KR20230042801A - Apparatus and method for estimating location - Google Patents

Abstract

These embodiments relate to a positioning device and method. In particular, it is possible to provide the positioning device and method for estimating the location and status of a terminal by identifying the environment around the terminal through a clustering technique in a wireless communication environment. Specifically, according to the provided positioning device and method, in a wireless network consisting of the terminal and a base station, a received signal is used to estimate the signal generation location and distinguish objects through clustering-based environment mapping, so as to estimate the location of the terminal.

Description

Positioning device and method {APPARATUS AND METHOD FOR ESTIMATING LOCATION}

The present embodiments relate to a positioning device and method.

Positioning technology is a positioning technology that grasps the location, speed, and path of a moving object. It does not stop at simply estimating the user's location, but is a communication technology using location information, autonomous driving technology, and spatial information construction technology based on location information. It is an important geotechnical technology that can be applied to countless future technologies such as Specifically, positioning technologies include positioning using GPS, positioning using mobile communications such as 4G LTE and 5G NR, and positioning technologies using short-range wireless communication systems such as Wi-Fi and ultra-wide band (UWB). In particular, millimeter wave (mmWave) positioning technology can enable positioning with high accuracy in a place where mobile communication is possible without restrictions of separate equipment and environment. With the development of such positioning technology, interest in autonomous driving technology is increasing in the field of vehicles such as automobiles, drones, and automatic guided vehicles (AGVs).

Accordingly, interest in Simultaneous Localization And Mapping (SLAM), which is an essential technology in autonomous driving technology, is recently increasing. SLAM (Simultaneous Localization And Mapping) is a technology that creates a map of the surrounding environment of a moving object while measuring the location of the moving object, and is emerging as a key technology for autonomous driving. Specifically, SLAM technology goes beyond the existing map-based location recognition or location-aware map-making technology to complement each other by simultaneously locating and building a map. It is very similar to exploring the unknown environment we do in our daily lives. may be in a similar way. In particular, since high-resolution measurement values can be obtained in the time and angle domains through a wireless network, it is possible to detect scattering objects and estimate states such as the user's position and direction by applying the characteristics of the wireless signal to the SLAM.

However, SLAMs using wireless signals have problems such as miss detection of an object due to a defect in a receiver, false alarms due to channel estimation errors, and unknown types of landmarks in the environment. Therefore, in a situation where radio signals of visible path and multi-path caused by multiple landmarks in the vehicle environment are mixed, there is a need for a positioning device and method for identifying the type by distinguishing the object from which each signal is reflected, and at the same time increasing location accuracy, there is.

Against this background, the present embodiments provide a positioning device and method for estimating the location of a terminal in real time by estimating the generation position of a signal using a received signal through clustering-based environment mapping and classifying objects.

In one aspect, the present embodiments receive a positioning reference signal (PRS) from a base station, and obtain a measurement value of the received positioning reference signal through channel estimation, a measurement value acquisition unit, a preset movement model, and a previous A state prediction unit for predicting a second state vector of the terminal using a first state vector of the terminal obtained at time, and a location reference signal existing on a reception path based on the measurement value of the position reference signal and the second state vector Using the environment mapping unit that estimates the creation positions of virtual anchors (VA) and scattering points (SP), clusters the creation positions, and maps them to the network environment map of the terminal, and the mapping result of the network environment map It is possible to provide a positioning device including a state correction unit that corrects the second state vector.

In another aspect, the present embodiments include a measurement value acquisition step of receiving a positioning reference signal (PRS) from a base station and obtaining a measurement value of the received positioning reference signal through channel estimation, a preset movement model, and a previous State prediction step of predicting the second state vector of the terminal using the first state vector of the terminal obtained at time, based on the measurement value of the position reference signal and the second state vector, which is present on the reception path of the position reference signal An environment mapping step of estimating the creation positions of virtual anchors (VA) and scattering points (SPs), clustering the creation positions, and mapping them to the network environment map of the terminal, and using the mapping result of the network environment map It is possible to provide a positioning method including a state correction step of correcting the second state vector.

According to the present embodiments, it is possible to provide a positioning device and method for estimating the position of a terminal in real time by estimating the generation position of a signal using a received signal through clustering-based environment mapping and classifying objects.

