A system and method for real-time channel coefficient generation

Abstract

The application discloses a system and method for real-time channel coefficient generation, which comprises a parameter module, a trigonometric function lookup table module, an antenna directional diagram item module, a moving speed item module, an electromagnetic wave transmission and reception phase item module and an intra-cluster ray power multiplication and summation module; based on the system, the method for generating channel coefficients selects a channel model according to a channel standard in a protocol and inputs channel scale parameters; according to the determined channel model and the channel scale parameters, parameter arrays of each ray in each cluster are determined, calculation is performed in each module, and finally, channel coefficients are generated. The hardware real-time channel coefficient system provided by the application can generate channel coefficients in real time, has the advantages of short generation time, flexible system and meeting the needs of real-time channel simulation compared with the traditional method for generating channel coefficients by using software.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and more specifically to a scheme for real-time generation of channel coefficients and channel simulation in a wireless channel simulation system. Background Technology

[0002] Wireless channels are the medium for information transmission in wireless communication systems, playing a crucial role in the overall performance of the system. Wireless channel simulators allow for the configuration of channel parameters in a laboratory setting to accurately simulate various channel models, realistically mimicking the transmission environment of wireless signals and enabling indoor performance testing of wireless systems. A typical structure of a wireless channel simulator is as follows: Figure 1 As shown.

[0003] A typical wireless channel simulator consists of two main parts: real-time processing and offline processing. The real-time processing part primarily includes radio frequency (RF) signal processing and channel simulation. The receiving channel converts the externally input RF signal into a baseband signal. The baseband channel performs channel simulation—calculating with the offline-generated channel coefficients and adding noise. Finally, the RF transmitting channel converts the processed baseband signal back into an RF signal and transmits it.

[0004] The offline processing mainly includes channel model selection, channel coefficient generation, and sampling rate matching. Traditional channel coefficient generation methods utilize the central processing unit (CPU) for computation. This method leverages the CPU's high operating frequency and strong ability to handle complex logical tasks to accurately and intuitively generate channel coefficients. However, with the application of large-scale MIMO technology, the complexity and refinement of channel modeling have continuously increased. The number of channel coefficients that need to be computed and stored has grown exponentially, significantly increasing the required time. The channel coefficient generation time can no longer meet the needs of real-time channel simulation; therefore, alternative channel coefficient generation methods or acceleration methods need to be considered.

[0005] The current mainstream solution to this problem is to accelerate the generation of channel coefficients at the software level—using a graphics processing unit (GPU) on a parallel computing CUDA system to accelerate the generation of channel coefficients. However, the results show that even with GPU acceleration, the generation speed of channel coefficients still cannot keep up with the consumption speed. Summary of the Invention

[0006] Technical Problem: In view of this, the purpose of this invention is to propose a system for real-time channel coefficient generation based on programmable logic modules. This hardware-based channel coefficient generation scheme can solve the problem of excessively long generation time of current software methods, achieving real-time channel coefficient generation.

[0007] Technical Solution: The present invention provides a real-time channel coefficient generation system that employs parallel computing of MIMO antennas. Channel coefficients are generated by serially calculating clusters and rays within the antenna. Specific modules in the system include a parameter module, a trigonometric function lookup table module, an antenna pattern module, a moving speed module, an electromagnetic wave transmission and reception phase module, and a cluster-based ray power multiplication and summation module. The parameter module determines the parameters of the channel scale. Subsequently, the trigonometric function lookup table module obtains the trigonometric function values ​​of the angles of each ray in each cluster. The moving speed module, electromagnetic wave transmission and reception phase module, and antenna pattern module calculate the parameter arrays of moving speed, electromagnetic wave transmission and reception phase, and antenna pattern required for channel coefficient generation. Finally, the cluster-based ray power multiplication and summation module generates the channel coefficients.

