CN116566450B

CN116566450B - Beam control algorithm implementation method based on ZYNQ

Info

Publication number: CN116566450B
Application number: CN202310833194.2A
Authority: CN
Inventors: 王伟; 王洪全; 李力力; 高远洪
Original assignee: Chengdu Huaxing Earth Technology Co ltd
Current assignee: Chengdu Huaxing Earth Technology Co ltd
Priority date: 2023-07-10
Filing date: 2023-07-10
Publication date: 2023-09-12
Anticipated expiration: 2043-07-10
Also published as: CN116566450A

Abstract

The application discloses a method for realizing a beam control algorithm based on ZYNQ in the technical field of wireless communication, which comprises the following steps: connecting the output end of PS with the input end of PL by adopting an AXI interface, and connecting the PS with the PL to form ZYNQ; step two: the PS receives and analyzes the external communication instruction; step three: transmitting the PS-resolved information to the PL through an AXI interface; step four: PL performs traversal calculations and adopts a pipeline design. The application reduces the development difficulty, shortens the development period, releases PL hardware resources and ensures the algorithm to realize time indexes.

Description

Beam control algorithm implementation method based on ZYNQ

Technical Field

The application belongs to the technical field of wireless communication, and particularly relates to a beam control algorithm implementation method based on ZYNQ.

Background

The beam steering algorithm is an algorithm that maximizes or minimizes power in a main direction when transmitting and receiving radio signals using a signal processing technique. The method aims to improve the signal strength and quality in a communication system and reduce energy consumption and multipath interference. Common beam steering algorithms include minimum mean square error (LMS) algorithms, kalman Filter (KF) algorithms, maximum Likelihood (ML) estimation algorithms, and the like.

In phased array antenna technology, a control chip (commonly used FPGA) needs to complete a beam resolving algorithm, calculates amplitude and phase values of each channel in an antenna array according to an input whole beam pointing instruction, and controls a phased array antenna amplitude-phase chip according to the calculated amplitude-phase values, so as to form a required space beam.

All the processes of angle preprocessing, angle correction, coordinate system conversion, channel amplitude and phase traversal calculation and the like in the conventional beam calculation algorithm are completed by using the FPGA, and pipeline design is needed in the channel amplitude and phase traversal calculation process so as to exert the advantages of FPGA parallel calculation and further meet the beam switching time requirement.

For example, chinese patent, publication No.: the application discloses a high-speed beam control method in a dynamic millimeter wave communication scene based on an FPGA (field programmable gate array), which comprises the following steps: initializing a wave beam at a receiving end of a communication system through full codebook scanning, and establishing an initial communication wave beam; establishing data communication between a beam control module and a baseband processing FPGA, and triggering a beam tracking process and carrying out beam measurement; establishing a beam training state machine to realize a beam tracking algorithm model; the antenna control information is transmitted through the SPI high-speed serial port to control the antenna by the beam control module; judging the beam quality after each beam training is finished, and if the beam selection is judged to be failed, carrying out beam recovery; and after the beam training is finished, selecting the optimal transmission beam for data transmission, and waiting for a training trigger signal of the next period. The application realizes the beam switching interval of microsecond level by defining the SPI high-speed mode writing protocol control antenna.

But in actual use, since all the computation processes are implemented in an FPGA. Although the FPGA has the advantage of parallel processing, the FPGA has the advantage only in the step of 'full-channel traversal calculation', and other steps are all disposable processes and have no advantage; due to the chip characteristics and fixed point computing characteristics of the FPGA, when complex mathematical algorithm operation is completed, compared with processors such as ARM, the FPGA has the disadvantages of more difficult development, longer development period, more occupied internal resources of the chip and the like.

And ZYNQ is a device series deduced by Sailingsi, and adopts an XilinxFPGA architecture based on an ARMCortex-A9 processor. This design fully combines high performance processing and programmable logic technology, allowing engineers to develop more easily complex system-level applications. The ZYNQ chip integrates a Processor Subsystem (PS) and programming logic resources (PL) and has the functions of a high-speed peripheral interface, a memory controller and the like. The ZYNQ has the advantages of low power consumption, high performance, strong flexibility, easy use, low cost and the like, and is widely applied to the fields of communication, industrial control, automobile electronics, video image processing and the like. Therefore, the application provides a beam control algorithm implementation method based on ZYNQ.