1 is a diagram showing the configuration of a positioning device according to an embodiment of the present disclosure.
2 is a diagram schematically illustrating a system to which a positioning device according to an embodiment of the present disclosure may be applied.
3 is a flowchart illustrating an operation of estimating a location of a terminal by a positioning device according to an embodiment of the present disclosure.
4 is a diagram for explaining an operation of estimating a creation position of a virtual anchor by a positioning device according to an embodiment of the present disclosure.
5 is a diagram for explaining an operation of estimating the generation position of a scattering point by the positioning device according to an embodiment of the present disclosure.
6 is a diagram for explaining an algorithm for mapping an environment map by a positioning device according to an embodiment of the present disclosure.
7 is a flowchart of a positioning method according to an embodiment of the present disclosure.

The present disclosure relates to a positioning device and method.

DETAILED DESCRIPTION Some embodiments of the present disclosure are described in detail below with reference to exemplary drawings. In adding reference numerals to components of each drawing, the same components may have the same numerals as much as possible even if they are displayed on different drawings. In addition, in describing the present embodiments, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present technical idea, the detailed description may be omitted. When "comprises", "has", "consists of", etc. mentioned in this specification is used, other parts may be added unless "only" is used. In the case where a component is expressed in the singular, it may include the case of including the plural unless otherwise explicitly stated.

Also, terms such as first, second, A, B, (a), and (b) may be used in describing the components of the present disclosure. These terms are only used to distinguish the component from other components, and the nature, sequence, order, or number of the corresponding component is not limited by the term.

In the description of the positional relationship of components, when it is described that two or more components are "connected", "coupled" or "connected", the two or more components are directly "connected", "coupled" or "connected". ", but it will be understood that two or more components and other components may be further "interposed" and "connected", "coupled" or "connected". Here, other components may be included in one or more of two or more components that are “connected”, “coupled” or “connected” to each other.

In the description of the temporal flow relationship related to components, operation methods, production methods, etc., for example, "after", "continued to", "after", "before", etc. Alternatively, when a flow sequence relationship is described, it may also include non-continuous cases unless “immediately” or “directly” is used.

On the other hand, when a numerical value or corresponding information (eg, level, etc.) for a component is mentioned, even if there is no separate explicit description, the numerical value or its corresponding information is not indicated by various factors (eg, process factors, internal or external shocks, noise, etc.) may be interpreted as including an error range that may occur.

In the present specification, the Positioning Reference Signal (PRS) is transmitted from each cell or base station to the terminal in order to measure the position of the user equipment (UE), and is received by the terminal at a specific time. could be a signal

1 is a diagram showing the configuration of a positioning device according to an embodiment of the present disclosure.

Referring to FIG. 1, a positioning device 100 according to an embodiment of the present disclosure receives a positioning reference signal (PRS) from a base station and obtains a measurement value of the received positioning reference signal through channel estimation. a measurement value acquisition unit 110 that predicts a second state vector of the terminal using a preset movement model and a first state vector of the terminal acquired at a previous time; a state prediction unit 120 that predicts a location reference signal measurement value Based on the second state vector and the second state vector, the locations of virtual anchors (VAs) and scattering points (SPs) existing on the reception path of the location reference signal are estimated, and the locations are clustered to form a network of terminals. It may include an environment mapping unit 130 that maps the environment map and a state correction unit 140 that corrects the second state vector using the mapping result of the network environment map.

According to an embodiment, the measurement value acquisition unit 110 receives a positioning reference signal (PRS) from a base station (BS) and obtains a measurement value of the received positioning reference signal through channel estimation. can For example, the measurement value acquisition unit 110 may receive a location reference signal including a signal reflected from a visible path, a wall in the environment, or a scattering point. In addition, since the measurement value acquisition unit 110 cannot classify the received location reference signal according to the path, it may receive all received signals and acquire measurement values through channel estimation. For example, the measurement value acquisition unit 110 may receive a 5G mmWave signal periodically transmitted from one base station and obtain a measurement value indexed to a signal path through channel estimation.

As another example, the measurement value acquisition unit 110 may generate a measurement model based on measurement values of a location reference signal received through a multipath. For example, the measurement value acquisition unit 110 may include a state vector of the UE, location information and measurement values of a virtual anchor or a scattering point existing in the network environment of the UE based on the measurement value of the location reference signal received through multi-path. A measurement model including errors can be created. This measurement model can be used to calibrate the state vector of the UE.

According to an embodiment, the state prediction unit 120 may predict the second state vector of the terminal using a preset movement model and the first state vector of the terminal obtained at the previous time. For example, the state prediction unit 120 inputs the first state vector of the terminal acquired at a previous time to a preset movement model, and applies process noise modeled as a zero mean Gaussian distribution with a known covariance to obtain the current time In this case, the second state vector of the terminal can be predicted. Here, the movement model may be set to a known transition function for time series data.