[0008] The parameters of the channel determined in the parameter module include the number of receiving antennas U, the number of transmitting antennas S, the base station height h_BS, the terminal height h_UT, the horizontal distance d_2D between the base station and the mobile station, the carrier frequency Fc, the speed of light light_speed, the mobile station speed speed_UT, the number of sampling points Datalen, the initial position coordinates of the base station antenna d_Tx, and the initial position coordinates of the mobile station antenna d_Rx, which are necessary parameters for generating channel coefficients.

[0009] The aforementioned trigonometric function lookup table module optimizes and improves upon the traditional trigonometric function lookup table, by using the address of the traditional trigonometric function lookup table... Compress to Within this cluster, 3 / 4 of the lookup table resources can be saved; when using it, the departure azimuth angle of the m-th ray within the n-th cluster... (AOD, Azimuth of Departure) (AOA, Azimuth of Arrival) (Arrival pitch angle) (ZOA, Zenith of Arrival) and departure pitch angle (ZOD, Zenith of Departure) After fixed-point transformation, the address is mapped to the address in the lookup table to obtain the trigonometric function value corresponding to the mapped address. Then, based on the mapping relationship between the addresses, the required function value is obtained using the trigonometric function transformation formula.

[0010] The aforementioned moving speed module uses a counter to iterate the sampling point time, thereby realizing the calculation and update of the moving speed term; then, the multiplication IP core is used to multiply the generated departure azimuth, arrival azimuth, arrival pitch angle and trigonometric function values ​​with the moving speed term, and finally sum them up.

[0011] The electromagnetic wave transmission and reception phase term module uses a multiplication IP core to multiply the trigonometric function values ​​of relevant angles generated by the optimized trigonometric function lookup table step by step, and then accumulates the values ​​in the three-dimensional coordinates of the transmission and reception phases to generate the electromagnetic wave transmission and reception phase terms; then it is added to the movement speed term, and the sum is normalized. Then, the trigonometric function lookup table is used to find the trigonometric function value of the angle, thereby generating the exponential value of the sum of the electromagnetic wave transmission and reception phase terms and the moving speed term.

[0012] After the exponential value of the sum of the electromagnetic wave transmission and reception phase term and the moving speed term is generated, it is multiplied with the radiation pattern term generated by the software in the antenna radiation pattern term module and stored in the hardware storage module to obtain the coefficient value of the ray within the cluster.

[0013] The cluster ray power multiplication and accumulation module accumulates the coefficient values ​​of the obtained cluster rays. A counter is used to count during the accumulation process. When the counter counts to the m-th ray, the accumulator outputs. The result of the accumulator output is multiplied with the cluster normalized power given by the software. After multiplication, the result is truncated and finally the channel coefficient at that sampling point is output.

[0014] The real-time channel coefficient generation method of the system adopts the following steps:

[0015] Step S1. Refer to the 3GPP TR38.901 standard channel model and the 5GNR communication standard to determine the channel scale parameters and other parameters;

[0016] Step S2: Use a trigonometric function lookup table to generate the angle of arrival and departure of each scattering path, thereby generating the moving speed term of the channel coefficient;

[0017] Step S3: Generate electromagnetic wave transmission and reception phase terms. Since the exponent value can ultimately be represented by trigonometric function values, the powers of the movement speed term and the electromagnetic wave transmission and reception phase terms are summed. The summation result is used to find the trigonometric function value through a trigonometric function lookup table, and then the exponent value is obtained.

[0018] Step S4: Combine the antenna pattern terms to obtain the parameter arrays for generating the channel coefficients. Finally, multiply and sum the ray power within the cluster to obtain the final channel coefficients.

