Channel tracking method for aerial multiple antenna communication system based on geometric prior information

Abstract

The application discloses a channel tracking method for an aerial multi-antenna communication system based on geometric prior information and belongs to the technical field of wireless communication. The method comprises the following steps: S1, constructing a ground-air channel model and a position transformation coordinate system; S2, initial ground-air channel estimation and reflector position estimation; and S3, geometric prior information assisted ground-air time-varying channel tracking. The aerial multi-antenna communication system channel tracking method based on geometric prior information is adopted, ground-air channel state information is tracked by using a small amount of pilots, ground-air channel angle prior information is constructed according to the prediction results of the position and attitude of an aircraft and the position of a reflector, low-dimensional multipath gain parameters of the ground-air channel are estimated, a high-dimensional channel matrix is reconstructed, the influence of flexible adjustment of the attitude of the aircraft on channel tracking is effectively coped with, and high-dimensional channel tracking is realized in a low pilot cost mode.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, and particularly relates to a channel tracking method for an aerial multi-antenna communication system based on geometric prior information. BACKGROUND

[0002] An aerial communication system is a key infrastructure for ensuring safe and efficient operation of air transportation. An aircraft flies at a height of more than 8000 meters in the cruising phase, and long-distance signal transmission causes severe path loss. Considering that multiple antennas can provide a large array gain, using multiple antenna technology can produce a directional beam to compensate for the low signal-to-noise ratio caused by path loss. However, the premise of fully utilizing multiple antenna technology is to obtain accurate channel state information. Due to the rapid change of the aircraft position and the flexible adjustment of the aircraft attitude, the ground-air high-dimensional channel changes rapidly, greatly increasing the difficulty of real-time acquisition of the channel state information at the ground station side. Therefore, the problem of real-time acquisition of the channel state information needs to be solved.

[0003] Existing work mainly studies the acquisition of channel state information under the condition of ground time variation, without considering the sparse multipath characteristics of the ground-air channel and the flexible change characteristics of the aircraft attitude. The ground channel acquisition scheme is not suitable for the high dynamic communication scenario of the aircraft. Specifically, on the one hand, the existing channel tracking method does not take into account the influence of the aircraft attitude on the ground-air channel, and the flexible adjustment of the aircraft attitude can cause the angle gain of the airspace channel to change suddenly. On the other hand, the existing channel tracking method does not effectively use the prior information such as the flight trajectory of the aircraft in the aerial communication scenario, and does not fully utilize the sparse multipath characteristics of the aerial communication channel. Therefore, the existing ground wireless communication channel tracking scheme cannot be directly applied in the aerial communication scenario.

[0004] In order to solve the above problems, the present application provides a channel tracking method for an aerial multi-antenna communication system based on geometric prior information, which effectively deals with the influence of the flexible adjustment of the aircraft attitude on the channel tracking. SUMMARY

[0005] The purpose of the present application is to provide a channel tracking method for an aerial multi-antenna communication system based on geometric prior information, which tracks the ground-air channel state information with a small amount of pilot, constructs the ground-air channel angle prior information according to the prediction results of the aircraft position and attitude and the position of the reflector, estimates the low-dimensional multipath gain parameters of the ground-air channel, and reconstructs the high-dimensional channel matrix, effectively dealing with the influence of the flexible adjustment of the aircraft attitude on the channel tracking, and realizing high-dimensional channel tracking in a low pilot cost manner.

[0006] To achieve the above purpose, the present application provides a channel tracking method for an aerial multi-antenna communication system based on geometric prior information, comprising the following steps:

[0007] S1, constructing the air-ground channel model and the position transformation coordinate system, as shown below:

[0008] S11, constructing the air-ground MIMO-OFDM channel model, using the aircraft position and attitude to describe the steering vector of the channel;

[0009] S12, constructing the reference coordinate system for position transformation, using the Euler angle sequence representing the aircraft attitude to establish a rotation matrix;

[0010] S13, deriving the steering vector of the aircraft side under the influence of the attitude according to the position and attitude integrated conversion formula;

[0011] S2, initial air-ground channel estimation and reflector position estimation, as shown below:

[0012] S21, in the initial channel estimation stage, using the initial high-dimensional channel estimation result to calculate the aircraft position, attitude and air-ground communication system scatterer position;