Disclosure of Invention

The application provides a ZYNQ-based beam control algorithm implementation method, which reduces development difficulty, shortens development period, releases PL hardware resources and ensures algorithm implementation time indexes.

In order to achieve the above object, the technical scheme of the present application is as follows: a method for realizing a beam control algorithm based on ZYNQ,

step one: and connecting the output end of the PS with the input end of the PL by adopting an AXI interface, and connecting the PS with the PL to form ZYNQ.

Step two: the PS receives, parses and pre-calculates external communication instructions.

Step three: and transmitting the PS-parsed information to the PL through an AXI interface.

Step four: PL performs traversal calculations and adopts a pipeline design.

After the scheme is adopted, the following beneficial effects are realized: the main meaning is that the quick and stable data interaction between the PS and the PL is realized by means of an AXI interface between the PS and the PL, so that the efficient information processing is realized.

Specifically, the AXI interface connection manner in the first step can make the data transmission between PS and PL more convenient and reliable; in the second step, the PS can parse and process various information according to the internal algorithm and the external communication instruction, and these information can be effectively transferred to the PL through the corresponding bus protocol based on the AXI interface; in the third step, through an AXI interface, PL can directly acquire various control signals, calculation results, state updating and other information generated by PS, and the information is used as the reference of self calculation, control and feedback, so that the processing efficiency and accuracy are improved; the pipeline design in the fourth step can realize multi-stage parallel processing, namely, data is divided into a plurality of stages, and different calculation tasks are independently and parallelly executed in each stage, so that the processing speed and the performance of the whole system are further improved.

The step flow of the application can reduce the system delay and furthest improve the response speed of the system: the problems of data errors and loss caused by complicated data transmission and the like can be avoided; the parallel processing efficiency and processing capacity of the system can be improved by utilizing the high computing performance of PL.

In a word, the implementation method of the beam control algorithm based on ZYNQ can reduce development difficulty, shorten development period, release PL hardware resources and ensure algorithm implementation time index.

Further, the PS analysis in the second step is: angle preprocessing, coordinate system conversion and angle correction.

Principle and beneficial effect: the specific principle of the PS is that after the external communication instruction is received, the PS performs angle preprocessing on data first, and key parameters such as azimuth angle, pitch angle and the like of a target object in an original coordinate system are identified. Then, the original data is converted from a motion coordinate system generated by the sensor into a fixed coordinate system or other needed coordinate systems by using a coordinate system conversion model, so that the influence caused by motion is eliminated, and more accurate target position information is obtained. Finally, the angle correction algorithm is utilized to adjust and fine tune the azimuth angle, pitch angle and other parameters of the target, and the accuracy and stability of control signals and data processing are improved.

The preprocessing and analyzing mode of PS can improve the data processing efficiency and the signal analyzing precision; meanwhile, the data analysis and processing time can be shortened, the calculation error is reduced, the system performance and stability are improved, and the reliability and the practicability of the equipment are further improved.

Further, the coordinate system is converted into:

s=acosd(cosd(a)*cosd(b))；

p=atand(tand(b)/sind(a))；

wherein: a is azimuth angle, b is pitch angle; s is the off-axis angle, p is the rotation angle, d is the unit of a, b in brackets.

Principle and beneficial effect: in the coordinate system conversion, s and p are off-axis angles and rotation angles, respectively, for describing a conversion relationship from one rectangular coordinate system to another rectangular coordinate system.

Let a and b be the azimuth and pitch angles in the original coordinate system, respectively, and s and p be the off-axis and rotational angles in the target coordinate system. Specifically, s represents the rotation angle of the target coordinate system with respect to the original coordinate system in the off-axis direction (i.e., the direction perpendicular to the original coordinate system), and p represents the rotation angle with respect to the original coordinate system after rotation about the axis.