[0019] Beneficial effects: This invention proposes a hardware system and method for real-time channel coefficient generation based on programmable logic. Its main idea is to achieve this by using different antennas in parallel channels, in a serial cluster or ray manner. This scheme has the following significant technical effects: it can be used to replace existing software channel coefficient generation schemes, realize real-time channel coefficient generation in hardware, and meet the testing requirements of long-term real-time simulation of wireless channels. Attached Figure Description

[0020] Figure 1 This is a typical system diagram of a wireless channel simulator;

[0021] Figure 2 The hardware generates an overall system diagram for the channel coefficients;

[0022] Figure 3 Implement a module diagram for the hardware system;

[0023] Figure 4 Real-time hardware generation system diagram for CDL-A model channel coefficients;

[0024] Figure 5 This is a comparison chart showing the generation time of channel coefficients for hardware and software at different sampling points under the same scale in the embodiment. Detailed Implementation

[0025] Wireless channels are mainly composed of multiple clusters (the number of clusters varies depending on the channel model), and each cluster contains multiple ray subpaths. In the hardware system for real-time generation of channel coefficients proposed in this invention, the large amount of data for real-time generation of channel coefficients is mainly related to the accumulation of sampling points. Therefore, considering speed and hardware resource consumption, a serial cluster / ray, parallel antenna system method is used to implement the channel coefficient generation hardware system within a sampling time interval, such as... Figure 2 As shown. The hardware system module diagram is as follows. Figure 3 As shown, the hardware system calculates the array of parameters needed to generate channel coefficients using the channel model and various parameters, and then multiplies these parameters to obtain the channel coefficients. Furthermore, the hardware channel coefficient generation module can be cascaded with the hardware channel simulation system without requiring software or hardware transmission of channel coefficients, thus reducing storage and transmission time.

[0026] Taking the CDL-A model in 3GPP TR38.901, the current mainstream 5G channel model, as an example, the formula for calculating the generated channel coefficients is as follows:

[0027]

[0028] in This represents the normalized power of the nth cluster. The meanings of the other symbols will be explained in detail in the specific modules.

[0029] The specific system for hardware-generated channel coefficients designed in this invention is as follows: Figure 4 As shown. The main modules of the hardware system are as follows.

[0030] Module 1: Parameter Module;

[0031] Select a channel model, determine the number of channel clusters N and the number of rays within a cluster M, determine the channel scale parameters and other parameters such as the height of the base station and the mobile station;

[0032] Module 2: Trigonometric Function Lookup Table Module;

[0033] We construct a trigonometric function lookup table that is important for generating channel coefficients using an IP core (Intellectual Property Core) and propose an optimization method.

[0034] Module 3: Antenna Pattern Item Module;

[0035] Antenna pattern items are stored using a hardware storage module:

[0036] .

[0037] in Let represent the field components of the elevation radiation pattern of the u-th receiving antenna. Let represent the field components of the azimuth radiation pattern of the u-th receiving antenna. The field components of the elevation radiation pattern of the s-th transmitting antenna are represented. Let represent the field components of the azimuth radiation pattern of the s-th transmitting antenna; Indicates the power ratio of linear cross-polarization; , , , Indicates pitch angle and azimuth The random initial phase of the m-th ray in the n-th cluster under four different polarization combinations; T represents the matrix transpose.

[0038] Module 4: Movement Speed ​​Item Module;

[0039] The moving speed term in the hardware-generated channel coefficients is formulated as follows: .

[0040] in This represents the spherical unit vector representing the azimuth angle of the m-th ray in the n-th cluster at the receiving end. This represents the velocity vector of the mobile station at the receiving end. Indicates the sampling point time. Indicates the carrier wavelength.

[0041] Module 5: Electromagnetic wave transmission and reception phase module;

[0042] Generate electromagnetic wave transmission and reception phase terms in hardware. .

[0043] in This represents a spherical unit vector representing the azimuth angle of the m-th ray in the n-th cluster at the emitting end. This represents the position vector of the u-th receiving antenna element. This represents the position vector of the s-th transmitting antenna element.

[0044] Module 6: Cluster-based X-ray Power Multiplication and Accumulation Module

[0045] In the hardware, the intra-cluster ray power is multiplied and accumulated to generate channel coefficients;

[0046] Furthermore, in module 1 (parameter module), the channel scale parameters include the number of receiving antennas U and the number of transmitting antennas S; among other parameters, the heights of the base station and the terminal are denoted as h_BS and h_UT, respectively; the horizontal distance between the base station and the terminal mobile station is denoted as d_2D; and the carrier frequency is denoted as Fc.

