

In order to achieve the above object, the intruder position estimation system according to the embodiment of the present invention transmits the position information and the placement angle information to the network at the start of chip particle detection, and during the detection, each of the intruder detected through the vibration sensor array. A plurality of sensor nodes for estimating the direction angle information of the intruder using the vibration signal and transmitting the information to the network; And a central server estimating the final position of the intruder using the position information, the placement angle information, and the direction angle information of the intruder received from each sensor node through the network.

The position information and the placement angle information are measured by a position measuring device and an electronic compass provided in each sensor node.

Each sensor node selects a reference sensor from among the vibration sensor arrays, calculates a difference in detection time of the vibration signal of the other vibration sensor with respect to the reference sensor, and uses the calculated detection time difference and the propagation speed of the vibration signal. Estimate the direction angle for.

Each sensor node estimates the propagation speed of the vibration signal using the intruder position information provided from the server, or estimates the propagation speed of the vibration signal using the vibration signal in the vibration sensor array.

The propagation speed of the vibration signal is estimated using a moving average method.

Each sensor node and server calculates the direction angle information of the intruder using the maximum likelihood estimation method or the least square method.

The server feeds back the estimated intruder position information to each sensor node, and each sensor node estimates the propagation speed of the vibration signal using the fed back intruder position information.

The present invention obtains a direction angle from the sensor node and transmits it to the server using at least three vibration sensor arrays for each sensor node, and the server estimates the position of the intruder based on the direction angle transmitted from each sensor node, and then the corresponding position. By feeding back information to each sensor node, you can detect and track intruders with greater precision.

Therefore, when the present invention is used in the field of defense, it can be applied to the surveillance reconnaissance sensor network to track the position of the intruder, and there is an effect that can be operated at low power using a vibration sensor which is a passive sensor. In addition, when the present invention is applied to the civilian field, it is possible to construct a system capable of detecting and tracking an intruder relatively precisely using a cheap vibration sensor and a low power node.

1 is a block diagram of each sensor node according to an embodiment of the present disclosure.
2 is a diagram illustrating an example of tracking an intruder in a sensor network composed of a plurality of sensor nodes.
Figure 3 is a flow chart showing the overall operation of performing the location of the intruder in the intruder location estimation system.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The present invention obtains a direction angle from the sensor node and transmits it to the server using at least three vibration sensor arrays for each sensor node, and the server estimates the position of the intruder based on the direction angle transmitted from each sensor node, and then the corresponding position. By feeding back information to each sensor node, you can precisely detect and track intruders. In this case, if two or more sensor nodes detect the same intruder, it is possible to estimate the position of the intruder relatively accurately.By feeding back the estimated intruder position to each sensor node, each sensor has three or four vibration sensor arrays. The propagation speed of the vibration signal can be estimated.

1 is a configuration diagram of a sensor node having a vibration sensor array.

As shown in FIG. 1, the sensor node 100 having a vibration sensor array includes three vibration sensors 10a to 10c for detecting vibrations by chip particles, and vibrations detected by the vibration sensors 10a to 10c. Communication and signal module 20 for estimating the direction angle to the ground chip chip particles, and the antenna 30 for transmitting the direction angle of the chip particles estimated by the communication and signal module 20 to the server (not shown) ). In this case, three vibration sensors 10a to 10c form an array of vibration sensors, and the antenna 30 may be built in.

2 illustrates an example of performing an intruder location tracking in a sensor network composed of a plurality of sensor nodes, and FIG. 3 is a flowchart illustrating an overall operation of performing an intruder location tracking in an intruder location estimation system.

As shown in Figure 2, each sensor node (a ~ d) is equipped with a device (for example, GPS) that can measure its own position, when installing each sensor node (a ~ d) vibration sensor The fields 10a to 10c should be installed in a predetermined shape and size.

In addition, each of the sensor nodes (a to d) is installed to be arranged in the same reference direction (for example, north), or is provided with an electronic compass to know the reference direction and the angle of placement. Each sensor node (a ~ d) is connected to the network when the installation is completed to form a sensor network.

Once connected to the network, as shown in FIGS. 3 and 3, each sensor node a to d is operated to measure its own position through GPS and an arrangement angle of the sensor node with respect to the reference direction through an electronic compass. After that, the measured position and placement angle are transmitted to the central server (hereinafter abbreviated as server) and detection is started (S10 to S12).

