CN105611627A

CN105611627A - Method for estimating AOA of WLAN access point based on double antennas

Info

Publication number: CN105611627A
Application number: CN201610012316.1A
Authority: CN
Inventors: 田增山; 张振源; 吴自鹏; 周牧; 廉颖慧; 李泽; 金悦; 林天瑜; 王嘉诚; 张千坤
Original assignee: Chongqing University of Post and Telecommunications
Current assignee: Chongqing University of Post and Telecommunications
Priority date: 2016-01-08
Filing date: 2016-01-08
Publication date: 2016-05-25
Anticipated expiration: 2036-01-08
Also published as: CN105611627B

Abstract

A method for estimating the AOA of a WLAN access point based on dual antennas, using a common two-antenna WLAN access point, and using the received CSI information at the receiving end to calculate the time of flight (Time?of?flight, TOF) and angle of arrival for joint estimation. The invention fully considers the influence of indoor complex multipath environment on angle estimation, and the average error of estimated angle is within 5 degrees, which can meet the requirement of direct path angle estimation under indoor multipath signals.

Description

Estimation Method of WLAN Access Point AOA Based on Dual Antennas

技术领域technical field

本发明涉及信号处理技术，无线定位技术领域，具体涉及基于双天线的WLAN接入点AOA的估计方法，适用于室内环境的定位系统。The invention relates to the field of signal processing technology and wireless positioning technology, in particular to a method for estimating AOA of a WLAN access point based on dual antennas, and is suitable for a positioning system in an indoor environment.

背景技术Background technique

在移动通信领域，人们对基于位置的服务(LocationBasedServices，LBS)的需求正在不断增长，而随着无线局域网络(WirelessLocalAreaNetworks，WLAN)的普及，基于WLAN的室内定位系统正是迎合了这种需求的一个新兴的研究热点。In the field of mobile communications, people's demand for location-based services (Location Based Services, LBS) is growing, and with the popularity of wireless local area networks (Wireless Local Area Networks, WLAN), the WLAN-based indoor positioning system just caters to this demand. An emerging research hotspot.

在室内环境下获取来波的AOA(angleofarrival，即波达角)信息可以为室内定位系统提供关键的定位参数，从而实现室内环境下的高精度定位。同时，和目前基于RSSI的定位技术相比，基于AOA的定位技术不需要进行位置指纹库构建，节省了大量的人力成本。同时，基于AOA的定位技术是一种被动定位技术可以轻松的实现基站的网络侧定位，不需要用户安装任何多余软件。因此，如何准确的估计来波方向的到达角是解决室内定位问题的关键。Obtaining AOA (angle of arrival, angle of arrival) information of incoming waves in an indoor environment can provide key positioning parameters for an indoor positioning system, thereby achieving high-precision positioning in an indoor environment. At the same time, compared with the current RSSI-based positioning technology, the AOA-based positioning technology does not need to construct a location fingerprint database, which saves a lot of labor costs. At the same time, the AOA-based positioning technology is a passive positioning technology that can easily realize the network-side positioning of the base station without requiring users to install any redundant software. Therefore, how to accurately estimate the angle of arrival of the incoming wave direction is the key to solving the indoor positioning problem.

传统的AOA估计算法如MUSIC算法，ESPRIT算法主要采用专用的多天线阵列设备，利用的信号子空间和噪声子空间的正交性实现角度估计。但是传统的角度估计算法也存在明显的缺陷，必须采用专用的大规模阵列天线才能准确的估计出AOA信息，这就为AOA估计在室内的应用推广设置了障碍。Traditional AOA estimation algorithms such as MUSIC algorithm and ESPRIT algorithm mainly use special multi-antenna array equipment to realize angle estimation by utilizing the orthogonality of signal subspace and noise subspace. However, the traditional angle estimation algorithm also has obvious defects. It is necessary to use a dedicated large-scale array antenna to accurately estimate the AOA information, which sets up obstacles for the application and promotion of AOA estimation indoors.