According to an embodiment, the environment mapping unit 130 may determine a virtual anchor (VA) and a scattering point existing on a reception path of the location reference signal based on the measurement value of the location reference signal and the second state vector. , SP) can be estimated. Also, the environment mapping unit 130 may perform clustering on the estimated generation location and map the clustering to the network environment map of the terminal. For example, the environment mapping unit 130 may map the measurement value of the location reference signal to the Euclidean space and map the creation locations of virtual anchors and scattering points based on a Dirichlet Process (DP) algorithm. In addition, the environment mapping unit 130 may select and cluster an existing cluster or a new cluster based on the probability of a generation location. For example, the environment mapping unit 130 may classify a line-of-sight (LOS) measurement signal from a multi-path signal based on a Dirichlet process algorithm to map creation positions of virtual anchors and scattering points. In addition, the environment mapping unit 130 may cluster the creation location by calculating a probability that data exists in the creation location and selecting a cluster having the highest probability. Here, the Dirichlet process algorithm may be a Bayesian nonparametric (BNP) model including an infinite number of parameters.

As another example, the environment mapping unit 130 clusters virtual anchors or scattering points into one or more clusters based on their creation locations, and maps the locations of one or more clusters on a network environment map representing the surrounding network environment of the terminal. . Here, the network environment map may include location information of at least one of a base station, a cluster of virtual anchors, and a cluster of scattering points. For example, the environment mapping unit 130 may estimate a generated location by assuming that the received location reference signal is a multipath signal due to a virtual anchor. Also, if the estimated generation location is displayed as a cluster at one or more fixed points, the environment mapping unit 130 may classify the corresponding cluster as a cluster of virtual anchors. For another example, the environment mapping unit 130 may estimate the generated location by assuming that the location reference signal is a multipath signal due to the scattering point, excluding the clusters of the separated virtual anchors. In addition, if the estimated generation location is displayed as a cluster at one or more fixed points, the environment mapping unit 130 may classify the corresponding cluster as a cluster of spawning points.

According to an embodiment, the state correction unit 140 may correct the second state vector by using a mapping result of the network environment map. For example, the state correction unit 140 may correct the second state vector by reflecting the generated positions of virtual anchors and scattering points mapped on the network environment map to the measured value of the location reference signal and applying a specific filter. . For example, the state correction unit 140 applies an Extended Kalman Filter (EKF) to the second state vector of the terminal and the measured value of the location reference signal mapped on the network environment map to obtain the second state in real time. Vectors can be calibrated. However, the extended Kalman filter is an example, and a specific filter is not limited thereto.

2 is a diagram schematically illustrating a system to which a positioning device according to an embodiment of the present disclosure may be applied.

Referring to FIG. 2 , it relates to a system to which a positioning device 100 for estimating a location of a terminal 200 according to an embodiment of the present disclosure can be applied, and the positioning device 100 includes one base station 210 It is possible to estimate the location and state of the terminal 200 in a network environment composed of and one terminal 200 . For example, the network environment of the terminal 200 is a base station 210 that periodically transmits a millimeter wave (mmWave) signal, a virtual anchor 220 designated as a wide surface that reflects signals such as a wall 240, or a signal It can consist of scattering points 230, which refer to small objects that scatter. Here, the location of the base station 210, the virtual anchor 220, or the scattering point 230 may be a fixed landmark.

According to an example, the positioning target may be a computer or PDA other than the terminal 200 as long as it is connected to a communication network and can communicate with the positioning device 100, and the positioning target may also be a vehicle. That is, the terminal 200 in this specification describes an example of a positioning target, and is not limited thereto.

According to an example, the positioning device 100 may directly include a wireless communication module or may be configured in a form interconnected through a wired or wireless network. Accordingly, the positioning device 100 may be installed in the terminal 200 to be positioned, or may be in the form of a separate device. For example, the positioning device 100 may be a terminal 200 that receives a position reference signal periodically transmitted from a base station. Specifically, the received position reference signal is a millimeter wave signal periodically transmitted from the base station and may include a signal reflected from a wide surface and scattered by a scattering point. Accordingly, the positioning device 100 may receive a position reference signal including a signal through the visible path 250 and a signal reflected from the wall 240 or the scattering point 230 in the environment at regular time intervals.

In addition, the positioning device 100 includes a storage unit such as a memory, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), and a programmable logic unit (PLU). may be based on one or more general purpose or special purpose computers, such as hardware such as a microprocessor, software including a set of instructions, or combinations thereof, or any other device capable of executing and responding to instructions. .

3 is a flowchart illustrating an operation of estimating a location of a terminal by a positioning device according to an embodiment of the present disclosure.