[0013] S22, in the channel tracking stage, using polynomial regression to predict the aircraft position, using quaternion method to represent the aircraft attitude, calculating the aircraft rotation angular velocity and predicting the attitude;

[0014] S3, geometric prior information aided air-ground time-varying channel tracking, as shown below:

[0015] S31, calculating the angle information of the air-ground channel according to the predicted spatial information, i.e. the geometric prior information for channel tracking;

[0016] S32, according to the Euler angle rotation sequence used to describe the attitude of the aircraft in step S1, using quaternion representation method to represent the aircraft attitude and perform attitude prediction;

[0017] S33, converting the high-dimensional channel estimation problem into a low-dimensional multipath parameter estimation problem, reducing the pilot overhead.

[0018] Preferably, in step S11 of constructing the air-ground MIMO-OFDM channel model, an LoS path and L NLoS paths are used to constitute the air-ground channel model, and it is assumed that the channel state information remains unchanged within a time block composed of consecutive time slots, then the space-frequency domain air-ground channel matrix on the xth subcarrier in the tth time block is as shown below:

[0019]

[0020] where, ρ R is the Rice factor, i.e. the power ratio between LoS path and NLoS path; is the steering vector of the ground station end, is the steering vector at the aircraft end; is the azimuth angle in the channel angle of arrival (AoA), is the elevation angle in the channel AoA; is the azimuth angle in the channel angle of departure (AoD), is the elevation angle in the channel AoD; denotes the complex gain of the LoS path, denotes the complex gain of the NLoS path, as follows:

[0021]

[0022] wherein, and are the large-scale fading and delay of the l-th NLoS path in the t-th time block, respectively.

[0023] Preferably, in step S12, the ground station geographical coordinate system, the aircraft geographical coordinate system, the aircraft body coordinate system and the aircraft antenna coordinate system are established in the reference coordinate system for position transformation;

[0024] The position of the aircraft in the ground station geographical coordinate system at the t-th time block is as follows:

[0025]

[0026] The attitude of the aircraft in the aircraft body coordinate system is usually described by a rotation sequence of Euler angles, i.e. a rotation sequence of the heading angle a, the pitch angle b and the roll angle g, and the attitude of the aircraft is as follows:

[0027] e t = (a t , b t , g t );

[0028] Therefore, the coordinate transformation matrix from the aircraft geographical coordinate system to the aircraft body coordinate system is as follows:

[0029]

[0030] wherein a is the heading angle, b is the pitch angle and g is the roll angle.

[0031] Preferably, in step S13, the position and attitude integrated conversion formula is defined as follows:

[0032]

[0033] Given the AoD of the LoS path of the air-ground channel, i.e. LAoD, the aircraft-side LoS path steering vector matrix is as follows:

[0034]

[0035] in,

[0036]

[0037] In the formula, and These represent the values ​​at antenna spacing d and carrier frequency λ, respectively. c In the case of along the coordinate axis x u y u Spatial frequency of direction; and Representing the x-axis respectively u y u The directional guide vector.

[0038] Preferably, in the initial channel estimation stage of step S21, the aircraft omnidirectionally transmits the i-th OFDM symbol in the q-th time slot. At the ground station, the signal received by the ground station on the x-th subcarrier As shown below:

[0039]

[0040] in, It is the aircraft-side omnidirectional precoding matrix. It is the signal transmitted on the xth subcarrier. It is additive white Gaussian noise, at the ground station side. It is a ground station simulation merging matrix.

[0041] Preferably, in the channel tracking stage of step S22, firstly, the azimuth and elevation angles at the reflector are calculated using the position transformation relationship:

[0042]

[0043] Then, the starting point is the actual position of the aircraft. Direction is The straight line is shown below:

[0044]

[0045] in, r∈{1,...,K} represents time;

[0046]

[0047] Finally, the position of the reflector is determined by the following system of equations:

[0048] A l p l =b l;

[0049] wherein,

[0050] Preferably, in step S31, the aircraft flight trajectory is fitted by polynomial fitting, and the aircraft position is predicted as follows:

[0051]

[0052] wherein, denotes the first derivative of f(x p ) with respect to x p ; x p denotes the x-axis of the p-coordinate system; T is the time block interval, and v is the velocity of the aircraft, and ||v|| can be calculated according to the Doppler shift and LAoA extracted in the high-dimensional air-ground channel estimation stage.