目前，移动通信系统中大多采用智能天线技术，因此服务基站将能提供较准确的电波到达角信息，并提供基于网络侧定位的位置服务。例如，目前基于LTE网络的AOA定位系统，该系统利用MIMO的预编码机制来实现AOA的获取。在室内应用中，目前已有Ubicarse系统，在WLAN环境下，采用SAR(SyntheticApertureRadar)的设计思路，通过旋转接收端天线模拟大天线阵列实现角度估计。另外，目前的另外一款系统ArrayPhaser系统通过级联WLAN接入点，实现多天线阵列构造的系统的设计。除此之外，DirectionFinding系统利用WLAN双天线，采用干涉仪测向的原理实现到达角测量。虽然上述系统能够准确的进行到达估计，但是上述系统都存在不同程度的缺陷。Ubicarse系统的实现需要对接收设备天线进行改进；ArrayPhaser系统需要对WLAN接入点设备做较大的改造，同时在级联设计中必须充分考虑到不同设备存在的同步误差，这也后期系统的维护提出了较大的挑战。DirectionFinding系统通过两天线进行角度估计，不能有效的估计出环境中存在的多径信号，同时估计出的角度分辨率较差。上述存在的这些问题都会给AOA定位系统在室内定位领域中的推广带来局限性。At present, most mobile communication systems use smart antenna technology, so the serving base station will be able to provide more accurate angle-of-arrival information of radio waves, and provide location services based on network-side positioning. For example, the current AOA positioning system based on the LTE network uses a MIMO precoding mechanism to achieve AOA acquisition. In indoor applications, there is currently a Ubicarse system. In a WLAN environment, the design idea of SAR (Synthetic Aperture Radar) is adopted to simulate a large antenna array by rotating the receiving end antenna to achieve angle estimation. In addition, another current system, the ArrayPhaser system, realizes the design of a system with multi-antenna array structure by cascading WLAN access points. In addition, the DirectionFinding system uses WLAN dual antennas and uses the principle of interferometer direction finding to measure the angle of arrival. Although the above-mentioned systems can accurately perform arrival estimation, all of the above-mentioned systems have defects to varying degrees. The realization of the Ubicarse system needs to improve the antenna of the receiving device; the ArrayPhaser system needs to make a major modification to the WLAN access point device. posed a greater challenge. The DirectionFinding system performs angle estimation through two antennas, which cannot effectively estimate the multipath signals existing in the environment, and the estimated angle resolution is poor. The above-mentioned problems will bring limitations to the promotion of the AOA positioning system in the field of indoor positioning.

发明内容Contents of the invention

本发明的目的是提供一种基于双天线的WLAN接入点AOA的估计方法，它估计角度的平均误差在5度左右，能远远满足室内多径信号的角度估计的需求。The purpose of the present invention is to provide a method for estimating the AOA of a WLAN access point based on dual antennas. The average error of the estimated angle is about 5 degrees, which can far meet the requirements for angle estimation of indoor multipath signals.

本发明所述的基于双天线的WLAN接入点AOA的估计方法，包括以下步骤：The estimation method of the WLAN access point AOA based on dual antennas of the present invention comprises the following steps:

步骤1、配置无线局域网；Step 1. Configure wireless LAN;

步骤2、在发射端采用多个子载波对原始数据进行正交调制；Step 2, using multiple subcarriers at the transmitting end to perform quadrature modulation on the original data;

步骤3、估计每个子载波的信道状态信息：在接收端接收到OFDM的CSI信道信息矩阵CSI_matrix：Step 3. Estimate the channel state information of each subcarrier: the CSI channel information matrix CSI _matrix of OFDM is received at the receiving end:

${CSI CSI}_{m m a a t t r r i i x x} = = [\begin{matrix} {csi csi}_{11,, 11} & {csi csi}_{11,, 22} & ... ... & {csi csi}_{11,, 1919} & {csi csi}_{11,, 2020} \\ {csi csi}_{22,, 11} & {csi csi}_{22,, 22} & ... ... & {csi csi}_{22,, 1919} & {csi csi}_{22,, 2020} \end{matrix}];;$

其中，csi_i,j为第i根天线上第j个子载波的信道信息值；Among them, csi _i,j is the channel information value of the jth subcarrier on the ith antenna;

步骤4、采用空间平滑算法对接收到的CSI信道矩阵进行去相关处理；Step 4, using a spatial smoothing algorithm to perform decorrelation processing on the received CSI channel matrix;

步骤5、利用去相关后的CSI信息对信号到达阵列天线的飞行时间TOF和到达角度估计算法进行联合估计，得到波达方向的估计值。Step 5. Using the de-correlated CSI information to jointly estimate the time-of-flight (TOF) of the signal arriving at the array antenna and the angle-of-arrival estimation algorithm to obtain an estimated value of the direction of arrival.

所述步骤3中，估计每个子载波的信道状态信息，采用最小二乘(LS)技术进行信道估计；In the step 3, estimate the channel state information of each subcarrier, and use the least squares (LS) technique to estimate the channel;

最小二乘技术是假设N-1个子载波是正交的，即没有ISI(InterSymbolInterference即码间串扰)，接收到的训练信号为Y[k],k＝0,1,2,…,N-1对信道H的估计为将代价函数最小化：The least squares technique assumes that N-1 subcarriers are orthogonal, that is, there is no ISI (InterSymbolInterference, intersymbol interference), and the received training signal is Y[k], k=0,1,2,...,N- 1 The estimate for the channel H is cost function minimize:

$J J ((\overset{^^}{H h})) = = | | | | Y Y - - X x \overset{^^}{H h} | | {| |}^{22} = = {Y Y}^{H h} Y Y - - {Y Y}^{H h} X x \overset{^^}{H h} - - {\overset{^^}{H h}}^{H h} {X x}^{H h} Y Y + + {\overset{^^}{H h}}^{H h} {X x}^{H h} A A \overset{^^}{H h};;$

令代价函数关于的偏导数等于0，得到LS信道估计的解为：let the cost function about The partial derivative of is equal to 0, and the solution of LS channel estimation is obtained as:

${\overset{^^}{H h}}_{L L S S} = = (({X x}^{H h})) {X x}^{- - 11} {X x}^{H h} Y Y = = {X x}^{- - 11} Y Y;;$

令表示中的元素，k＝0,1,2,…,N-1，由无ICI的假设条件下可知X为对角矩阵，因此每个子载波上的LS信号估计能表示为：make express The elements in , k=0,1,2,...,N-1, under the assumption of no ICI, it can be known that X is a diagonal matrix, so the LS signal estimation on each subcarrier can be expressed as:

${\overset{^^}{H h}}_{L L S S} [[k k]] = = \frac{Y Y [[k k]]}{X x [[k k]]},, k k = = 00,, 11,, 22,, ... ... N N - - 1. 1.$

所述步骤5中，去相关后的CSI信息对信号到达阵列天线的飞行时间TOF和到达角度估计算法进行联合估计时构造包含AOA和TOF信息的二维方向矩阵：In the step 5, the CSI information after decorrelation constructs a two-dimensional direction matrix containing AOA and TOF information when the time-of-flight TOF and the angle-of-arrival estimation algorithm of the signal arrival array antenna are jointly estimated:

$A A = = [[\overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{11},, {τ τ}_{11})),, \overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{22},, {τ τ}_{22})),, ... ...,, \overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{N N},, {τ τ}_{N N}))]];;$

其中是M*1的方向向量，θ_k是第k条路径的到达角，τ_k是第k条路径的飞行时间，最后建立了利用OFDM多载波信息对TOF和AOA进行联合估计的系统；in is the direction vector of M*1, θ _k is the angle of arrival of the k-th path, τ _k is the flight time of the k-th path, and finally a system for joint estimation of TOF and AOA using OFDM multi-carrier information is established;

AOA估计方法是利用MUSIC算法完成到达角度的估计；MUSIC算法是根据N个接收信号矢量得到协方差矩阵的估计值；The AOA estimation method is to use the MUSIC algorithm to complete the estimation of the angle of arrival; the MUSIC algorithm is to obtain the estimated value of the covariance matrix according to the N received signal vectors;

$R R = = \frac{11}{N N} {Σ Σ}_{n no = = 11}^{N N} x x ((n no)) {x x}^{H h} ((n no));;$

其中，R为协方差矩阵，然后对其进行特征值分解R＝UΣU^H，按照特征值的大小顺序，把与信号个数K相等的最大特征值对应的特征向量看作信号子空间，把剩下的(M-K)个特征值对应特征向量看作噪声子空间，则 $R = U_{S} Σ_{S} U_{S}^{H} + U_{N} Σ_{N} U_{N}^{H},$ 再使角度遍历360度，按照 $P_{M U S I C} = 1 / (a (θ) U_{N} U_{N}^{H} a (θ))$ 来计算谱函数P_MUSIC，，其中a(θ)为通过寻求峰值来得到波达方向的估计值θ。Among them, R is the covariance matrix, and then perform eigenvalue decomposition on it R=UΣU ^H , according to the order of the size of the eigenvalues, the eigenvector corresponding to the largest eigenvalue equal to the number of signals K is regarded as the signal subspace, and the remaining The following (MK) eigenvalues correspond to eigenvectors as noise subspaces, then $R = u_{S} Σ_{S} u_{S}^{h} + u_{N} Σ_{N} u_{N}^{h},$ Then make the angle traverse 360 degrees, according to $P_{m u S I C} = 1 / (a (θ) u_{N} u_{N}^{h} a (θ))$ to calculate the spectral function P _MUSIC , where a(θ) is the estimated value θ of the direction of arrival obtained by seeking the peak value.

本发明具有以下优点：基于现有WiFi设备进行AOA的估计，本发明利用接收到的OFDM正交子载波的信道状态信息(ChannelStateInformation，CSI)矩阵对多径信号到达阵列天线的飞行时间(TimeofFlight，TOF)和AOA进行联合估计，相比于传统的WLAN测角技术，本发明无需更改任何的收发设备，能够达到相同的测角精度，估计角度的平均误差在5度左右，能够远远满足室内多径信号的角度估计的需求，具有较高的推广价值。The present invention has the following advantages: the estimation of AOA is carried out based on the existing WiFi equipment, and the present invention utilizes the channel state information (ChannelStateInformation, CSI) matrix of the OFDM orthogonal subcarriers received to arrive at the flight time (TimeofFlight, CSI) matrix of the multipath signal to the array antenna TOF) and AOA for joint estimation. Compared with the traditional WLAN angle measurement technology, the present invention does not need to change any transceiver equipment, and can achieve the same angle measurement accuracy. The average error of the estimated angle is about 5 degrees, which can far meet indoor The demand for angle estimation of multipath signals has high promotional value.

附图说明Description of drawings

图1本发明的原理框图。Figure 1 is a schematic block diagram of the present invention.