Referring to FIG. 3 , the measurement value acquisition unit 110 of the positioning device according to an embodiment of the present disclosure may receive a location reference signal from a base station and obtain a measurement value through channel estimation (S310). For example, the measurement value acquisition unit 110 may receive a multi-path signal from each landmark every time and obtain a measurement value through channel estimation. Also, the measurement value acquisition unit 110 may generate a measurement model based on the measurement value of the received multipath signal. For example, the measurement value acquisition unit 110 may generate a measurement model of the measurement value of the obtained i-th signal as shown in Equation 1.

이고, 측정값 오차는

일 수 있다. 구체적으로,

는 도래 시간(time of arrival, TOA)이고,

는 방위각과 고도에 대한 도래각(difference of Arrival, DOA)이고,

는 방위각과 고도에 대한 출발 방향(direction of departure, DOD) 을 의미할 수 있다. 즉, xⁱ와 m은 가상 앵커와 산란 지점과 같이 다중 경로를 구성하는 물체의 위치와 유형를 의미하고, S_k는 단말의 위치 벡터와 방향각을 포함하는 상태 벡터를 의미할 수 있다. At this time,

, and the measurement error is

can be Specifically,

is the time of arrival (TOA),

is the difference of arrival (DOA) for the azimuth and altitude,

may mean a direction of departure (DOD) for azimuth and altitude. That is, x ⁱ and m denote positions and types of objects constituting a multipath, such as virtual anchors and scattering points, and S _k denotes a state vector including a position vector and azimuth angle of a terminal.

일 수 있고,

는 각각의 3차원 위치, α_k는 방향각, ζ_k는 이동 속도, ξ_k는 회전율 및 B_k는 클록 바이어스를 의미할 수 있다. 또한, 시간 k에서의 이동 모델은 수학식 2와 같이 표현할 수 있다. The state predictor 120 of the positioning device according to an embodiment may predict the state vector of the terminal using a preset movement model (S320). For example, the state predictor 120 may predict the second state vector of the terminal using a preset movement model and the first state vector of the terminal obtained at the previous time. For example, the state predictor 120 may predict the state vector of the terminal at time k using a preset movement model. Here, the state vector of the terminal at time k may have the same meaning as the second state vector. Specifically, the state vector S _k of the terminal is

can be,

is each three-dimensional position, α _k is the bearing angle, ζ _k is the movement speed, ξ _k is the rotation rate, and B _k is It may mean clock bias. In addition, the movement model at time k can be expressed as Equation 2.

를 가지는 k_max 시간에 대한 동적 모델일 수 있다. Here, the movement model g(•) may be a known transfer function, and q _k may mean process noise modeled as a zero mean Gaussian distribution using a known covariance Q. In addition, the migration model has a known transition density

It may be a dynamic model for k _max time with .

The environment mapping unit 130 of the positioning device according to an embodiment may use the measured value of the location reference signal and the predicted state vector of the terminal to map the network environment map of the terminal through clustering (S330). For example, the environment mapping unit 130 may map measurement values of the location reference signal to a 3D Euclidean space, and map creation locations of virtual anchors and scattering points based on a Dirichlet Process (DP) algorithm. For example, the environment mapping unit 130 may use a Dirichlet process algorithm to calculate a probability that data corresponding to measurement values of location reference signals belong to each cluster. Specifically, the prior distribution when the number of clusters is finite can be expressed as Equation 3.

는 각각 i 번째 측정값의 클러스터 인덱스, 총 데이터 개수, j 번째 클러스터에 할당된 데이터 개수, 집중도 매개변수 및 감마 함수를 의미할 수 있다. 또한, J는 총 클러스터의 개수를 의미할 수 있다. 다만, 디리클레 프로세스 알고리즘을 사용하면 J는 유한한 경우뿐만 아니라 무한한 경우에도 총 클러스터의 개수를 의미할 수 있다. here,

may mean the cluster index of the i th measurement value, the total number of data, the number of data allocated to the j th cluster, the concentration parameter, and the gamma function, respectively. Also, J may mean the total number of clusters. However, when using the Dirichlet process algorithm, J may mean the total number of clusters not only when it is finite but also when it is infinite.

Also, in detail, the environment mapping unit 130 may calculate a conditional prior probability for l ⁱ . The conditional prior probability for l ⁱ can be expressed as Equation 4.

Here, l ^-i may mean a set of l with all indicators except i. Also, if J is ∞, the conditional prior probability can be expressed as Equation 5.

In this way, the conditional prior probability for the new cluster can be expressed as Equation 6.