[0053] Preferably, in step S32, the aircraft attitude is represented by quaternion parameters as follows:

[0054]

[0055] wherein, i b , j b , k b are unit vectors on the x b , y b , z b axes, respectively;

[0056] The predicted actual position of the aircraft is converted to as follows:

[0057]

[0058] According to the reflector position and the predicted converted position the next time NLoS path is as follows:

[0059]

[0060] Preferably, in step S33, in estimating the low-dimensional channel parameters, the received signal on the xth subcarrier in the ith OFDM symbol in the qth time slot is as follows:

[0061]

[0062] wherein,

[0063] i.e.

[0064] The pilot overhead of the reconstructed channel fast prediction stage is as follows:

[0065]

[0066] Wherein, Q represents the total number of time slots;

[0067] Based on the predicted and estimated parameters of the reconstructed channel fast prediction stage Obtain the reconstructed high-dimensional channel matrix

[0068] Therefore, the present application adopts the above-mentioned channel tracking method for an aerial multi-antenna communication system based on geometric prior information, which has the following beneficial effects:

[0069] (1) Considering the flexible adjustment of the aircraft attitude, a reference coordinate system is established and a position and posture integrated conversion formula is proposed to describe the relationship between the spatial state information such as the aircraft attitude and position and the ground-air information angle information;

[0070] (2) Considering the regularity of the aircraft flight trajectory, the spatial state information such as the position and attitude of the aircraft is predicted in the channel tracking stage to further calculate the angle information in the ground-air high-dimensional channel;

[0071] (3) Considering the sparse multipath characteristics of the ground-air channel, the high-dimensional channel estimation is converted into low-dimensional sparse multipath parameter estimation in the channel tracking stage, so as to realize high-dimensional channel tracking with low pilot overhead.

[0072] The technical solutions of the present application will be further described in detail below through the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0073] Figure 1 It is a flow chart of the channel tracking method for an aerial multi-antenna communication system based on geometric prior information;

[0074] Figure 2 It is a schematic diagram of the reference coordinate system g-coordinate system, a-coordinate system and b-coordinate system;

[0075] Figure 3 It is a schematic diagram of the channel tracking frame structure based on geometric prior information;

[0076] Figure 4 It is a flow chart of the channel tracking algorithm based on geometric prior information;

[0077] Figure 5 It is a schematic diagram of the g-coordinate system and p-coordinate system in a two-dimensional plane. DETAILED DESCRIPTION

[0078] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the embodiments of the present application are further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present application and not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0079] As shown in the figure, the channel tracking method of an aerial multi-antenna communication system based on geometric prior information comprises the following steps: Figure 1

[0080] S1, constructing an air-ground channel model and a position transformation coordinate system;

[0081] S2, initial air-ground channel estimation and reflector position estimation;

[0082] S3, geometric prior information assisted air-ground time-varying channel tracking.

[0083] Embodiments

[0084] S1, constructing an air-ground channel model and a position transformation coordinate system.

[0085] The present application considers that there is an aircraft and a ground station in an aerial MIMO-OFDM communication system, and the aircraft and the ground station are both equipped with uniform planar antenna arrays, which are composed of K A =M A *N A and K B =M B *N B antennas respectively. The aircraft and the ground station both adopt a hybrid analog architecture to save hardware cost, that is, under the full connection architecture of the aircraft, there are radio frequency chains connected to K A antennas through phase shifters; under the full connection architecture of the ground station, there are radio frequency chains connected to K B antennas through phase shifters. The aerial communication system adopts an orthogonal frequency division multiplexing (OFDM) transmission scheme with X subcarriers and a sampling period of T S , and transmits N S independent data streams on each subcarrier.

[0086] S11, constructing an air-ground MIMO-OFDM channel model, and using the aircraft position and attitude to describe the steering vector of the channel;

[0087] ​Considering that in the high-altitude cruise scenario of an aircraft, there are a limited number of reflectors near the ground station and the aircraft, the present application uses a LoS path and L NLoS paths to form a ground-air channel model. Assuming that the channel state information remains unchanged within a time block composed of consecutive time slots, the space-frequency domain ground-air channel matrix on the xth subcarrier in the tth time block is As shown below:

[0088]

[0089] wherein ρ R is the Rice factor, i.e., the power ratio between the LoS path and the NLoS path.