具体实施方式detailed description

下面结合附图和具体实施例对本发明进行详细说明。The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

如图1所示，系统整体原理框图。As shown in Figure 1, the overall functional block diagram of the system.

(1)在WLAN发射网卡上产生原始数据，经过OFDM基带调制后通过射频发出。(1) Raw data is generated on the WLAN transmitting network card, and sent out through radio frequency after being modulated by OFDM baseband.

(2)在WLAN接收网卡上接收经过多径信道的射频信号，然后进行下变频得到基带信号。(2) Receive the radio frequency signal through the multipath channel on the WLAN receiving network card, and then perform down-conversion to obtain the baseband signal.

(3)获取基带信号后，首先对基带信号进行解调，获取接收信号的CSI信息。(3) After the baseband signal is acquired, the baseband signal is first demodulated to acquire the CSI information of the received signal.

(4)利用获取到的CSI信息构建观测方程，利用TOF与AOA联合估计算法估计AOA信息。(4) Construct the observation equation by using the obtained CSI information, and use the joint estimation algorithm of TOF and AOA to estimate the AOA information.

步骤1、配置无线局域网：Step 1. Configure WLAN:

中心频点，带宽，阵列天线个数(2个)，天线间距，正交调制子载波个数，子载波频率间距，信噪比等。Center frequency point, bandwidth, number of array antennas (2), antenna spacing, number of quadrature modulation subcarriers, subcarrier frequency spacing, signal-to-noise ratio, etc.

步骤2、采用多个子载波对原始数据进行正交调制。Step 2, using multiple sub-carriers to perform quadrature modulation on the original data.

步骤3、估计每个子载波的信道状态信息：Step 3. Estimate the channel state information of each subcarrier:

假设所有子载波是正交的，即没有ICI，可以将N个子载波的训练符号表示成矩阵形式：Assuming that all subcarriers are orthogonal, that is, there is no ICI, the training symbols of N subcarriers can be expressed in matrix form:

其中，X[k]表示第k个子载波上的导频信号，满足E{X[k]}＝0，Var{X[k]}＝σ²,k＝0,1,2,…N-1。因为假设所有的子载波都是正交的，所以X是一个对角矩阵。给定k个载波的信道增益[H[0]H[1]…H[k-1]]^T，接收到的k个训练信号[Y[0]Y[1]…Y[k-1]]^T能够表示为：Among them, X[k] represents the pilot signal on the kth subcarrier, satisfying E{X[k]}=0, Var{X[k]}=σ ² , k=0,1,2,...N- 1. Since all subcarriers are assumed to be orthogonal, X is a diagonal matrix. Given the channel gains [H[0]H[1]…H[k-1]] ^T of k carriers, the received k training signals [Y[0]Y[1]…Y[k-1] ] ^T can be expressed as:

其中，H为信道向量，H＝[H[0],H[1],…,H[k-1]]^T；Z为噪声向量Z＝[Z[0],Z[1],…,Z[k-1]]^T，Y为接收的训练信号向量，Y＝[Y[0]Y[1]…Y[k-1]]^T，满足E{Z[k]}＝0，k＝0,1,…N-1。Wherein, H is a channel vector, H=[H[0], H[1],...,H[k-1]] ^T ; Z is a noise vector Z=[Z[0], Z[1],..., Z[k-1]] ^T , Y is the received training signal vector, Y=[Y[0]Y[1]…Y[k-1]] ^T , satisfying E{Z[k]}=0, k=0,1,...N-1.

为了得到信道估计LS信道估计法需要最小化下面的代价函数 In order to get the channel estimate The LS channel estimation method needs to minimize the following cost function

令代价函数关于的偏导数等于0，即：let the cost function about The partial derivative of is equal to 0, that is:

$\frac{\partial \partial J J ((\overset{^^}{H h}))}{\partial \partial \overset{^^}{H h}} = = - - 22 {(({X x}^{H h} Y Y))}^{* *} + + 22 (({X x}^{H h} X x \overset{^^}{H h})) = = 00;;$

然后可以得到由此得到LS信道估计的解为：Then you can get This leads to the solution of the LS channel estimation for:

令表示中的元素，k＝0,1,2,…,N-1。由无ICI的假设条件下可知X为对角矩阵，因此第k个子载波上的LS信号估计可以表示为：make express The elements in k=0,1,2,...,N-1. It can be known that X is a diagonal matrix under the assumption of no ICI, so the LS signal estimation on the kth subcarrier can be expressed as:

${\overset{^^}{H h}}_{L L S S} = = \frac{Y Y [[k k]]}{X x [[k k]]},, k k = = 00,, 11,, 22,, ... ... N N - - 11;;$

即在接收端接收到OFDM的CSI(ChannelStateInformation)信道信息矩阵CSI_matrix：That is, the CSI (ChannelStateInformation) channel information matrix CSI _matrix of OFDM is received at the receiving end:

其中，csi_i,j为第i根天线上第j个子载波的信道信息值。Among them, csi _i,j is the channel information value of the jth subcarrier on the ith antenna.

