Electromagnetic channel calculation method for large-scale scene

Abstract

The invention discloses an electromagnetic channel calculation method for a large-scale scene, and the method comprises the following steps: S1, converting a reflector in the scene into a discrete surface element, carrying out the modeling of a reflection channel, and completing the modeling of a primary reflection channel of a receiver and the iterative recursive transformation of the modeling of a multi-reflection channel; s2, calculating a reflection channel: carrying out channel calculation on a primary reflection channel between the transmitter and the receiver at each moment according to a primary reflection modeling result; the method comprises the following steps of: for a multi-reflection channel between a transmitter and a receiver, converting channel modeling into a recursion form, calculating by multiplexing a result at a previous moment to reduce the recursion times, and searching by combining a low-complexity approximation method, channel correction and a defined search function to complete the calculation of the multi-reflection channel. According to the method, the channel calculation process is optimized in an iterative recursion mode, and the method has the advantages of being low in calculation complexity, high in calculation speed and high in time delay resolution.

Description

Technical Field

[0001] The present invention relates to the field of electromagnetic waves, and in particular to an electromagnetic channel calculation method for large-scale scenes. Background Art

[0002] In the military and civilian fields, electromagnetic spectrum equipment needs to be tested and trained to improve its performance and reliability. Traditional testing and training methods need to be carried out in actual outdoor scenes, which is costly and complicated to organize. Therefore, it is an effective way to test and train by constructing a digital electromagnetic environment of a large-scale outdoor scene to simulate the channel conditions of equipment in actual scenes. However, as the simulation scene becomes more complex, higher requirements are placed on the speed and accuracy of electromagnetic channel calculation.

[0003] In the calculation of electromagnetic channels in large-scale scenes, the terrain is complex and changeable, and there are a large number of irregular rough surfaces. It is necessary to simulate diffuse reflection and scattering channels. Traditional deterministic algorithms are generally used for simulation, such as ray tracing and time-domain radiosity algorithms. The traditional time-domain radiosity algorithm converts the reflectors in the scene into discrete surface elements and calculates the radiosity between the surface elements to simulate the reflection channel. However, the computational complexity of this method is O(N m ), the O symbol (usually represented as O) is used to describe the time complexity or space complexity of an algorithm or function. It provides an upper bound estimate of the rate at which a function grows, and is used to indicate the rate at which the performance of an algorithm grows as the size of the input data grows. It describes the worst-case performance of an algorithm, where N is the number of bins and m is the number of simulated channel reflections. As m increases, the computational complexity increases exponentially, resulting in low computational efficiency and exponential growth in storage space requirements. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an electromagnetic channel calculation method for large-scale scenarios, which optimizes the channel modeling process by iterative recursion and has the characteristics of low computational complexity, fast calculation speed and high delay resolution.

[0005] The objective of the present invention is achieved through the following technical solution: A method for calculating an electromagnetic channel for a large-scale scene, comprising the following steps:

[0006] S1. Convert the reflectors in the scene into discrete surface elements, and then model the reflection channel (for example, the time domain radiosity algorithm for scattering channel modeling), and complete the iterative transformation of the modeling of the receiver's single reflection channel and the modeling of multiple reflection channels;

[0007] S2. Calculate the single reflection channel and multiple reflection channel between the transmitter and the receiver respectively:

[0008] The channel calculation is performed on the single reflection channel between the transmitter and the receiver at each time according to the single reflection modeling result;

[0009] For the multiple reflection channels between the transmitter and the receiver, the channel modeling is converted into a recursive form, and the number of recursions is reduced by reusing the results of the previous moment for calculation. At the same time, a low-complexity approximation method and channel correction are combined, and a search function is defined to search and complete the calculation of the multiple reflection channels.

[0010] The beneficial effect of the present invention is that the present invention optimizes the channel modeling process by iterative recursion. The method utilizes the principle of spatial coherence, reuses the channel data of the previous moment, and significantly reduces the computational complexity by correcting the calculation used for the current moment. The core advantages of the algorithm include: 1) Compared with the traditional algorithm, the method proposed in the present invention achieves an order of magnitude increase in the computational speed while ensuring the computational accuracy, meeting the high-speed computing requirements in large-scale scenarios; 2) It has a wide range of applicability and can adapt to the dynamic changes of electromagnetic devices in the scene, such as mobile vehicles and drones, which improves practicality; 3) Low storage space requirements: By optimizing the data structure, the storage space requirements are reduced, so that the algorithm can run on limited hardware resources; 4) By selecting the number of channels stored in each face element, a compromise between computational accuracy and computational speed and storage space is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 Schematic diagram of the reflection channel between the transceiver and the scene element;