[0090] and are the steering vectors at the ground station end and the aircraft end, respectively, is the azimuth angle in the channel angle of arrival (AoA), is the elevation angle in the channel AoA, is the azimuth angle in the channel angle of departure (AoD), is the elevation angle in the channel AoD, a B and the expressions of (·) and will be explained in detail later. represents the complex gain of the LoS path, represents the complex gain of the NLoS path, which is affected by the path gain, the path delay, and the Doppler frequency shift, as shown below:

[0091]

[0092] wherein and are the large-scale fading and delay of the lth NLoS path in the tth time block.

[0093] The present application utilizes the channel reciprocity of a time division multiplexing system, i.e., the channel parameters from the aircraft to the ground station are the same as the channel parameters from the ground station to the aircraft . In the signal transmission phase, taking the forward link as an example, the channel tracking method is explained.

[0094] S12, a reference coordinate system for position transformation is constructed, and a rotation matrix is established by using a sequence of Euler angles representing the attitude of the aircraft;

[0095] Since the AoA of the ground-air channel is determined by the positions of the aircraft and the reflectors, and the AoD of the ground-air channel is determined by the positions of the aircraft, the positions of the reflectors, and the attitude of the aircraft, in order to accurately describe the relationship between the positions and the attitude of the aircraft and the AoA and the AoD of the channel, as shown in Figure 2 , four reference coordinate systems are established:

[0096] • Ground station geographic coordinate system (g-coordinate system): Its origin is the centroid of the ground station array antenna, and its x-axis (x... g ) and y-axis (y g Aligned with north and west respectively, z-axis (z g Perpendicular to the ground and pointing upwards;

[0097] • Aircraft geographic coordinate system (b-coordinate system): Its origin is the aircraft's center of gravity, and its x-axis (x... b ), y-axis (y b ) and z-axis (z b Parallel to x g y g and z g ;

[0098] • Aircraft body coordinate system (a-coordinate system): Its center of gravity is the aircraft's center of gravity, and its x-axis (x... a ) and y-axis (y a ) respectively aligned with the starboard wing and nose of the aircraft, z-axis (z a Perpendicular to x a and y a Zhang Cheng's plane points downwards;

[0099] • Aircraft antenna coordinate system (u-coordinate system): Its origin is the centroid of the aircraft array antenna, and its x-axis (x... u ) and y-axis (y u ) respectively aligned with the row and column arrangement directions of the aircraft array antenna elements, z-axis (z u Perpendicular to x u and y u Zhang Cheng's plane points downwards.

[0100] Given the four coordinate systems above, the ground station's position in the g-coordinate system is fixed at (0,0,0). In the t-th time block, the aircraft's position in the g-coordinate system, i.e., its actual position, is as follows:

[0101]

[0102] In the α-coordinate system, the attitude of an aircraft is usually described by a sequence of Euler angle rotations, namely a sequence of rotations of heading angle α, pitch angle β, and roll angle γ. The attitude of the aircraft is as follows:

[0103] e t =(α t ,β t ,γ t (4);

[0104] Given the Euler angle coordinates of the aircraft, the coordinate transformation matrix from the b-coordinate system to the a-coordinate system is as follows:

[0105]

[0106] S13, deriving the guidance vector of the aircraft side under the influence of the attitude according to the integrated pose conversion formula;

[0107] Taking the LoS path as an example, in order to better describe the AoD (LAoD) of the LoS path of the air-ground channel and its relationship with the position and attitude of the aircraft, the integrated pose conversion formula is given, which is defined as follows:

[0108]

[0109] Specifically, the LAoA of the air-ground channel is equal in the following two cases:

[0110] Case one: when the position coordinates of the aircraft are and the attitude angle coordinates are e t =(α t ,β t ,γ t );

[0111] Case two: when the position coordinates of the aircraft are and the attitude angle coordinates are e t =(0,0,0).

[0112] On this basis, the guidance vector of the ground station end and the aircraft end is derived as follows.