For another example, the environment mapping unit 130 may calculate how much data corresponding to the measurement value of the location reference signal is separated from each cluster. The degree of separation can be quantified using the likelihood of a Gaussian distribution through centroid and covariance. Specifically, the environment mapper 130 may calculate a probability that the measurement value y ⁱ occurs from the j-th cluster having a density of p _j (y) or a new cluster having a density of p ₀ (y). The conditional distribution can be expressed as Equations 7 and 8.

Through this, the measured value y ⁱ is It can be set as a cluster with the highest probability. Details of the Dirichlet process algorithm will be described later with reference to FIG. 6 .

The state correction unit 140 of the positioning device according to an embodiment may correct the predicted state vector using the mapping result of the environment map (S340). For example, the state correction unit 140 may correct the predicted second state vector of the terminal using a mapping result on a network environment map and a measurement value of a location reference signal. For example, the state correction unit 140 may correct the second state vector by reflecting the generated positions of the virtual anchors and scattering points mapped on the network environment map to the measured value of the position reference signal and applying a specific filter. there is. Here, the specific filter may be an Extended Kalman Filter (EKF), but is not limited thereto. The position and state of the terminal may be estimated based on the second state vector of the terminal calculated through the above series of processes.

4 is a diagram for explaining an operation of estimating a creation position of a virtual anchor by a positioning device according to an embodiment of the present disclosure.

Referring to FIG. 4 , in the positioning device according to an embodiment, the environment mapping unit 130 may estimate a creation position of a virtual anchor in a network environment of a terminal using a Dirichlet process algorithm. As an example, the environment mapping unit 130 uses the Dirichlet process algorithm in a network environment composed of 4 virtual anchors 220 and 4 scattering points due to 1 base station 210 and 4 walls 240 as shown in FIG. 4 . can be mapped using For example, the environment mapping unit 130 may estimate the creation location of the virtual anchor by assuming that the location reference signal is a multipath signal reflected by the four walls 240 and received from the virtual anchor. Then, the environment mapping unit 130 may cluster into one or more clusters using the Dirichlet process algorithm when the estimated generation location is intensively formed at one or more fixed points. Accordingly, the clustered cluster may be divided into clusters of virtual anchors. In addition, a measurement value of a line-of-sight (LOS) signal may indicate a location of a base station on a network environment map. Specifically, the creation position of the virtual anchor in the network environment map of the terminal may be displayed by forming a cluster around VA1 to VA4. On the other hand, spawning points are not marked as fixed points and can be distinguished from clusters of virtual anchors. Accordingly, the environment mapping unit 130 may distinguish between a multi-path signal due to the virtual anchor and a signal not caused by the virtual anchor.

5 is a diagram for explaining an operation of estimating the generation position of a scattering point by the positioning device according to an embodiment of the present disclosure.

Referring to FIG. 5 , in the positioning device according to an embodiment, the environment mapping unit 130 may estimate the location of spawning points in the network environment of the terminal by using the Dirichlet process algorithm. As an example, the environment mapping unit 130, as shown in FIG. 5 , in a network environment composed of one base station 210 and four virtual anchors 220 and four scattering points 230 due to four walls 240, Dirichlet It can be mapped using a process algorithm. For example, the environment mapping unit 130 may estimate the spawning location of the scattering point by assuming that the location reference signal is a multipath signal scattered by the four scattering points 230 and received from the scattering point. Then, the environment mapping unit 130 may cluster into one or more clusters using the Dirichlet process algorithm when the estimated generation location is intensively formed at one or more fixed points. Accordingly, the clustered clusters may be divided into clusters of scattering points. Specifically, the environment mapping unit 130 may display the location of spawning points in the network environment map of the terminal by forming clusters around SP1 to SP4, excluding the location of the virtual anchor cluster and the location of the base station. On the other hand, the generation location of clutter is not marked as a fixed point, so it can be distinguished from a cluster of scattering points. Accordingly, the environment mapping unit 130 may distinguish between a multipath signal due to a scattering point and a non-multipath signal.

6 is a diagram for explaining an algorithm for mapping an environment map by a positioning device according to an embodiment of the present disclosure.

이고, 시간 k에서 단말 상태의 사후 밀도는

를 의미할 수 있다. 예를 들어, 상태 예측부(120)는 수학식 9에 의해 시간 k에서 단말 상태의 사후 밀도를 예측할 수 있다. Referring to FIG. 6 , the positioning device according to an embodiment may map objects in a network environment of a terminal and predict a state vector of the terminal in a situation in which a plurality of multi-path signals are received. For example, the state prediction unit 120 may predict the a posteriori density of the terminal state at time k, given the a posteriori density of the terminal state at time k−1. where the a posteriori density of the terminal state at time k-1 is

and the a posteriori density of the terminal state at time k is

can mean For example, the state predictor 120 may predict the a posteriori density of the UE state at time k by Equation 9.