[0047] Furthermore, the optimization method for the trigonometric function lookup table proposed in Module 2 (Trigonometric Function Lookup Table Module) refers to implementing the trigonometric function lookup table using a data compression method. Only one lookup table is needed to complete the lookup and evaluation of all trigonometric functions for sin and cos. In specific implementation, [the following is a more detailed explanation of the method]. The address is fixed-point converted and used as the address of the lookup table; the parameter for generating the channel coefficients—the departure azimuth angle of the m-th ray in the n-th cluster—is used. azimuth of arrival Reaching pitch angle and leave pitch angle After fixed-point conversion, the resulting address is mapped to the address in the lookup table, yielding the trigonometric function value corresponding to the mapped address. Then, based on the mapping relationship between addresses, the required function value is obtained using the trigonometric function conversion formula.

[0048] Furthermore, in module 3 (antenna pattern item module), the antenna pattern item consists of the receiving antenna field pattern, the transmitting field pattern, and the random initial phase of the ray, which are collectively referred to as the pattern item. Since the pattern item is highly complex and the hardware implementation result is complicated, it is calculated by software and stored in the hardware storage module.

[0049] Furthermore, in module 4 (movement speed module), a counter is first used in the hardware to iterate the sampling point time, thereby realizing the calculation and update of the speed term; then, the multiplication IP core is used to multiply the trigonometric function values ​​of the relevant angles generated by the optimized trigonometric function lookup table proposed in this invention with the speed term, and finally the summation is performed.

[0050] Furthermore, in module 5 (electromagnetic wave transmission and reception phase term module), the phase merging of transmission and antenna is implemented: in the hardware, the trigonometric function values ​​of the relevant angles generated by the optimized trigonometric function lookup table are multiplied step by step using a multiplication IP core, and then the values ​​in the three-dimensional coordinates of the transmission and reception phases are accumulated. After accumulation, they are added to the moving speed term generated by module 4, and then normalized. Then, the trigonometric function lookup table is used to find the trigonometric function value of the angle, thereby generating the exponential term.

[0051] Furthermore, in module 6 (the cluster-internal ray power multiplication and accumulation module), a complex IP core is used to multiply the exponential term generated in module 5 and the antenna pattern term in module 3. The multiplication is then accumulated, with a counter used for counting. When the counter reaches the m-th ray, the accumulator outputs its result. The accumulator output is multiplied by the cluster normalized power given by the software, and then truncated to finally output the channel coefficient at that sampling point.

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with the embodiments and accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0053] This invention constructs a hardware system and method for real-time generation of channel coefficients. Compared with existing technologies that use CPU to generate channel coefficients in software, this invention implements the key calculation point of channel coefficients—the small-scale fading model—in hardware, and performs multiplication and superposition on the corresponding arrays. This can greatly reduce the generation time of channel coefficients and realize real-time generation of channel coefficients.

[0054] This embodiment provides a specific method for generating channel coefficients using hardware, based on a selected channel scenario and various channel parameters. The steps are as follows:

[0055] Step S1: Determine the channel scale parameters and other parameters by referring to the 3GPP TR38.901 channel model and the 5GNR communication standard;

[0056] Specifically, the parameters determined in this embodiment include: the number of receiving antennas U, the number of transmitting antennas S, the base station height h_BS, the terminal height h_UT, the horizontal distance d_2D between the base station and the mobile terminal, and the carrier frequency F. CLight speed, mobile station speed, number of sampling points (Datalen), initial position coordinates of base station antenna (d_Tx), and initial position coordinates of mobile station antenna (d_Rx);

[0057] Step S2: Use a trigonometric function lookup table to generate the arrival and departure angles of each scattering path, thereby generating the moving speed term of the channel coefficient. ;

[0058] Specifically, in this embodiment, the arrival angle and departure angle of each scatterer path within the cluster are... After generation, the optimized trigonometric function lookup table designed in this invention is used to find its trigonometric function values, and then the spherical unit vector of the receiving antenna is calculated using the formula. and the velocity vector of the receiving mobile station The multiplication is performed using the multiplication IP core in the movement speed module of the hardware system designed in this invention. Sampling point time... This is achieved by a counter – the sampling time can be obtained by multiplying the sampling time interval Ts by the number of sampling points counted.