The conversion mode of the coordinate system can accurately process and estimate the data according to the specific mathematical model and the characteristics of the coordinate system, thereby improving important technical indexes such as sensor control precision, positioning precision, navigation precision and the like; the data processing time can be shortened, the calculation error is reduced, the system performance and stability are improved, and the reliability and the practicability of the equipment are further improved; the method can adapt to transformation and conversion among different physical quantities, and is convenient for users to carry out quantization processing and comparative analysis in practical application, thereby exerting greater application value.

Further, the angle correction uses two-dimensional linear interpolation.

Principle and beneficial effect: the angle correction uses two-dimensional linear interpolation, i.e. given a set of angle correction parameters (typically including pitch angle and azimuth angle), the correction value at the target point is predicted from the input actual pitch angle and azimuth angle. The basic principle of the two-dimensional linear interpolation algorithm is that the relation among four angle correction parameters is estimated through bilinear interpolation, and the approximate value of the target point correction parameters is calculated according to the position and weight of control points around the target point.

The advantage of two-dimensional linear interpolation is that it enables more accurate estimation of correction values at different positions and reduces situations of poor equipment control accuracy due to sampling errors and insufficient data. Compared with the traditional interpolation algorithm, the two-dimensional linear interpolation is more in line with the actual physical process, can better process the nearly continuous control parameters, and can also carry out smooth processing on surrounding data in the estimation process, thereby improving the reliability and stability of the algorithm.

r_ab=u-floor(u/1.28)*1.28；

r_ac=v-floor(v/1.28)*1.28；

u_itab=u_b_in*r_ab+u_a_in*(1.28-r_ab)；

u_itcd=u_d_in*r_ab+u_c_in*(1.28-r_ab)；

u_it=u_itcd*r_ac+u_itab*(1.28-r_ac)；

u_out=u_it/1.28/1.28

U_FPGA=(u_out+u)*1.28

v_itab=v_b_in*r_ab+v_a_in*(1.28-r_ab)；

v_itcd=v_d_in*r_ab+v_c_in*(1.28-r_ab)；

v_it=v_itcd*r_ac+v_itab*(1.28-r_ac)；

v_out=v_it/1.28/1.28

V_FPGA=(v_out+v)*1.28

Where u and v are azimuth and pitch angles input by external instructions, u_a_in, u_b_in, u_c_in, u_d_in, v_a_in, v_b_in, v_c_in and v_d_in are 4 pairs of parameters required by two-dimensional linear interpolation.

The calculation method of the principle two-dimensional linear interpolation is mainly used for estimating the numerical value of the adjacent position according to the input azimuth angle and the pitch angle under the condition of giving four control points. Specifically, the input azimuth and pitch angles are converted into coordinates r_ab and r_ac within a square area of 1.28×1.28 by normalization and truncation processing. And then performing bilinear interpolation on the parameters according to the known 4 pairs of parameters to obtain an approximate value at the target point.

The principle of bilinear interpolation is that after a rectangular area where a target point is located is determined, weighted average is carried out on four vertexes in the area according to a certain weight, so that an approximate value of the target point is obtained. The calculation method is mainly applied to the fields of image processing, numerical analysis and the like, and is very useful in transforming or estimating a local area.

The algorithm can greatly improve the calculation speed and efficiency, and can adapt to different application scenes and precision requirements by adjusting the coefficients of the U_FPGA and the V_FPGA; meanwhile, the calculation cost and the energy consumption can be greatly reduced, and further more efficient algorithm realization and application are realized.

Further, the traversal calculation in the step four is:

C(dx,dy)=-2π(dx*sinθ*cosφ+dy*sinθ*sinφ)/λ

ux=MUX_F*sinθ*cosφ

vx=MUX_F*sinθ*sinφ

MUX_F=-(360/C)*f*2^22/128=-1.2*f*2^15

wherein, C is the light speed=300×10ζ9mm/s, (dx, dy) is the physical coordinate value of the antenna unit in the rectangular coordinate system, the unit is mm, θ, phi is the off-axis angle and rotation angle parameter of the antenna under the spherical coordinate system, λ is the signal wavelength, f is the signal frequency, and the unit is Ghz; the three variables ux, vx, mux_f are intermediate variables of the calculation process.