Also, the state predictor 120 may calculate s _k-1 and V _k-1 by Equations 10 and 11.

Here, Gk may be a Jacobbian matrix of the movement model g(•) calculated by Equation 12.

Accordingly, the calculated s _k and V _k may be corrected by identifying the measured value of the line-of-sight (LOS) signal and applying a specific filter.

,

및

로 표현할 수 있다. 구체적으로, 시간 k=0 일 때, 환경 매핑부(130)는 감지된 물체와 클러스터는 없지만, 기지국의 위치는 알 수 있다. 이에 따라, 환경 매핑부(130)는 단말의 네트워크 환경 맵을 초기화할 수 있다. For example, the environment mapping unit 130 may cluster the locations of objects in the network environment using the Dirichlet process algorithm, classify and map the types. For example, the environment mapping unit 130 may initialize a cluster of objects in a network environment. Here, the number of clusters according to object type m = {BS, VA, SP} at time k can be expressed as J _k,m . In addition, the set of centroids, covariances, and number of clusters according to type m of object at time k is

,

and

can be expressed as Specifically, at time k = 0, the environment mapping unit 130 has no detected object and cluster, but can know the location of the base station. Accordingly, the environment mapping unit 130 may initialize the network environment map of the terminal.

를 추정할 수 있다. As another example, the environment mapping unit 130 may estimate a generation location where measurement values of each location reference signal are converted into virtual anchors and scattering points before clustering. As a specific example, the environment mapping unit 130 uses Equations 13 and 14 to generate locations where measurement values are converted into virtual anchors.

can be estimated.

Here, H _x,k and H _s,k may be Jacobian matrices represented by ∂h/∂s _k and ∂h/∂x _k , respectively.

를 추정할 수 있다. For another specific example, the environment mapping unit 130 uses Equations 15 and 16 to determine the location where the measured value is converted into a scattering point.

can be estimated.

의 확률을 계산할 수 있다. 각각의 확률은 수학식 17 및 수학식 18을 이용하여 계산할 수 있다. As another example, the environment mapping unit 130 may calculate a probability of each generation location in an existing cluster or a new cluster and perform clustering by comparing the probabilities with a network environment map. For another specific example, the environment mapping unit 130 is each i-th generation position included in the existing cluster or the new cluster at time k.

can calculate the probability of Each probability can be calculated using Equations 17 and 18.

가 속한 클러스터의 인덱스일 수 있다. 또한, μ⁰은 환경 내에서 생성된 모든 생성 위치의 중심을 나타내는 지점(ex: 원점)이고, Σ⁰은 신규 클러스터의 고정된 큰 공분산을 의미할 수 있다.where l is the i-th generation position

It may be the index of the cluster to which . In addition, μ ⁰ is a point (ex: origin) representing the center of all generated positions generated in the environment, and Σ ⁰ may mean a fixed large covariance of a new cluster.

As another specific example, the environment mapping unit 130 may compare the calculated probabilities and map the generation location to the cluster having the highest probability. Probabilities can be compared using Equation 19.

In this case, if j ^* is less than or equal to J _k−1,m, the environment mapping unit 130 may select an existing cluster and update the covariance and centroid of the J ^* th cluster. The covariance and centroid can be updated using Equations 20 and 21.

Also, when a new cluster is selected, the environment mapping unit 130 may use the center and covariance of the estimated generation location as the center and covariance of the cluster. Also, the environment mapping unit 130 may classify clusters in which the number of objects is greater than or equal to a threshold value as landmarks.

For example, the state correction unit 140 may predict the a posteriori density of the terminal state at time k based on the measured density. Here, the posterior density can be predicted using Equation 22.

Here, η can be a regularization term.

For example, the state correction unit 140 may correct the state vector of the terminal using Equations 23 and 24.

는 각각 단말 상태의 공분산과 예측된 공분산을 의미할 수 있다. 이에 따라, 측정 장치는 위의 과정을 통해 단말의 위치 및 상태를 추정할 수 있다. 또한, H_k는 시간 k에서 이동 모델 h(·)의 Jacobian matrix일 수 있다.where K _k is the Kalman gain,

may mean the covariance of the terminal state and the predicted covariance, respectively. Accordingly, the measurement device may estimate the location and state of the terminal through the above process. Also, H _k may be a Jacobian matrix of the movement model h(·) at time k.