[0059] Step S3: Generate electromagnetic wave transmission and reception phase terms .

[0060] Specifically, in this embodiment, the arrival angle and departure angle of each scatterer path within the cluster are... After generation, the trigonometric function values ​​are looked up using the optimized trigonometric function lookup table designed in this invention, and then the spherical unit vector of the transmitting antenna is calculated using the formula. , spherical unit vector of the receiving antenna Position vector of transmitting antenna element Position vector of the receiving antenna element The multiplication IP core in the electromagnetic wave transmission and reception phase terms of the hardware system designed in this invention is used to perform cumulative multiplication and summation.

[0061] Since the exponent value can ultimately be represented by trigonometric function values, the powers of the moving speed term and the electromagnetic wave transmission and reception phase terms are summed. The summation result is used to look up the trigonometric function values ​​in a trigonometric lookup table to obtain the exponent value.

[0062] Step S4: Call the antenna pattern entries in the hardware storage system, multiply and sum the ray power within the cluster, and calculate the final channel coefficient.

[0063] Specifically, in this embodiment, in the intra-cluster ray power multiplication and summation module of the present invention, the exponential value and the antenna pattern item of the nth cluster and the mth ray generated in step S3 are multiplied, and after multiplication, they are summed. A counter is used for counting during the summation process. When the counter counts to the mth ray, the accumulator outputs. The result output by the accumulator is multiplied by the cluster normalization power given by the software, and after multiplication, truncation is performed, and finally the channel coefficient at the sampling point moment is output.

[0064] To verify the effectiveness of the present invention, the channel coefficients generated in the embodiment are verified. Under the condition of ensuring that all parameters are the same, the channel coefficient values generated by hardware and the channel coefficient values generated by software are compared, and it is found that the variance of the real part of the channel coefficient and the variance of the imaginary part of the channel coefficient are controlled within 10 -3 Below, the overall deviation is not large, indicating that the method of generating channel coefficients by hardware in the present invention ensures the accuracy of the channel coefficients.

[0065] To verify the performance of the real-time generation hardware system proposed by the present invention, different central processing units (CPUs) and graphics processing units (GPUs) are selected to generate software channel coefficients respectively, and the time for the hardware system with different clock frequencies to generate channel coefficients is compared. The results are as Figure 5 shown. For a single-input single-output channel, the relationship of the channel coefficient calculation time is basically hardware < GPU < CPU. When the number of sampling points is small, the time advantage of hardware calculation is very obvious. However, when the number of sampling points continues to increase, the error between the high-performance GPU and the hardware calculation time will decrease. On the other hand, it can be seen that whether it is CPU, GPU or hardware, when the channel configuration is the same, the calculation time will increase approximately linearly with the increase of the number of sampling points, corresponding to the channel coefficient generation method.

[0066] The parts not detailed in the present invention are all well-known technologies in the art. The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative labor. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present invention through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art shall be within the protection scope determined by the claims.