After the intrusion occurs, the vibration signal by the intruder is detected by the vibration sensor (10a ~ 10c) (N can be installed) of each sensor node (a ~ d) (S13). At this time, since the propagation speed of the detected vibration signals 1 to 3 is limited by the medium, even the same vibration signal is detected with a time difference between the vibration sensors 10a to 10c.

Each sensor node (a to d) is a vibration sensor (10a ~ 10c), that is, randomly selected one vibration sensor in the vibration sensor array (for example, vibration sensor 10a) set as a reference sensor and then the other A time difference of arrival (TDoA) of the vibration sensor is calculated by using a cross correlation method (S14). In this case, as an embodiment, the vibration sensor that first receives the vibration signal may be selected as the reference sensor. Therefore, when N vibration sensors are used, N-1 independent TDoA measurements can be obtained.

, 센서어레이(N개) 중 기준센서(0번 센서)의 위치를

, 센서어레이 중 기준센서를 제외한 임의의 진동센서(i번째 센서, i=1, 2,..,N-1)의 위치를

, 진동신호의 전파속도를 v 라고 하자.
최소제곱(Least square) 방법
상기 단계(S14)에서 i번째 진동센서와 기준센서의 도착시간차이(TDoA)를 구하면 식(1)과 같다.

..................식(1)

......식(2)
상기 식(2)에서 좌변의 두번째 항을 이용하여 양변을 제곱하여 정리하면 다음 식(3)과 같다.

...식(3)
상기 식(3)을 각 진동센서(i=1, 2,..,N-1)에 적용하면 식(4)와 같이 된다.

...식(4)
상기 식(4)에서 침입자의 위치(X)와 방향각(θ)을 구하면 식(5) 및 (6)과 와 같다.

..................식(5)

.............식(6)
상기 식(5)를 통해 추정한 침입자의 위치는 진동센서의 오차를 감안하면 실제 위치와 굉장히 큰 오차를 갖지만, 상기 식(6)을 통해 추정한 방향각(θ)은 비교적 매우 정확하다.
최대우 추정(Maximum Likelihood estimation)방법
상기 주어진 침입자의 위치, 기준센서의 위치, i번째 센서의 위치 그리고 진동신호의 전파속도에 대해 신호 도착시간 차이(TDoA)에 대한 확률은 일반적으로 아래 식(7)과 같이 나타낼 수 있다.

..........식(7)
상기 식(7)을 이용하여 침입자의 위치에 대한 최대우(Likelihood)를 구하면 다음 식(8)과 같다.

................식(8)
상기 식(7)에서 각 다른 i에 대해 TDoA에 대한 확률이 서로 독립이고 동일한 분포(I.I.D)를 갖는다면 N-1개의 측정된 TDoA에 대한 침입자의 위치(Xs)의 log Likelihood는 식 (9)와 같이 나타낼 수 있다.

........식(9)
상기 최대우(Maximum Likelihood)를 갖는 침입자의 위치(Xs)를 찾기 위하여 다음 식(10)을 만족하는 Xs를 찾으면 된다.

..............식(10)
이와같이 최대우 추정방법(MLE)을 통해 Xs를 구하면 식(6)을 이용하여 방향각(θ)을 구할 수 있게 된다. 상기 방법에 의해 추정된 방향각에 의한 방향선(a~c)는 도 2에 도시되어 있다. The communication and signal module of each sensor node (a ~ d) using the time difference of arrival (TDoA) and the propagation speed of the vibration signal of the two measured (N-1 when using N vibration sensors) Is estimated by the least square method or the maximum likelihood estimation method (S15). Hereinafter, a direction angle of an intruder is obtained by using a least square method or a maximum likelihood estimation method.
First, find out the location of the intruder.

, The position of reference sensor (sensor 0) among N sensor arrays

, Position of any vibration sensor (i-th sensor, i = 1, 2, .. N-1) except reference sensor among sensor array

Let v be the propagation speed of the vibration signal.
Least square method
In step S14, when the arrival time difference (TDoA) of the i-th vibration sensor and the reference sensor is obtained, Equation (1) is obtained.

Equation (1)

Expression (2)
In both equations (2), using both terms on the left side, the squares of both sides are summarized as in the following equation (3).