步骤4、采用空间平滑算法对接收到的CSI信道矩阵进行去相关处理：Step 4, using a spatial smoothing algorithm to decorrelate the received CSI channel matrix:

将等距线阵分成若干个相重叠的子阵列。若各子阵列的阵列流形相同(这一假设适用于等距线阵)，则子阵列协方差矩阵可以相加后平均取代原来意义上的阵列协方差矩阵R_s。将M元的等距线阵用滑动分成L个子阵，每个子阵有N个单元，其中N＝M-L+1。第l个前向子阵输出为：Divide an equidistant linear array into several overlapping subarrays. If the array manifolds of each subarray are the same (this assumption applies to equidistant linear arrays), the subarray covariance matrices can be added and averaged to replace the original array covariance matrix R _s . The equidistant linear array of M elements is divided into L sub-arrays by sliding, and each sub-array has N units, where N=M-L+1. The output of the lth forward subarray is:

${x x}_{l l}^{f f} ((t t)) = = {[[{x x}_{l l} ((t t)),, {x x}_{l l + + 11} ((t t)),, ... ...,, {x x}_{l l + + N N - - 11} ((t t))]]}^{T T} = = {A A}_{M m} {D D.}^{11 - - 11} s the s ((t t)) + + {n no}_{l l} ((t t)),, 11 \leq \leq l l \leq \leq L L;;$

其中s(t)为接收平面波的复振幅，n_l(t)为噪声向量，A_M为N×M维的方向矩阵，其列为N维的导向矢量a_M(θ_i)(i＝1,2,…,K)，where s(t) is the complex amplitude of the received plane wave, n _l (t) is the noise vector, A _M is the N×M dimensional direction matrix, and its column is the N-dimensional steering vector a _M (θ _i ) (i=1 ,2,...,K),

$D D. = = d d i i a a g g (({e e}^{j j \frac{22 π π d d}{λ λ} {sinθ sinθ}_{11}},, {e e}^{j j \frac{22 π π d d}{λ λ} {sinθ sinθ}_{22}},, ... ...,, {e e}^{j j \frac{22 π π d d}{λ λ} {sinθ sinθ}_{K K}}));;$

d为阵列天线之间的间隔，λ为接收信号的波长，θ_i为第i个信源的来波方向，所以，第l个前向子阵的协方差矩阵为：d is the interval between array antennas, λ is the wavelength of the received signal, and θ _i is the incoming wave direction of the i-th source, so the covariance matrix of the l-th forward sub-array is:

${R R}_{l l}^{f f} = = E E. [[{x x}_{l l}^{f f} ((t t)) {x x}_{l l}^{f f} {((t t))}^{H h}]] = = {A A}_{M m} {D D.}^{l l - - 11} {R R}_{S S} {(({D D.}^{l l - - 11}))}^{H h} {A A}_{M m}^{H h};;$

则前向空间平滑协方差矩阵R_f为：Then the forward spatial smoothing covariance matrix R _f is:

${R R}_{f f} = = \frac{11}{L L} {Σ Σ}_{l l = = 11}^{L L} {R R}_{l l}^{f f};;$

在此基础上，考察直线阵的倒序阵(按M,M-1,…,2,1顺序排列)。同理可以得到后向空间平滑的协方差矩阵R_b为：On this basis, investigate the reverse order of the linear array (arranged in the order of M, M-1,...,2,1). Similarly, the covariance matrix R _b for backward space smoothing can be obtained as:

${R R}_{b b} = = \frac{11}{L L} {Σ Σ}_{l l = = 11}^{L L} {R R}_{l l}^{b b};;$

其中，第l个后向子阵的协方差矩阵其实R_b就是R_f的共轭倒序阵，他们之间的这种关系就是常说的共轭倒序不变性。因此，前后向平滑协方差矩阵为：Among them, the covariance matrix of the lth backward subarray In fact, R _b is the conjugate reverse matrix of R _f , and the relationship between them is the so-called conjugate reverse invariance. Therefore, the forward and backward smoothing covariance matrix is:

$\overset{~ ~}{R R} = = \frac{11}{22} (({R R}_{f f} + + {R R}_{b b}));;$

共轭倒序不变性的优点在于可以增加子阵数目，但相对原阵列而言，阵列的有效孔径还是减少了，因为子阵列比原阵列小。尽管存在这一孔径损失，但它改变了基于天线阵列协方差矩阵的特征分解类AOA算法的局限性。The advantage of conjugate inversion invariance is that the number of sub-arrays can be increased, but compared with the original array, the effective aperture of the array is still reduced because the sub-arrays are smaller than the original array. Despite this aperture loss, it changes the limitations of eigendecomposition-like AOA algorithms based on the antenna array covariance matrix.