Claims

1. A system for real-time channel coefficient generation, characterized in that: The system employs a parallel computing MIMO antenna, with clusters and rays being calculated serially within the antenna to generate channel coefficients. Specific modules within the system include a parameter module, a trigonometric function lookup table module, an antenna pattern module, a moving speed module, an electromagnetic wave transmission and reception phase module, and a cluster-based ray power multiplication and summation module. The parameter module determines the channel scale parameters, then obtains the trigonometric function values ​​of each angle for each ray in each cluster through the trigonometric function lookup table module. The moving speed module, electromagnetic wave transmission and reception phase module, and antenna pattern module calculate the necessary parameter arrays for generating channel coefficients, and finally, the cluster-based ray power multiplication and summation module generates the channel coefficients. The aforementioned trigonometric function lookup table module optimizes and improves upon the traditional trigonometric function lookup table, extending the traditional lookup table's address range from [0, 2π] to [0, π / 2]. When used, it retrieves the departure azimuth angle of the m-th ray within the n-th cluster. azimuth of arrival Reaching pitch angle and leave pitch angle After locating, the address is mapped to the address in the lookup table to obtain the trigonometric function value corresponding to the mapped address. Then, based on the mapping relationship between the addresses, the trigonometric function conversion formula is used to obtain the trigonometric function value of the required angle. The aforementioned moving speed module uses a counter to iterate the sampling point time, thereby realizing the calculation and update of the moving speed term; then, the multiplication IP core is used to multiply the generated trigonometric function values ​​of departure azimuth, arrival azimuth, arrival pitch angle and departure azimuth with the moving speed term, and finally sum them up.

2. The system for real-time channel coefficient generation according to claim 1, characterized in that, The parameters of the channel determined in the parameter module include the number of receiving antennas U, the number of transmitting antennas S, the base station height h_BS, the terminal height h_UT, the horizontal distance d_2D between the base station and the mobile station, the carrier frequency Fc, the speed of light light_speed, the mobile station speed speed_UT, the number of sampling points Datalen, the initial position coordinates of the base station antenna d_Tx, and the initial position coordinates of the mobile station antenna d_Rx. These are the parameters necessary for generating the channel coefficients.

3. The system for generating real-time channel coefficients according to claim 2, further characterized in that, The electromagnetic wave transmission and reception phase term module uses a multiplication IP core to multiply the trigonometric function values ​​of the relevant angles generated by the optimized trigonometric function lookup table step by step, and then accumulates the values ​​in the three-dimensional coordinates of the transmission and reception phases to generate the electromagnetic wave transmission and reception phase terms. Then, add it to the moving speed term, and normalize the sum to [0, -π / 2]. Then, use a trigonometric function lookup table to find the trigonometric function value of the angle, thereby generating the exponential value of the sum of the electromagnetic wave transmission and reception phase term and the moving speed term.

4. The system for real-time channel coefficient generation according to claim 3, characterized in that, After the exponential value of the sum of the electromagnetic wave transmission and reception phase term and the moving speed term is generated, it is multiplied with the radiation pattern term generated by the software in the antenna radiation pattern term module and stored in the hardware storage module to obtain the coefficient value of the ray within the cluster.

5. The system for real-time channel coefficient generation according to claim 4, further characterized in that, The cluster ray power multiplication and accumulation module accumulates the coefficient values ​​of the obtained cluster rays. A counter is used to count during the accumulation process. When the counter counts to the m-th ray, the accumulator outputs. The result of the accumulator output is multiplied with the cluster normalized power given by the software. After multiplication, the result is truncated and finally the channel coefficient at that sampling point is output.

6. A real-time channel coefficient generation method for the system as described in claim 1, characterized in that, The method employs the following steps: Step S1. Refer to the 3GPP TR38.901 standard channel model and the 5GNR communication standard to determine the channel scale parameters and other parameters; Step S2: Use a trigonometric function lookup table to generate the angle of arrival and departure of each scattering path, thereby generating the moving speed term of the channel coefficient; Step S3: Generate electromagnetic wave transmission and reception phase terms. Since the exponent value can ultimately be represented by trigonometric function values, the powers of the movement speed term and the electromagnetic wave transmission and reception phase terms are summed. The summation result is used to find the trigonometric function value through a trigonometric function lookup table, and then the exponent value is obtained. Step S4: Combine the antenna pattern terms to obtain the parameter arrays for generating the channel coefficients. Finally, multiply and sum the ray power within the cluster to obtain the final channel coefficients.

Images