[0113] The azimuth angle and the elevation angle of the LAoA of the air-ground channel can be expressed as a function of the actual position of the aircraft, as follows:

[0114]

[0115] At the tthtime block, according to the position conversion relationship, is converted to Further, the azimuth angle and the elevation angle of the LAoD of the air-ground channel can be expressed as a function of the converted aircraft position, as follows:

[0116]

[0117] By replacing the position of the ground station with the position of the reflector or replacing the position of the reflector with the position of the aircraft, the departure angle or the arrival angle of other reflection paths can also be obtained by a similar method.

[0118] Given the LAoD, the LoS path guidance vector matrix of the aircraft side is as follows:

[0119]

[0120] where, and denote the spatial frequencies along the coordinate axes x c , y u , respectively, for the case of antenna spacing d, carrier frequency u

[0121] and denote the steering vectors along the coordinate axes x u , y u , respectively, as follows:

[0122]

[0123] The steering vectors of the NLoS paths of the ground-air channel and the steering vectors of the LoS and NLoS paths have similar expressions.

[0124] S2, Initial ground-air channel estimation and scatterer position estimation.

[0125] As shown in Figure 3 , in the channel tracking frame structure, there are two stages: initial channel estimation and channel tracking.

[0126] i) Initial channel estimation stage: In this stage, high-dimensional channel estimation is first performed, and the ground-air channel angle information and the actual position of the aircraft are extracted according to the estimation results, which are used to calculate the attitude of the aircraft.

[0127] ii) Channel tracking stage: The position and attitude of the aircraft are predicted, the ground-air channel angle information is calculated, and it is used as geometric prior information to further estimate the low-dimensional multipath gain parameters to reconstruct the ground-air high-dimensional channel.

[0128] S21, In the initial channel estimation stage, the position, attitude of the aircraft and the position of the scatterer of the ground-air communication system are calculated using the initial high-dimensional channel estimation results.

[0129] In the initial channel estimation stage, it is assumed that the ground station needs to receive pilot symbols using G time slots, and each time slot contains T P OFDM symbols. The time block index t is omitted, and the aircraft transmits the i-th OFDM symbol to the ground station in the q-th time slot, and the received signal of the ground station on the x-th subcarrier is as follows:

[0130]

[0131] where, ​​is an aircraft-side omni-directional precoding matrix, is a transmitted signal on the xth subcarrier, is an additive white Gaussian noise. is an analog precoding matrix, is a digital precoding matrix.

[0132] on the ground station side, is a ground station analog combining matrix, which integrates pilot signals received in G time slots, and through existing channel estimation algorithms and channel parameter extraction algorithms, the AoA and AoD of the channel can be obtained.

[0133] In order to explain the prediction method of the NLoS path parameters, the present application assumes that the positions of the reflection bodies of the multipath signals of the aircraft communication system do not change and are reflected only once in the channel tracking stage, and the positions of the reflection bodies are calculated under this condition. If the multipath signals are reflected multiple times, the positions of the "first hop" of the signals are calculated using the method in the present application.

[0134] S22, in the channel tracking stage, the aircraft position is predicted using polynomial regression, the aircraft attitude is represented using quaternion method, the aircraft angular velocity is calculated and the attitude is predicted;

[0135] In the channel tracking stage, the AoA of the NLoS path will not change, and the AoD of the NLoS path will depend on the reflection body position, the aircraft position and the aircraft attitude. According to the AoD of the NLoS path at K moments r∈{1,...,K} and the actual position of the aircraft The reflection body position p related to the lth NLoS path is calculated l =(x l ,y l ,z l ). Specifically, K reflection vectors are determined, the starting point of the vectors is the actual position of the aircraft, the direction of the vectors is the AoD of the NLoS path at multiple moments, and the intersection point of the K vectors is the reflection body position.

[0136] First, the azimuth and elevation angles at the reflection body (the azimuth and elevation angles are based on a coordinate system with the origin at the reflection body position, and the three coordinate axes are aligned with the coordinate axes x g ,y g and z g ) are calculated using the position conversion relationship, as follows:

[0137]

[0138] Then, the straight line with the starting point at the actual position of the aircraft and the direction of is as follows:

[0139]

[0140] in,

[0141]

[0142] Finally, the position of the reflector is determined by the following system of equations:

[0143] A l p l =b l (18);

[0144] in,

[0145] S3. Geometric prior information-assisted ground-space time-varying channel tracking, such as Figure 4 As shown.