... (3)
Applying equation (3) to each vibration sensor (i = 1, 2, ..., N-1) is the same as equation (4).

... formula (4)
The position X of the intruder and the direction angle θ in Equation (4) are obtained as in Equations (5) and (6).

Equation (5)

............. Equation (6)
The position of the intruder estimated by Equation (5) has a very large error with the actual position in consideration of the error of the vibration sensor, but the direction angle θ estimated by Equation (6) is relatively very accurate.
Maximum Likelihood Estimation
With respect to the position of the intruder, the position of the reference sensor, the position of the i-th sensor, and the propagation speed of the vibration signal, the probability of the difference in signal arrival time (TDoA) can be expressed as Equation (7) below.

.......... Equation (7)
Using the equation (7) to obtain the maximum right (Likelihood) for the position of the intruder is as shown in the following equation (8).

Equation (8)
If the probability for TDoA for each other i in the equation (7) is independent of each other and has the same distribution (IID), then the log likelihood of the intruder's position (Xs) for N-1 measured TDoAs is given by equation (9). Can be expressed as:

Expression (9)
In order to find the position Xs of the intruder having the maximum likelihood, the Xs satisfying the following equation (10) can be found.

Equation (10)
As described above, when Xs is obtained through the maximum right estimation method MLE, the direction angle θ can be obtained using Equation (6). Direction lines a to c by the direction angle estimated by the method are shown in FIG. 2.

The propagation speed of the vibration signal is a speed estimated by using the intruder's location information transmitted from the server, and when five or more vibration sensors (N = 5) are used, the propagation of the vibration signal is also performed at each sensor node (a to d). Speed estimation is possible.

, 침입자의 방향각을

, 침입자의 위치를

라고 하면, 다음 식(11)과 같이 나타낼 수 잇다.

.........식(11)
다른 센서노드에서 방향각이 수신되어 식(11)이 두개이상 구해지면, 다음의 식(12)와 같이 행렬식으로 나타낼 수 있다.

.........식(12)
따라서, 상기 식(12)를 정리하면 다음의 식(13)과 같이 각 센서노드(a~d)의 위치정보, 배치각 및 침입자에 대한 방향각 정보를 이용하여 침입자의 위치를 추정할 수 있게 된다. The communication and signal module 20 of each sensor node (a ~ d) transmits the estimated direction angle to the server (not shown) through the antenna 30 (S16), the server receives each sensor node (a ~) The position of the intruder is estimated using the position information, the placement angle, and the direction angle information of the intruder in step S17 and S18. As the estimation method, the stochastic method (MLE) or the geometric method (Least square) is used.
Estimate Intruder Location on Server
The server that receives the intruder's direction angle from each sensor node can estimate the intruder's location using the following Least Square method, which is a known technique.
First, the position of the kth (k = 1, 2, .., K) sensor node among the sensor nodes that detected one intruder

Intruder's direction angle

Intruder's position

Can be expressed as in the following equation (11).

......... (11)
If a direction angle is received from another sensor node and two or more equations (11) are obtained, they may be expressed in a matrix form as shown in Equation (12) below.

......... Equation (12)
Therefore, the equation (12) can be summarized so that the position of the intruder can be estimated using the position information, the placement angle, and the direction angle information of the intruder as shown in the following equation (13). do.

When each sensor node (a ~ d) uses three or four vibration sensors, the server sends the calculated intruder's location information back to the sensor nodes (a ~ d) that detected the intruder, and then estimates the propagation speed of the vibration signal. To be used (S19).

On the other hand, in order to estimate the direction angle of the intruder in each sensor node (a ~ d), it is necessary to propagate the vibration signal. However, in order to calculate the propagation speed of the vibration signal through the actual measurement together with the direction angle with respect to the chip particles, five or more (N ≥ 5) vibration sensor arrays should be used. However, the propagation speed of the vibration signal is not a constant, and may vary greatly depending on a state of a medium (for example, soil), that is, a place and a time. In addition, if the propagation speed is misunderstood, a large error is shown in estimating the intruder's direction angle.