Hereinafter, a positioning method that can be performed by the positioning device described with reference to FIGS. 1 to 6 will be described. However, detailed descriptions of some embodiments or some operations described in FIGS. 1 to 6 may be omitted below, but this is only to prevent duplication of description, so the positioning method may provide the same positioning device as described above. can

7 is a flowchart of a positioning method according to an embodiment of the present disclosure.

Referring to FIG. 7 , the positioning method according to an embodiment of the present disclosure may include a measurement value obtaining step of obtaining a measurement value of a location reference signal (S710). For example, the positioning device may receive a positioning reference signal (PRS) from a base station and obtain a measurement value of the received positioning reference signal through channel estimation. For example, a positioning device may receive a position reference signal that includes a signal reflected from a line of sight, a wall in the environment, or a scattering point. Also, since the positioning device cannot classify received position reference signals according to paths, it is possible to receive all received signals and obtain measurement values through channel estimation. Specifically, the positioning device may receive a 5G mmWave signal periodically transmitted from one base station and obtain a measurement value indexed to a signal path through channel estimation.

As another example, the positioning device may generate a measurement model based on a measurement value of a location reference signal received through a multipath. For example, the positioning device measures a state vector of the terminal based on a measurement value of a position reference signal received through a multi-path, position information of a virtual anchor or scattering point existing in a network environment of the terminal, and a measurement value error. model can be created.

The positioning method according to an embodiment may include a state prediction obtaining step of predicting a state vector of a terminal (S720). For example, the positioning device may predict a second state vector of the terminal using a preset movement model and a first state vector of the terminal obtained at a previous time. For example, the positioning device inputs the first state vector of the terminal obtained at a previous time to a preset movement model, and applies process noise modeled as a zero-mean Gaussian distribution with a known covariance to obtain the first state vector of the terminal at the current time. Two-state vectors can be predicted. Here, the movement model may be set to a known transition function for time series data.

The positioning method according to an embodiment may include an environment mapping step of mapping a network environment map (S730). For example, the positioning device generates positions of a virtual anchor (VA) and a scattering point (SP) existing on a reception path of the position reference signal based on the measurement value of the position reference signal and the second state vector. can be estimated. And, the positioning device may cluster the estimated generated location and map it to the network environment map of the terminal. For example, the positioning device may map measurement values of position reference signals to Euclidean space, and map creation positions of virtual anchors and scattering points based on a Dirichlet Process (DP) algorithm. Also, the positioning device may perform clustering by selecting an existing cluster or a new cluster based on the probability of the generated location.

As another example, the positioning device may cluster virtual anchors or scattering points into one or more clusters based on the generated location, and map the locations of the one or more clusters on a network environment map representing a network environment around the terminal. Here, the network environment map may include location information of at least one of a base station, a cluster of virtual anchors, and a cluster of scattering points. For example, the positioning device may estimate the generated position by assuming that the received position reference signal is a multipath signal due to a virtual anchor. In addition, when the estimated generated position is displayed as a cluster at one or more fixed points, the positioning device may classify the corresponding cluster as a cluster of virtual anchors. For another example, the positioning device may estimate the generated position by assuming that the position reference signal is a multi-path signal due to the scattering point, excluding clusters of divided virtual anchors. In addition, when the estimated generated position is displayed as a cluster at one or more fixed points, the positioning device may classify the corresponding cluster as a cluster of scattering points.

The positioning method according to an embodiment may include a state correction step of correcting the state vector of the terminal (S740). For example, the positioning device may correct the second state vector by using a mapping result of the network environment map. For example, the positioning device may reflect generated positions of virtual anchors and scattering points mapped on a network environment map to measurement values of the position reference signal, and apply a specific filter to correct the second state vector. Specifically, the positioning device may correct the second state vector in real time by applying an Extended Kalman Filter (EKF) to the second state vector of the terminal and the measured value of the position reference signal mapped on the network environment map. there is. However, the extended Kalman filter is an example, and a specific filter is not limited thereto.

As described above, according to the present disclosure, a positioning device and method can be provided. In particular, it is possible to provide a positioning device and method for estimating the position and state of a terminal by identifying the environment around the terminal through a clustering technique in a wireless communication environment. Specifically, to provide a positioning device and method for estimating the position of a terminal in real time by estimating the generation position of a signal and classifying objects using received signals through clustering-based environment mapping in a wireless network composed of terminals and base stations. can

The above description is merely illustrative of the technical idea of the present disclosure, and various modifications and variations can be made to those skilled in the art without departing from the essential characteristics of the technical idea. In addition, since the present embodiments are not intended to limit the technical idea of the present disclosure, but to explain, the scope of the present technical idea is not limited by these embodiments. The scope of protection of the present disclosure should be construed by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of rights of the present disclosure.