步骤5、利用去相关后的CSI信息对信号到达阵列天线的飞行时间和到达角度估计算法进行联合估计：Step 5. Use the de-correlated CSI information to jointly estimate the time-of-flight of the signal arriving at the array antenna and the angle-of-arrival estimation algorithm:

去相关后的CSI信息对信号到达阵列天线的飞行时间TOF(TimeofFlight)和到达角度估计算法进行联合估计定位时构造包含AOA和TOF信息的二维方向矩阵：The CSI information after decorrelation is used to jointly estimate the TOF (TimeofFlight) of the signal arriving at the array antenna and the angle of arrival estimation algorithm to construct a two-dimensional direction matrix containing AOA and TOF information:

其中是M*1的方向向量，θ_k是第k条路径的到达角，τ_k是第k条路径的飞行时间，最后建立了利用OFDM多载波信息对TOF和AOA进行联合估计的系统。in is the direction vector of M*1, θ _k is the angle of arrival of the k-th path, and τ _k is the flight time of the k-th path. Finally, a system for joint estimation of TOF and AOA using OFDM multi-carrier information is established.

AOA估计方法是利用MUSIC算法完成到达角度的估计。假设相邻两天线间距为d，当信号经过不同的传播路径后，设某条多径的信号一定的角度θ_k入射到接收天线阵列时，由于入射角度θ_k的存在，相邻两根天线上会引入和θ_k相关的路程差dsinθ_k，由该路程差造成的相位偏移为2πfdsinθ_k/c。如果以第一根天线为参考，设由第v条路径到达天线1的波前信号为s_k(t)，则第i个天线收到的第k条径的波前信号为：The AOA estimation method is to use the MUSIC algorithm to complete the estimation of the angle of arrival. Assuming that the distance between two adjacent antennas is d, when the signal passes through different propagation paths, when a certain multipath signal is incident on the receiving antenna array at a certain angle θ _k , due to the existence of the incident angle θ _k , the adjacent two antennas A path difference dsinθ _k related to θ _k will be introduced above, and the phase shift caused by this path difference is 2πfdsinθ _k /c. If the first antenna is taken as a reference, and the wavefront signal from the vth path to antenna 1 is set as s _k (t), then the wavefront signal of the kth path received by the ith antenna is:

s_i,k(t)＝a_ks_k(t)exp(-jw₀(i-1)dsinθ_k/c)；s _i,k (t)=a _k s _k (t)exp(-jw ₀ (i-1)dsinθ _k /c);

其中s_i,k(t)代表第i个天线收到的第k条路径的波前信号，a_k代表第k条路径的幅度衰减，w₀代表载波的角速度，d代表天线间隔，c代表光速。考虑到第i条天线的噪声的影响，假设环境中共有N条多径信号，那么第i个天线收到的波前信号是多条多径信号的总和，即：Where s _i,k (t) represents the wavefront signal of the k-th path received by the i-th antenna, a _k represents the amplitude attenuation of the k-th path, w ₀ represents the angular velocity of the carrier, d represents the antenna spacing, and c represents speed of light. Considering the influence of the noise of the i-th antenna, assuming that there are N multi-path signals in the environment, the wavefront signal received by the i-th antenna is the sum of multiple multi-path signals, namely:

${s the s}_{i i} ((t t)) = = {Σ Σ}_{k k = = 11}^{N N} {s the s}_{i i,, k k} ((t t)) = = {Σ Σ}_{k k = = 11}^{N N} {a a}_{k k} {s the s}_{k k} ((t t)) exp exp ((- - {jw jw}_{00} ((i i - - 11)) d d {sinθ sinθ}_{k k} / / c c)) + + {n no}_{i i} ((t t));;$

假定阵列天线数为M，各天线的噪声是均值为0，方差为σ²的平稳白噪声过程，并且噪声之间互不相关，同时噪声和信号之间不相关，对各天线上接收到的波前信号写成向量形式为：Assuming that the number of array antennas is M, the noise of each antenna is a stationary white noise process with a mean value of 0 and a variance of ^σ2 , and the noises are not correlated with each other, and the noise and the signal are not correlated at the same time. The wavefront signal is written in vector form as:

X(t)＝AS(t)+N(t)；X(t)=AS(t)+N(t);

其中，X(t)＝[x₁(t),x2(t),...,x_M(t)]^T为M维的接收数据向量，S(t)＝[S₁(t),S₂(t),...,S_N(t)]^T为N维信号向量，为M*N维方向矩阵。为M维方向向量，τ_k＝dsinθ_k/c，N(t)＝[n₁(t),n₂(t),...,n_M(t)]^T为M维噪声，其中N为多条路径的个数。Wherein, X(t)=[x ₁ (t), x2(t),...,x _M (t)] ^T is the received data vector of M dimension, S(t)=[S ₁ (t), S ₂ (t),...,S _N (t)] ^T is an N-dimensional signal vector, is an M*N dimensional direction matrix. is the M-dimensional direction vector, τ _k =dsinθ _k /c, N(t)=[n ₁ (t),n ₂ (t),...,n _M (t)] ^T is the M-dimensional noise, where N is the number of multiple paths.

