WO2010026503A1

WO2010026503A1 - Angle of arrival (basis selection) and location estimation system

Info

Publication number: WO2010026503A1
Application number: PCT/IB2009/053645
Authority: WO
Inventors: Gunes Kurt
Original assignee: Turkcell Iletisim Hizmetleri Anonim Sirketi
Priority date: 2008-09-03
Filing date: 2009-08-18
Publication date: 2010-03-11
Also published as: TR200806652A1

Abstract

The present invention is related to 2- and/or 3-dimensional location estimation of mobile communication devices by means of antenna arrays (2), angle of arrival estimation and basis selection algorithms. In addition, said location estimation system (1) enables to denote the location of the mobile communication device within the coordinates as well as location estimation.

Description

ANGLE OF ARRIVAL (BASIS SELECTION) AND LOCATION ESTIMATION SYSTEM

Field of the Invention

This invention is related to 2- and/or 3-dimensional location estimation of the mobile communication devices by means of an antenna array.

Prior Art

Several location estimation methods have been used in the prior art. Most of these methods, which are currently in use, are time of arrival (TOA) based systems. In TOA, distance between a receiver and a transmitter is used as a time scale by using the connection between the velocity of the light in the space and a carrier frequency of a signal. For example, in US-Patent 6163696, signals designed to fulfill this objective are used to estimate location information.

Today, TDOA (time difference of arrival), GPS (Global Positioning System) based methods and RF (radio frequency) fingerprinting methods are used in estimation of location other than TOA (time of arrival).

In KR20040067533, a method used to estimate the location of a mobile phone is disclosed. In this method, location estimate is calculated by processing the radio signals received from the adjacent antennas. In the KR20040067533 known as RF fingerprinting method, more than one base station is required in order to estimate location of a mobile phone, which increases the number of the receivers. In addition, measurements are required to be previously made in the region in order to determine a base station. In the systems using RF Fingerprinting method, measurements must be made again if the environmental conditions change.

In US4750147, a method used to estimate location of a signal source is disclosed. In this method, signals transmitted from the source are determined by the receivers. In the system, location of the signal source is estimated by using ESPRIT algorithm. Location estimation performance of the system decreases when algorithm used in this method encounters with a high correlation in case of multi-path channels. Moreover, if the differences between the incidence angles of the signals transmitted from the sources are higher than a certain degree, the used algorithm cannot resolve these angles.

In US2002190902, a FFT (Fast Fourier Transform) based method is used to estimate location of the signal source. Calculations in this method are not as much as in ESPRIT while the resolution performance of the system is rather low. In addition, the performance drops further if there are more than one channel in the system. Resolution performance is defined as the ability to differentiate (resolve) the incoming signals from the close angles. In addition, the closer the angles to be resolved are, the higher the resolution performance is.

Summary of the Invention

The objective of the present invention is to realize a location estimation system used for 2- and/or 3-dimensional location estimation of the mobile communication devices.

Another objective of the invention is to realize a location estimation system enabling to denote the location of the mobile communication device within the coordinates as well as location estimation.

A further objective of the present invention is to realize a location estimation system allowing use of the results obtained by means of location estimation system in developing position based services in the wireless communication devices. Detailed Description of the Invention

The Location Estimation System realized to fulfill the objective of the present invention is illustrated in the accompanying figures wherein,

Figure 1 is a view of a location estimation system.

Figure 2 is a view of receiving the signals emanated from a source in the far- field region by a set of antennas.

Figure 3 is a view of making 2-dimensional location estimation. Figure 4 is a view making 3-dimensional location estimation. Figure 5 is a view of flow chart of a location estimation system.

1. Location Estimation System

2. Antenna arrays 3. Sampler

4. DOA (Direction of Arrival) estimator

5. Location estimator

In the location estimation system (1), the location is estimated by using antenna arrays (2). In DOA estimators (4), incidence angle is obtained by using BS (basis selection) algorithms, and the obtained incidence angle is converted into location information in the location estimator (5).