Principle and beneficial effect: the algorithm described above is one implementation of a traversal calculation for calculating the phase offset of an electromagnetic wave signal received or transmitted in a particular direction.

Specifically, the algorithm obtains the phase offset value of the corresponding position by calculating the weighted sum of the components (i.e., dx, dy) of the incident signal in two orthogonal directions and performing phase adjustment according to the incident angle.

Where mux_f is a constant coefficient, the value of which can be calculated by a given parameter, in relation to the frequency F. As the frequency changes, mux_f will also change accordingly, thereby affecting the final phase offset result. Therefore, the algorithm has the advantage that different MUX_F values can be set according to different frequencies, so that flexible design and optimization of antennas of different frequency bands are realized.

In addition, the 360/C value in the algorithm is related to the wavelength λ, which can be understood as a unit wavelength phase offset. Therefore, the denominator C in the algorithm represents a phase increment caused by the light speed, and the larger the value is, the smaller the phase difference corresponding to the unit distance is, so that finer phase control and adjustment can be realized, and the accuracy and sensitivity of the received or transmitted signal are improved.

Further, in the first step, the PS input interface is connected to an external command interface.

Principle and beneficial effect: the PS input end interface and the external instruction interface are connected, so that more flexible control and operation modes can be provided; the ZYNQ can be easily integrated into various applications, and the function expansion or performance optimization can be conveniently and rapidly carried out.

The PS input interface may be connected to various input devices such as USB, ethernet, SD card, HDMI, etc., and processed and controlled by an ARM processor. The external instruction interface can directly send instructions to the PS by using an external hardware module, so that dynamic adjustment and optimization of system-level application are realized, FPGA logic is not required to be reprogrammed, and development efficiency is greatly improved.

Further, in the first step, the PL output interface is connected to the amplitude-phase chip control interface.

Principle and beneficial effect: the PL output end interface is connected with the amplitude-phase chip control interface, so that output signals of the FPGA logic can be connected with and processed by the analog circuit; the process of converting the digital control signal into the analog control signal can be rapidly completed through the PL output end interface, so that the response speed of the system is improved; with the programmability of PL, the use of additional hardware such as analog controllers can be reduced, thereby reducing system costs.

Drawings

Fig. 1 is a schematic diagram of a hardware platform architecture of an embodiment of a beam control algorithm implementation method based on ZYNQ of the present application.

Fig. 2 is a schematic diagram of division of the implementation of the algorithm in the embodiment of the implementation method of the beam control algorithm based on ZYNQ.

Description of the embodiments

Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the application.

In the description of the present application, it should be understood that the terms "longitudinal," "transverse," "vertical," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate or are based on the orientation or positional relationship shown in the drawings, merely to facilitate description of the application and to simplify the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be configured and operated in a particular orientation, and thus should not be construed as limiting the application.

In the description of the present application, unless otherwise specified and defined, it should be noted that the terms "mounted," "connected," and "coupled" are to be construed broadly, and may be, for example, mechanical or electrical, or may be in communication with each other between two elements, directly or indirectly through intermediaries, as would be understood by those skilled in the art, in view of the specific meaning of the terms described above.

The following is a further detailed description of the embodiments:

example 1, substantially as shown in figures 1 and 2 of the accompanying drawings: a method for realizing a beam control algorithm based on ZYNQ,

Step four: PL performs traversal calculations and adopts a pipeline design.

The specific implementation process is as follows:

step one: connecting PS and PL by adopting an AXI interface, and connecting PS and PL to form ZYNQ; the PS input end interface is connected with an external instruction interface; the PL output end interface is connected with the amplitude-phase chip control interface.

Step two: the PS receives the external communication command and performs analysis and pre-calculation such as angle preprocessing, coordinate system conversion, angle correction, etc., and the analysis and pre-calculation performed by the PS includes, but is not limited to, these 3 steps, and all algorithms that do not require traversal calculation are applicable, such as histogram equalization, color space conversion, bit operation, etc.

Wherein the angle pretreatment: there is no fixed algorithm, depending on different requirements.