[0146] S31. Calculate the angle information of the air-to-ground channel based on the predicted spatial information, which is the geometric prior information used for channel tracking.

[0147] In this invention, polynomial fitting is used to fit the aircraft's flight trajectory, thereby predicting the aircraft's position. This invention assumes that the aircraft moves along a two-dimensional curve at a constant speed over a period of time. Under this assumption, the actual position of the aircraft in time block t+1 is... Located in A defined two-dimensional plane Above. (As shown) Figure 5 As shown, in order to better describe the two-dimensional plane Constructing the aircraft trajectory coordinate system (p-frame):

[0148] • The aircraft trajectory coordinate system (p-coordinate system) has its origin at the aircraft's position in time block t. Its x-axis (x p )and Orientation alignment, y-axis (y p Perpendicular to x p .

[0149] Given the actual position of an aircraft in three-dimensional space Transform it into a two-dimensional plane The position above As shown below:

[0150]

[0151] in, The starting point is The destination is A three-dimensional vector. is the distance between , is the angle between .

[0152] Given , the aircraft trajectory f(x p ) can be calculated by polynomial fitting, the predicted position is shown as follows:

[0153]

[0154] where denotes the first derivative of f(x p ) with respect to x p , T is the time block interval, v is the speed of the aircraft, and ||v|| can be calculated according to the Doppler shift and LAoA extracted in the high-dimensional geo-air channel estimation stage.

[0155] To obtain the actual position of the aircraft in three-dimensional space , the starting point is , and the direction is , the unit vector, as shown below:

[0156]

[0157] where is a two-dimensional vector with the starting point and the ending point . is the angle between and , is the angle between and .

[0158]

[0159] where

[0160] The position of the aircraft in the two-dimensional plane is converted to the actual position in three-dimensional space as shown below:

[0161]

[0162] where is the distance between and .​

[0163] In the channel tracking phase, the actual position of the aircraft in step S1 and the azimuth and elevation angles of the air-ground channel LAoD The converted aircraft position is as follows:

[0164]

[0165] Further, e t may be derived as follows:

[0166]

[0167] wherein denotes the pseudo-inverse, [A] i,j denotes the element of the matrix A in the i-th row and j-th column.

[0168] S32, according to the Euler angle rotation sequence in step S1 to describe the attitude of the aircraft, using the quaternion representation method to represent the aircraft attitude and attitude prediction again;

[0169] In step S1, the attitude of the aircraft is described by using the Euler angle rotation sequence, in order to further efficiently predict the attitude, the quaternion representation method with small calculation amount and high precision is used to represent the aircraft attitude and predict the attitude, according to the quaternion parameters of the last time and the current time, the aircraft rotation angular velocity is calculated, and the next time aircraft attitude is predicted, finally the predicted quaternion parameter result is converted into Euler angle parameter. The attitude of the aircraft can be represented by the quaternion parameter, as follows:

[0170]

[0171] wherein i b , j b , k b represent the unit vectors on the x b , y b , z b axis respectively. The rotation angular velocity ω b of the aircraft on the three axes x b , y b , z t may be represented as The Euler angle parameter of the aircraft can be converted into the quaternion parameter, and the conversion relationship is as follows:

[0172]

[0173] According to the quaternion parameters in the two time blocks, the angular velocity of the aircraft can be calculated, and the predicted quaternion parameters are as follows:

[0174]

[0175] It is worth noting that the angular velocity at the current time is calculated from the attitude at the current time and the attitude at the previous time, assuming that the angular velocity of the aircraft does not change over time.

[0176] Finally, the Euler angle parameters can be expressed as a function of the quaternion parameters as follows:

[0177]

[0178] The final predicted actual position of the aircraft is converted to:

[0179]

[0180] According to the reflector position and the predicted converted position the next time NLoS path is as follows:

[0181]

[0182]

[0183] S33, the high-dimensional channel estimation problem is converted into a low-dimensional multipath parameter estimation problem, reducing the pilot overhead;

[0184] Combined with the geometric prior information, i.e. AoA and AoD, the channel multipath complex gain is estimated When estimating the complex gain coefficient, the dimension is the number of physical paths, i.e. L+1. Therefore, compared with estimating the high-dimensional channel matrix, the computational complexity is significantly reduced. When estimating the low-dimensional channel parameters, the time block index value is omitted, and the received signal on the xth subcarrier of the ith OFDM symbol in the qth time slot is as follows:

[0185]

[0186] where,

[0187] i.e.