With the inventive location estimation system (1), 2-dimensional or 3-dimensional location estimation of the mobile communication devices are made. In said system

(1), location information is estimated by processing signals received from the antennas with one of the BS (basis selection) algorithms. In said system (1), antennas included in the antenna arrays (2) detect the incoming waves, these waves are then processed by the sampler (3), and a matrix is generated. This generated matrix is resolved by a BS (basis selection) algorithm in a DOA estimator (4) and direction of arrival of the wave is estimated. This process proceeds at least in two antenna arrays. Direction of arrival information is converted into location information by being processed in the location estimator

(5).

Said system (1) comprises two or more antenna arrays (2). Each antenna array (2) has at least two antennas located in a rectangle fashion. Each antenna array (2) can be viewed as a part of a base station. System (1) performance and capacity is increased by processing the outputs/signals received from the several antennas in the system.

Signals of the mobile communication device are received by the antennas in the antenna array (2) (100) and are converted into a matrix by means of two and/or more samplers (3) (200). Matrix outputs received from the samplers (3) are fed into the BS (basis selection) algorithms in the DOA estimators (4) (300). Basis selection algorithms determine angles and probabilities of the signals sent by mobile communication device (400). Location is estimated by converting said angles into the location information in the location estimators (5) (500).

The aim of the BS (basis selection) algorithms used in DOA estimators (4) is to resolve the signal. The aim of the resolution realized by BS (basis selection) algorithms such as MP (Matching Pursuit), OMP (Orthogonal Matching Pursuit) and FTB-OMP (Flexible Tree-search Based OMP) is to form the input signal by using the minimum number of basis vector as possible.

In the inventive location estimation system (1), two antenna arrays (2) are used to estimate the incidence angle. Assuming that there are N antennas in the antenna arrays (2), array output is converted into a NxI dimensional x vector. Phase difference between i^th and j^th components of the defined x vector varies depending on the location of/ andy antennas and incidence angle of the signal. An A dictionary is defined according to location of the antennas in the antenna array (2). Preferably, this dictionary is prepared according to M incidence angles.

Rx-I Rx-2 Rx-N

Incidence angle

1 1 1 e^Jψl e^{Jψ 2} _eJψM

I O A=

J(N-I)ψl j(N-l)_Ψ 2 j(N-l)ψM

In the defined A dictionary, ψ_t , Q₁ are phase difference to be formed between the antennas according to incidence angle. In said equation, M different incidence

15 angles are dealt. Cosine values of the angles are calculated as M=COs(^ ) by using BS (basis selection) algorithms.

Outputs of antenna array (2) are the linear combinations of columns of the A matrix depending on the angles of incidence and multi-path channel. Hence, the 0 problem of DOA (direction of arrival) estimation is enabled to be reduced to selection of correct linear columns. It is sufficient to select the closest column for the signal arriving from a certain direction. The correct vector combination is required for the signals arriving from different directions. Complexity level of this solution is high. MP and OMP algorithms can be used as a resolution method 5 having lower complexity level.

In the inventive location estimation system (1), signals fed to the antenna arrays (2) are sampled by means of two or more samplers (3). The samplers (3) operate synchronously. The signal received by the samplers (3) can be OFDM (Orthogonal frequency division multiplexing), single carrier enabling data transmission by changing the parameters sent in each symbol or CDMA (Code division multiple access) modulated. In addition, waveforms received by array antennas are planar. The requirement for waveforms to be planar is easily satisfied in the far field region, that is, the location in which the signal is received. Signals are received as planar since the signals are received by the receiver from a far- field location.

Received samples at each antenna array (2) component have a phase difference based on the source location. Received samples can be resolved in the DOA estimators (4) by means of the algorithm and location estimation can be made.

where j is the complex exponent, M is the number of distinct point sources, h.(t) is the associated information bit and n(t) is the AWGN (additive white Gaussian noise) component. The center frequency is w₀, and -(§, .αΛ is the time delay between the reference sensor (first sensor) and the (/, k)^th sensor. The azimuth and inclination DOAs of z^th source are denoted by §, and a., respectively. The DOA (direction of arrival) values are elements of the set (-π/2, π/2] . The phase difference due to time delay is a function of the physical location of the sensors receiving RF signals in the antenna array (2). Moreover, no restrictions are imposed on the shape of the antenna array (2) used in the location estimation system (1).