MUSIC算法是利用接收信号X(t)的自相关矩阵的奇异值分解，将特种空间划分为相互正交信号子空间和噪声子空间，利用子空间的正交特性获得谱估计的结果。假设，X(t)的自相关矩阵为R_xx(t)，The MUSIC algorithm uses the singular value decomposition of the autocorrelation matrix of the received signal X(t), divides the special space into mutually orthogonal signal subspaces and noise subspaces, and uses the orthogonal characteristics of the subspaces to obtain the spectral estimation results. Suppose, the autocorrelation matrix of X(t) is R _xx (t),

R_xx(t)＝E[X(t)X^H(t)]＝R_s+R_noise；R _xx (t)=E[X(t)X ^H (t)]=R _s +R _noise ;

其中R_xx(t)为秩为M的M*M对称阵，R_s＝E[S(t)S^H(t)]为秩为N的N*N矩阵，R_noise为秩为M-N的矩阵。通过对R_xx(t)求特征值和特征向量，并按大小对特征值进行排序，取M-N个最小的特征值对应的特征向量构成噪声子空间，记为E_n＝{v_N+1,v_N+2,...,v_M}，由于噪声子空间和信号子空间正交，所以在信号方向θ_k上，显然有：Where R _xx (t) is an M*M symmetric matrix with rank M, R _s =E[S(t) ^SH (t)] is an N*N matrix with rank N, and R _noise is a matrix with rank MN . By calculating the eigenvalues and eigenvectors of R _xx (t), and sorting the eigenvalues by size, the eigenvectors corresponding to the MN smallest eigenvalues are taken to form a noise subspace, which is recorded as E _n ={v _N+1 , v _N+2 ,...,v _M }, since the noise subspace is orthogonal to the signal subspace, in the signal direction θ _k , obviously:

${E E.}_{N N}^{H h} \overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{k k})) = = 00;;$

为了实现N个多径信号的来波方向估计，通过连续改变θ值，利用上式进行谱峰搜索，获得N个θ对应的谱峰即为多径信号的来波方向：In order to realize the estimation of the direction of arrival of N multipath signals, by continuously changing the value of θ, use the above formula to search for spectral peaks, and obtain the spectral peaks corresponding to N θ, which is the direction of arrival of multipath signals:

${P P}_{m m u u s the s i i c c} ((θ θ)) = = \frac{11}{{\overset{&RightArrow; &Right Arrow;}{a a}}^{H h} ((θ θ)) {E E.}_{N N} {E E.}_{N N}^{H h} \overset{&RightArrow; &Right Arrow;}{a a} ((θ θ))} . .$

进一步，所述步骤3估计每个子载波的信道状态信息，采用最小二乘(LS)技术进行信道估计。Further, the step 3 estimates the channel state information of each subcarrier, and adopts the least square (LS) technique for channel estimation.

最小二乘技术是假设N-1个子载波是正交的，即没有ICI，接收到的训练信号为Y[k],k＝0,1,2,…,N-1对信道H的估计为最小化下面的代价函数：The least squares technique assumes that N-1 subcarriers are orthogonal, that is, there is no ICI, and the received training signal is Y[k], k=0,1,2,..., N-1 estimates the channel H as Minimize the following cost function:

$J J ((\overset{^^}{H h})) = = | | | | Y Y - - A A \overset{^^}{H h} | | {| |}^{22} = = {Y Y}^{H h} Y Y - - {Y Y}^{H h} X x \overset{^^}{H h} - - {\overset{^^}{H h}}^{H h} {X x}^{H h} Y Y + + {\overset{^^}{H h}}^{H h} {X x}^{H h} X x \overset{^^}{H h};;$

令上面的代价函数关于的偏导数等于0，得到LS信道估计的解为：Let the cost function above be about The partial derivative of is equal to 0, and the solution of LS channel estimation is obtained as:

令表示中的元素，k＝0,1,2,…,N-1。由无ICI的假设条件下可知X为对角矩阵，因此每个子载波上的LS信号估计可以表示为：make express The elements in k=0,1,2,...,N-1. It can be known that X is a diagonal matrix under the assumption of no ICI, so the LS signal estimation on each subcarrier can be expressed as:

${\overset{^^}{H h}}_{L L S S} = = \frac{Y Y [[k k]]}{X x [[k k]]},, k k = = 00,, 11,, 22,, ... ... N N - - 1. 1.$

进一步，所述步骤5去相关后的CSI信息对信号到达阵列天线的飞行时间TOF和到达角度估计算法进行联合估计定位时构造包含AOA和TOF信息的二维方向矩阵：Further, the CSI information after the de-correlation in step 5 performs joint estimation and positioning on the time-of-flight TOF and the angle-of-arrival estimation algorithm of the signal arriving at the array antenna to construct a two-dimensional direction matrix containing AOA and TOF information:

MUSIC算法是根据N个接收天线接收信号得到协方差矩阵的估计值，The MUSIC algorithm is to obtain the estimated value of the covariance matrix according to the signals received by N receiving antennas.