[0188] Therefore, equation (40) can be rewritten as:

[0189]

[0190] where,

[0191] Integrate the pilot signals received in Q time slots,

[0192]

[0193] The integrated pilot signals are as follows:

[0194]

[0195] Wherein, And

[0196] According to formula (42), the least square solution is obtained

[0197] In order to ensure the channel estimation performance, the pilot overhead of the reconstructed channel fast prediction stage is as follows:

[0198]

[0199] Finally, the parameters predicted and estimated in the reconstructed channel fast prediction stage

[0200] are brought into formula (1), and the high-dimensional channel matrix is reconstructed

[0201] Therefore, the present application adopts the above-mentioned channel tracking method of the aerial multi-antenna communication system based on geometric prior information, tracks the ground-air channel state information by using a small amount of pilots, constructs the ground-air channel angle prior information according to the prediction results of the aircraft position and attitude and the reflector position, estimates the low-dimensional multipath gain parameters of the ground-air channel, and reconstructs the high-dimensional channel matrix, effectively deals with the influence of the flexible adjustment of the aircraft attitude on the channel tracking, and realizes the high-dimensional channel tracking in a low pilot overhead mode.

[0202] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application but not to limit it, although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that: it can still modify or equivalently replace the technical solutions of the present application, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present application.

Claims

1. A channel tracking method for an airborne multi-antenna communication system based on geometric prior information, characterized in that, Includes the following steps: S1. Construct the ground-to-air channel model and the position transformation coordinate system, as shown below: S11. Construct a ground-to-air MIMO-OFDM channel model and use the aircraft's position and attitude to describe the channel's steering vector; S12. Construct a reference coordinate system for position transformation and establish a rotation matrix using the Euler angle sequence that characterizes the aircraft attitude. S13. Derive the guidance vector on the aircraft side under attitude influence based on the attitude integration conversion formula; S2, Initial Ground-to-Air Channel Estimation and Reflector Location Estimation, are shown below: S21. In the initial channel estimation stage, the aircraft position, attitude and the position of the ground-to-air communication system scatterer are calculated using the initial high-dimensional channel estimation results. S22. During the channel tracking phase, multinomial regression is used to predict the aircraft position, quaternion method is used to represent the aircraft attitude, the aircraft rotation angular velocity is calculated and the attitude is predicted. S3. Geometric prior information-assisted ground-space time-varying channel tracking, as shown below: S31. Calculate the angle information of the air-to-ground channel based on the predicted spatial information, which is the geometric prior information used for channel tracking. S32. Based on the Euler angle rotation sequence used in step S1 to describe the aircraft's attitude, the aircraft's attitude is re-represented using quaternion representation and attitude prediction is performed. S33. Transform the high-dimensional channel estimation problem into a low-dimensional multipath parameter estimation problem to reduce pilot overhead.