The coordinate system is converted into:

s=acosd(cosd(a)*cosd(b))；

p=atand(tand(b)/sind(a))；

The angle correction adopts two-dimensional linear interpolation:

r_ab=u-floor(u/1.28)*1.28；

r_ac=v-floor(v/1.28)*1.28；

u_itab=u_b_in*r_ab+u_a_in*(1.28-r_ab)；

u_itcd=u_d_in*r_ab+u_c_in*(1.28-r_ab)；

u_it=u_itcd*r_ac+u_itab*(1.28-r_ac)；

u_out=u_it/1.28/1.28

U_FPGA=(u_out+u)*1.28

v_itab=v_b_in*r_ab+v_a_in*(1.28-r_ab)；

v_itcd=v_d_in*r_ab+v_c_in*(1.28-r_ab)；

v_it=v_itcd*r_ac+v_itab*(1.28-r_ac)；

v_out=v_it/1.28/1.28

V_FPGA=(v_out+v)*1.28

where u, v are the azimuth, pitch angle, u_a_in, u_b_in, u_c_in, u_d_in, v_a_in, v_b_in, v_c_in, v_d_in of the external command input are 4 pairs of parameters required for two-dimensional linear interpolation (in this embodiment, 4 pairs of parameters are obtained from experimental tests).

Step four: PL performs traversal calculations and adopts a pipeline design.

The traversal calculation is as follows:

C(dx,dy)=-2π(dx*sinθ*cosφ+dy*sinθ*sinφ)/λ

ux=MUX_F*sinθ*cosφ

vx=MUX_F*sinθ*sinφ

MUX_F=-(360/C)*f*2^22/128=-1.2*f*2^15

The foregoing is merely exemplary of the present application and the specific structures and/or characteristics of the present application that are well known in the art have not been described in detail herein. It should be noted that modifications and improvements can be made by those skilled in the art without departing from the structure of the present application, and these should also be considered as the scope of the present application, which does not affect the effect of the implementation of the present application and the utility of the patent. The protection scope of the present application is subject to the content of the claims, and the description of the specific embodiments and the like in the specification can be used for explaining the content of the claims.

Claims

1. A method for realizing a beam control algorithm based on ZYNQ is characterized by comprising the following specific steps:

step one: connecting the output end of PS with the input end of PL by adopting an AXI interface, and connecting the PS with the PL to form ZYNQ;

step two: the PS receives, analyzes and pre-calculates the external communication instruction, and analyzes the external communication instruction into: angle preprocessing, coordinate system conversion and angle correction, wherein the coordinate system conversion comprises the following steps:

s=acosd(cosd(a)*cosd(b))；

p=atand(tand(b)/sind(a))；

wherein: a is azimuth angle, b is pitch angle; s is the off-axis angle, p is the rotation angle, d represents the units of a and b in brackets as degrees;

step three: transmitting the PS-resolved information to the PL through an AXI interface;

step four: PL performs a traversal calculation and adopts a pipeline design, the traversal calculation is:

C(dx,dy)=-2π(dx*sinθ*cosφ+dy*sinθ*sinφ)/λ

ux=MUX_F*sinθ*cosφ

vx=MUX_F*sinθ*sinφ

MUX_F=-(360/C)*f*2^22/128=-1.2*f*2^15

wherein C is the light speed=300×109mm/s, (dx, dy) is the physical coordinate value of the antenna unit in a rectangular coordinate system, and the unit is mm; θ, φ is off-axis angle and rotation angle parameters under the spherical coordinate system of the antenna, λ is signal wavelength, f is signal frequency, and the unit is Ghz; the three variables ux, vx, mux_f are intermediate variables of the calculation process.

2. The implementation method of the beam control algorithm based on the ZYNQ of claim 1, wherein the implementation method is characterized by: the angle correction uses two-dimensional linear interpolation.

3. The implementation method of the beam control algorithm based on the ZYNQ of claim 1, wherein the implementation method is characterized by: the PS input interface is an external command interface.

4. The implementation method of the beam control algorithm based on the ZYNQ of claim 1, wherein the implementation method is characterized by: the PL output end interface is an amplitude-phase chip control interface.