$R R = = \frac{11}{N N} {Σ Σ}_{n no = = 11}^{N N} x x ((n no)) {x x}^{H h} ((n no))$

本发明充分考虑了室内复杂的多径环境对角度估计造成的影响，采用该系统和方法，估计角度的平均误差在5度左右，能够远远满足室内多径信号的角度估计的需求。The present invention fully considers the influence of indoor complex multipath environment on angle estimation, adopts the system and method, and the average error of estimated angle is about 5 degrees, which can far meet the requirement of indoor multipath signal angle estimation.

Claims

1. a kind of estimation method based on the WLAN access point AOA of double antenna, it is characterized in that, may further comprise the steps:

Step 1. Configure wireless LAN;

Step 2, using multiple sub-carriers at the transmitting end to perform quadrature modulation on the original data;

Step 3. Estimate the channel state information of each subcarrier:

The CSI channel information matrix CSI _matrix of OFDM is received at the receiving end:

{CSI CSI}_{m m a a t t r r i i x x} = = [\begin{matrix} {csi csi}_{11,, 11} & {csi csi}_{11,, 22} & ... ... & {csi csi}_{11,, 1919} & {csi csi}_{11,, 2020} \\ {csi csi}_{22,, 11} & {csi csi}_{22,, 22} & ... ... & {csi csi}_{22,, 1919} & {csi csi}_{22,, 2020} \end{matrix}];;

Among them, csi _i,j is the channel information value of the jth subcarrier on the ith antenna;

Step 4, using a spatial smoothing algorithm to perform decorrelation processing on the received CSI channel matrix;

Step 5. Using the de-correlated CSI information to jointly estimate the time-of-flight (TOF) of the signal arriving at the array antenna and the angle-of-arrival estimation algorithm to obtain an estimated value of the direction of arrival.

2. the estimation method of the WLAN access point AOA based on dual antenna according to claim 1, is characterized in that: in described step 3, estimate the channel state information of each subcarrier, adopt least squares (LS) technique to carry out channel estimation;

The least squares technique assumes that N-1 subcarriers are orthogonal, that is, there is no ICI, and the received training signal is Y[k], k=0,1,2,..., N-1 estimates the channel H as cost function minimize:

J J ((\overset{^^}{H h})) = = | | | | Y Y - - X x \overset{^^}{H h} | | {| |}^{22} = = {Y Y}^{H h} Y Y - - {Y Y}^{H h} X x \overset{^^}{H h} - - {\overset{^^}{H h}}^{H h} {X x}^{H h} Y Y + + {\overset{^^}{H h}}^{H h} {X x}^{H h} X x \overset{^^}{H h};;

let the cost function about The partial derivative of is equal to 0, and the solution of LS channel estimation is obtained as:

{\overset{^^}{H h}}_{L L S S} = = (({X x}^{H h})) {X x}^{- - 11} {X x}^{H h} Y Y = = {X x}^{- - 11} Y Y;;

make express The elements in , k=0,1,2,...,N-1, under the assumption of no ISI, it can be known that X is a diagonal matrix, so the LS signal estimation on each subcarrier can be expressed as:

{\overset{^^}{H h}}_{L L S S} [[k k]] = = \frac{Y Y [[k k]]}{X x [[k k]]},, k k = = 00,, 11,, 22,, ... ... N N - - 1. 1.

3. the method for estimating the WLAN access point AOA based on dual antennas according to claim 1 or 2, characterized in that: in the step 5, the time-of-flight (TOF) and When the angle of arrival estimation algorithm performs joint estimation, a two-dimensional direction matrix containing AOA and TOF information is constructed:

A A = = [[\overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{11},, {τ τ}_{11})),, \overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{22},, {τ τ}_{22})),, ... ...,, \overset{&RightArrow; &Right Arrow;}{a a} (({θ θ}_{N N},, {τ τ}_{N N}))]];;

in is the direction vector of M*1, θ _k is the angle of arrival of the k-th path, τ _k is the flight time of the k-th path, and finally a system for joint estimation of TOF and AOA using OFDM multi-carrier information is established;

The AOA estimation method is to use the MUSIC algorithm to complete the estimation of the angle of arrival; the MUSIC algorithm is to obtain the estimated value of the covariance matrix according to N received signal vectors x(n);

R R = = \frac{11}{N N} {Σ Σ}_{n no = = 11}^{N N} x x ((n no)) {x x}^{H h} ((n no));;

Among them, R is the covariance matrix, and then perform eigenvalue decomposition R=UΣU ^H on it, according to the order of the size of the eigenvalue U, the eigenvector corresponding to the largest eigenvalue equal to the number of signals K is regarded as the signal subspace, and The remaining (MK) eigenvalues correspond to eigenvectors as noise subspaces, then

R = u_{S} Σ_{S} u_{S}^{h} + u_{N} Σ_{N} u_{N}^{h},

Then make the angle traverse 360 degrees, according to

P_{m u S I C} = 1 / (a (θ) u_{N} u_{N}^{h} a (θ))

to calculate the spectral function P _MUSIC , where a(θ) is the estimated value θ of the direction of arrival obtained by seeking the peak.