2. The channel tracking method for an airborne multi-antenna communication system based on geometric prior information according to claim 1, characterized in that, In S12, in constructing the reference coordinate system for position transformation, a ground station geographic coordinate system, an aircraft geographic coordinate system, an aircraft body coordinate system, and an aircraft antenna coordinate system are established. In the The aircraft's location in the ground station's geographic coordinate system is shown below for each time block: ； In the aircraft's body coordinate system, the aircraft's attitude is usually described by a sequence of Euler angle rotations, i.e., the heading angle. Pitch angle and roll angle The rotation sequence gives the aircraft's attitude as follows: ； Therefore, the coordinate transformation matrix from the aircraft geographic coordinate system to the aircraft body coordinate system is as follows: ； in, For heading angle, The pitch angle, For roll angle; In S13, the pose-integrated transformation formula is defined as follows: ； Given the AoD (Average Direction of Distance) of the air-to-ground channel LosS path, i.e., LAoD, the aircraft-side LosS path steering vector matrix is ​​as follows: ； in, This represents the aircraft-side Loss-of-Support (LoS) directional steering vector matrix; This represents the azimuth angle in the departure angle of the air-to-ground channel; This represents the pitch angle in the departure angle of the air-to-ground channel; ； ； ； ； In the formula, and These respectively represent the antenna spacing as carrier frequency In the case of coordinate axes along the aircraft antenna coordinate system , Spatial frequency of direction; and Representing along the coordinate axes , The directional guide vector; This indicates the number of array elements in the horizontal direction of the area array antenna; This indicates the number of array elements in the vertical direction of the area array antenna; In step S22, during the channel tracking phase, firstly, the azimuth and elevation angles at the reflector are calculated using the position transformation relationship: ； ； Then, the starting point is the actual position of the aircraft. , direction is The straight line is shown below: ； in, , Indicates time; ； Finally, the position of the reflector is determined by the following system of equations: ； in, , ; When estimating low-dimensional channel parameters by incorporating geometric prior information, in the th... The first time slot, the first The first OFDM symbol The received signals on each subcarrier are shown below: ； in, , , ,Right now ; This indicates the guidance vector at the ground station end where the time block index value is omitted; Indicates omitting time block index values The guidance vector at the ground station end of each NLoS path; In S33, the pilot overhead for the fast prediction stage of the reconstructed channel is as follows: ； in, This represents the total number of time slots. Indicates the number of OFDM symbols. Indicates the number of radio frequency links at the ground station; Based on the parameters predicted and estimated during the fast prediction stage of the reconstructed channel Obtain the reconstructed high-dimensional channel matrix ; in, This represents the estimated channel multipath complex gain; Indicates the first Estimate the complex gain of the channel Loss path at each time step; Indicates the first Estimate the complex gain of the NLoS path of the channel at each time step; Indicates the number of NLoS paths; Indicates the first NLoS path.

3. The channel tracking method for an airborne multi-antenna communication system based on geometric prior information according to claim 2, characterized in that, In step S11, when constructing the ground-to-air MIMO-OFDM channel model, a Loss path and A ground-to-air channel model is formed by NLoS paths. Assuming the channel state information remains unchanged within a time block composed of consecutive time slots, then the... The first time block The spatial frequency domain ground-to-space channel matrix on each subcarrier As shown below: ； in, It is the Rice factor, which is the power ratio between the LoS path and the NLoS path; It is the guide vector at the ground station end. It is the guidance vector at the aircraft end; , Both are azimuth angles in the channel angle of arrival (AoA). , All are elevation angles in channel AoA; , All are azimuth angles in the channel departure angle (AoD). , All are elevation angles in the channel AoD; Indicates the first The complex gain of the Loss path at each moment Indicates the first The complex gain of the NLoS path at each time point is shown below: ； in, and They are the first The first time block Large-scale fading and delay of NLoS paths.

4. The channel tracking method for an airborne multi-antenna communication system based on geometric prior information according to claim 3, characterized in that, In step S21, the initial channel estimation phase, the aircraft in the... The omnidirectional transmission in the [number] time slot OFDM symbols Upon arrival at the ground station, the ground station is at the... Received signal on each subcarrier As shown below: ； in, It is the aircraft-side omnidirectional precoding matrix. It is in the Transmitted signals on each subcarrier It is additive white Gaussian noise, at the ground station side. It is a ground station simulation merging matrix. This refers to the number of radio frequencies used in aircraft.

5. The channel tracking method for an airborne multi-antenna communication system based on geometric prior information according to claim 4, characterized in that, In step S22, a multinomial fitting method is used to fit the aircraft's flight trajectory and predict the aircraft's position, as shown below: ； in, express about The first derivative, Represents the x-axis of the p-coordinate system; It is the time block interval. It is the speed of the aircraft. The Doppler shift and LAoA extracted during the high-dimensional air-to-ground channel estimation stage can be used for calculation.

6. The channel tracking method for an airborne multi-antenna communication system based on geometric prior information according to claim 5, characterized in that, In step S32, quaternion parameters are used to represent the aircraft attitude, as shown below: ； in, , , They represent , , Unit vector on the axis, , , These are the x-axis, y-axis, and z-axis of the aircraft's geographic coordinate system, respectively. , , and Represents quaternion parameters; Predicted actual aircraft position Converted to As shown below: ； Based on the position of the reflector and the predicted converted position The AoD of the NLoS path at the next moment As shown below: ； 。

Images