Therefore, in the present invention, the position information of the intruder finally estimated by the server is transmitted to each sensor node (a ~ d), each sensor node (a ~ d) receives the position information of the intruder from the server to perform a moving average method By estimating the propagation speed of the vibration signal using the three or four vibration sensor arrays, the propagation speed of the vibration signal is estimated (S20 and S21). That is, each sensor node (a ~ d) can calculate the direction angle by using a random propagation velocity initial value before correction through the information of the server, it is possible to set the value set through the experiment as the initial value. When the exact position information of the intruder is fed back from the server, each sensor node (a ~ d) can correct the speed of the medium by calculating the propagation speed according to the type and state of the installed medium through the above formula (1). This can increase the accuracy of the estimation of the direction angle when the next intruder occurs.

Each of the sensor nodes a to d displays the location information of the intruder transmitted from the server on a screen of a display unit (not shown) so that the administrator can know. This series of steps, in real time, allows the intruder to be tracked in real time.

As described above, the present invention can obtain a direction angle from the sensor node using three or more vibration sensor arrays for each sensor node, and when two or more sensor nodes detect the same intruder, it is possible to accurately estimate the position of the intruder. By feeding back the estimated intruder position to each sensor node, it is possible to estimate the propagation speed of the vibration signal for each sensor using only three or four vibration sensor arrays.

Therefore, when the present invention is used in the field of defense, it can be applied to the surveillance reconnaissance sensor network to track the position of the intruder, and there is an effect that can be operated at low power using a vibration sensor which is a passive sensor. In addition, when the present invention is applied to the civilian field, it is possible to construct a system capable of detecting and tracking an intruder relatively precisely using a cheap vibration sensor and a low power node.

The intruder location estimation method of the sensor network described above is not limited to the configuration and method of the embodiments described above, the embodiments are all or part of each of the embodiments is optional so that various modifications can be made It may be configured in combination.

10a ~ 10c: vibration sensor 20: communication and signal module
30: antenna 100: sensor node

Claims (14)

Transmitting, at each sensor node, its own location information and placement angle information to a server;
Detecting, at each sensor node, a vibration signal generated by an intruder's intrusion through the vibration sensor array;
Calculating, at each sensor node, a detection time difference between vibration signals of another vibration sensor with respect to the reference sensor in the vibration sensor array;
At each sensor node, estimating a direction angle with respect to the intrusion position of the intruder by using the calculated detection time difference and the propagation speed of the vibration signal and transmitting the same to the server;
Estimating the position of the intruder using the direction angle information of the intruder received from each sensor node, the position information and the placement angle information of each sensor node;
At the server, feeding back the estimated intruder location information to each sensor node; And
Estimating the propagation speed of the vibration signal according to the medium using the intruder position information fed back from each sensor node to estimate a direction angle with respect to the intrusion position of the intruder. The method of claim 1, wherein the reference sensor
Intruder position estimation method characterized in that the randomly selected from the vibration sensor array or the vibration signal is selected as the first arrival vibration sensor. The method of claim 1, wherein the inclination angle of the intruder
Intruder position estimation method characterized in that the calculation by the maximum likelihood estimation method or least square method. delete delete delete According to claim 1, wherein the propagation speed of the vibration signal is
Intruder location estimation method estimated using moving average method. At the start of chip particle detection, location information and placement angle information are transmitted to the network, and during detection, the reference sensor is selected from the vibration sensor array, and the difference of vibration signal detection time of other vibration sensors with respect to the reference sensor is calculated. A plurality of sensor nodes for estimating a direction angle with respect to the intrusion position of the intruder using the difference and the propagation speed of the vibration signal and transmitting the estimated angle to the network; And
And a central server estimating the location of the intruder using the location information, the placement angle information, and the direction angle information of the intruder received from each sensor node through the network.
The server feeds back the estimated intruder position information to each sensor node, and each sensor node uses the feedback intruder position information to estimate the propagation speed of a vibration signal according to a medium and use the calculated intruder. Positioning system. The method of claim 8, wherein the location information and the placement angle information
Intruder position estimation system, characterized in that measured by the position measuring device and the electronic compass provided in each sensor node. delete delete The method of claim 8, wherein each sensor node
An intruder position estimation system comprising estimating the propagation speed of a vibration signal using a moving average method. The method of claim 8, wherein each sensor node and server
An intruder position estimation system, comprising calculating the direction angle information of the intruder and the position of the intruder using a maximum likelihood estimation method or a least square method.
delete

Images