Claims (16)

a measurement value acquisition unit that receives a positioning reference signal (PRS) from a base station and acquires a measurement value of the positioning reference signal received through channel estimation;
a state prediction unit for predicting a second state vector of the terminal using a preset movement model and a first state vector of the terminal obtained at a previous time;
Estimating creation positions of virtual anchors (VA) and scattering points (SPs) existing on a reception path of the location reference signal based on the measurement value of the location reference signal and the second state vector; , an environment mapping unit clustering the generated location and mapping the clustering location to a network environment map of the terminal; and
and a state correction unit correcting the second state vector by using a mapping result of the network environment map. According to claim 1,
The measurement value acquisition unit,
Based on the measurement value of the location reference signal received through multipath, a measurement model including a state vector of the UE, location information of the virtual anchor or the scattering point existing in the network environment of the UE, and measurement error is generated. Positioning device characterized in that for generating. According to claim 1,
The environment mapping unit,
Mapping the measurement value of the position reference signal to Euclidean space, and mapping the virtual anchor and the spawning position of the scattering point based on a Dirichlet Process (DP) algorithm,
The positioning device characterized in that for clustering by selecting an existing cluster or a new cluster based on the probability of the generated position. According to claim 1,
The environment mapping unit,
The positioning device characterized by clustering the virtual anchor or the scattering point into one or more clusters based on the generated location, and mapping the locations of the one or more clusters on the network environment map indicating a network environment around the terminal. . According to claim 4,
The network environment map,
and location information of at least one of the base station, the virtual anchor cluster, and the scattering point cluster. According to claim 4,
The environment mapping unit,
Assuming that the received position reference signal is a multipath signal due to the virtual anchor, if the generated position is displayed as a cluster at one or more fixed points, the positioning device classifies the cluster as a cluster of the virtual anchor. . According to claim 6,
The environment mapping unit,
Excluding the cluster of the virtual anchor, assuming that the position reference signal is a multipath signal due to the scattering point, if the generated position is displayed as a cluster at one or more fixed points, the cluster is classified as a cluster of the scattering point. A positioning device characterized in that According to claim 1,
The state correction unit,
Wherein the second state vector is corrected by reflecting the generated positions of the virtual anchor and the scattering point mapped on the network environment map to the measured value of the position reference signal, and applying a specific filter. A measurement value acquisition step of receiving a positioning reference signal (PRS) from a base station and obtaining a measurement value of the positioning reference signal received through channel estimation;
a state prediction step of predicting a second state vector of the terminal using a preset movement model and a first state vector of the terminal obtained at a previous time;
Estimating creation positions of virtual anchors (VA) and scattering points (SPs) existing on a reception path of the location reference signal based on the measurement value of the location reference signal and the second state vector; , an environment mapping step of clustering the generated locations and mapping them to a network environment map of the terminal; and
and a state correction step of correcting the second state vector by using a mapping result of the network environment map. According to claim 9,
The measurement value acquisition step,
Based on the measured value of the location reference signal received through multi-path, the state vector of the terminal, the self-solution in the network environment of the terminal is a measurement model including the location information of the virtual anchor or the scattering point and a measurement error. Positioning method characterized in that for generating. According to claim 9,
The environment mapping step,
Mapping the measurement value of the position reference signal to Euclidean space, and mapping the virtual anchor and the spawning position of the scattering point based on a Dirichlet Process (DP) algorithm,
The positioning method characterized in that for clustering by selecting an existing cluster or a new cluster based on the probability of the creation position. According to claim 9,
The environment mapping step,
The positioning method characterized in that the virtual anchor or the scattering point is clustered into one or more clusters based on the generated location, and the location of the one or more clusters is mapped on the network environment map indicating a network environment around the terminal. . According to claim 12,
The network environment map,
and location information of at least one of the base station, the virtual anchor cluster, and the scattering point cluster. According to claim 12,
The environment mapping step,
Assuming that the received location reference signal is a multipath signal due to the virtual anchor, if the generated location is displayed as a cluster at one or more fixed points, the cluster is classified as a cluster of the virtual anchor. . According to claim 13,
The environment mapping step,
Excluding the cluster of the virtual anchor, assuming that the position reference signal is a multipath signal due to the scattering point, if the generated position is displayed as a cluster at one or more fixed points, the cluster is classified as a cluster of the scattering point. A positioning method characterized in that for doing. According to claim 9,
In the state correction step,
Wherein the second state vector is corrected by reflecting the generated positions of the virtual anchor and the scattering point mapped on the network environment map to the measurement value of the position reference signal, and applying a specific filter.

Images