Received samples are converted into matrices in the samplers (3) based on their location within the sensor array. These matrices collected at the output of these samplers are fed into 2-Dimensional DOA estimators (4). The 2-Dimensional DOA (direction of arrival) estimator (4) is based on the 1-Dimensional DOA (direction of arrival) estimation method, and uses the BS (basis selection) algorithms.

Subsets are chosen from a large redundant set of vectors to match a given data. The BS (basis selection) algorithms are suitable for selection of basis for signal decomposition by determining the subsets. This problem has various applications such as time/frequency representations, speech coding, and spectral estimation.

In DOA (direction of arrival) estimation, the set of vectors are modeled as possible outputs of the antenna array elements when the signal is arriving from a certain direction. Problem of selecting correct linear combination of these elements is equivalent to the problem of selecting correct DOA (direction of arrival). The DOA (direction of arrival) estimation methodology does not require a priori knowledge on the number of angles to be resolved and uses very small amount of snapshots for convergence. The output of the algorithm is directly the angles of arrivals and their corresponding amplitudes; hence it does not require any further process as in the case of conventional DOA (direction of arrival) estimators.

The DOA (direction of arrival) estimator (4) makes use of the known phase difference between antenna elements and uses iterative BS algorithms. These algorithms are MP (Matching Pursuit), OMP (Orthogonal Matching Pursuit) or FTB-OMP (Flexible Tree-Search Based Orthogonal Matching Pursuit).

The main advantages of application of BS algorithms to DOA (direction of arrival) problem are the decreased complexity and increased resolution. The BS algorithms give DOA' s and their corresponding amplitudes as output, so they do not require post-processing, and the power values at different angles can be calculated by means of the outputs. In 3 -dimensional location estimation, information necessary for application of the BS algorithms is generated by using the location information of the sensors of the antenna arrays (2) for both inclination and azimuth angles. Outputs of these 2- dimensional DOA (direction of arrival) estimators (4) are the inclination and the azimuth angles of the received multipath components and possibly co-channel interferers. These obtained angles are fed into the location estimator (5) along with their amplitude values as generated by the basis selection algorithms.

Location estimator (5) combines the outputs of two or more 2-dimensional DOA (direction of arrival) estimators (4) in order to make 3-dimensional location estimation. The relative positions of the antenna arrays (2) must be available in the location estimator system (1). Outputs of the location estimators (5) are the possible location estimates of the signal source with an associated probability value. The probability value can be estimated based on the normalized received signal amplitude values.

Within the framework of this basic concept, it is possible to develop a wide variety of applications. The inventive "Location Estimation System (1)" cannot be limited to the examples provided above to explain the invention. The present invention is essentially according to the claims.

Claims

1. A location estimation system (1) used for 2 and/or 3 dimensional location estimation of mobile communication devices, comprising antenna arrays (2), sampler (3), DOA (direction of arrival) estimator (4), location estimator (5), characterized in that signal of the mobile communication device is received by the antennas in the antenna array (2) (100), signal is converted into matrix by means of two and/or more samplers (3) (200), matrix outputs received from the samplers (3) are fed into the BS (basis selection) algorithms in the DOA estimators (4) (300), basis selection algorithms determine angles and probabilities of the signals sent by mobile communication device (400), the angles are converted into the location information in the location estimators (5) (500).

2. A location estimation system (1) according to Claim 1, characterized by an antenna array (2) comprising more than one antenna which increases the performance and capacity of the system and enables reception of signals/outputs.

3. A location estimation system (1) according to any one of the preceding claims, characterized by DOA (direction of arrival) estimators (4) which enables location information estimation by processing the signals received from the antennas with one of the BS (basis selection) algorithms.

4. A location estimation system (1) according to any one of the preceding claims, characterized by a DOA (direction of arrival) estimator (4) which proceeds at least in two directions to determine angle of incidence.