[0012] Figure 2 This is a schematic diagram of channel correction;

[0013] Figure 3 It is a flowchart of the iterative time-domain radiometry algorithm system;

[0014] Figure 4 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0015] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0016] like Figure 1 As shown in FIG. , it is a schematic diagram of the reflection channel between the transceiver and the scene element. Figure 4 As shown, a method for calculating electromagnetic channels for large-scale scenes includes the following steps:

[0017] S1. Convert the reflectors in the scene into discrete surface elements, then model the reflection channel (for example, the time domain radiometry method for scattering channel modeling), and complete the iterative recursive transformation of the modeling of the single reflection channel and the multiple reflection channel of the receiver;

[0018] S101. Convert the reflectors in the scene into discrete surface elements, referred to as surface elements, and the reflection channel between the transmitter and the receiver Considered as the signal received by each surface element in the scene, after reflection, it finally converges at the receiver to form:

[0019] For any facet i, it receives the direct signal from the transmitter and the signal reflected from other facets, and the sum of the channels formed is Surface element i then reflects the received signal to the receiver, forming a reflection channel between the transmitter and the receiver;

[0020] S102. Define a function F to characterize the channel characteristics. This function is also called a shape factor. For the shape factor F between any bin k and any bin i, ki By the attenuation component D ki , delay component T ki and the visual function V ki composition:

[0021]

[0022] in is the propagation delay, d ki is the distance between the facets, c is the speed of light, ρ k is the reflection coefficient, θ i and θ k is the angle between the surface element normal vector and the line connecting the two, A i is the receiving area, visible function

[0023] The shape factor of the transmitter and the bin i is denoted as F Ti , the shape factor of the bin i and the receiver is expressed as F iR ;

[0024] S103. Modeling the reflection channel, including single reflection channel modeling and multi-reflection channel modeling.

[0025] The single reflection channel modeling includes:

[0026] For a single reflection channel, the direct channel from the transmitter to each facet in the scene is expressed as is the channel received by the plane element i without reflection, that is, the channel F of the transmitter directly irradiating the plane element i Ti , The direct channel from each bin to the receiver is represented by L R =[F 1R …F NR ], where N is the total number of facets;

[0027] Each bin reflects the direct signal from the transmitter to the receiver, forming all the primary reflection channels between the transceiver:

[0028]

[0029] Define the calculation diagonalization function diag([c 1,1 … c 1,N ])=[c 1,1 … c 1,N ] diag , where c 1,1 to c 1,N For the 1st to Nth submatrices in the block matrix, define the calculation in [C] i,j is the element in the i-th row and j-th column of matrix C, [A] i,k is the element in the i-th row and k-th column of matrix A, [B] k,j is the element in the kth row and jth column of matrix B. Where * is the convolution operation.

[0030] The multiple reflection channel modeling includes:

[0031] For the modeling of multiple reflection channels, the single reflection channel from bin k to bin i is The first reflection channel from other pixels in pixel i is represented by the vector Elements in a vector represents the first reflection channel transmitted from the transmitter, reflected by the bin 1 and finally reaching the bin i. The same is true for other elements. Therefore, the first reflection channels in each bin form an N×N matrix.

[0032]

[0033] The matrix The element F in row i and column j ij is the shape factor between face element i and face element j, and this matrix is ​​defined as the transfer matrix L H , define the function and and will Recorded as Represents the first reflection channel received by each facet from other facets. Each facet reflects the first reflection signal from other facets to the receiver, forming the sum of the second reflection channels between the transceiver and the receiver. is the direct channel matrix L from each bin to the receiver R Diagonalization of ;

[0034] Similarly, the secondary reflection channel from bin j to bin i is The two reflection channels in the bin i from other bins are represented as vectors The secondary reflection channel in each bin constitutes an N 2 ×N matrix

[0035]

[0036] in Represents the secondary reflection channel received by each bin from other bins. The secondary reflection signal in each bin is reflected to the receiver, forming the sum of the three-reflection channels between the transceiver and the receiver. Therefore, t i The sum of the reflected channels between the transceivers at the moment It is expressed as:

[0037]

[0038] Define matrix concatenation calculation t i The direct channel in each bin at the moment and the reflected channel with a reflection number less than m-1, where m-1 is the upper limit of the reflection number.

[0039] S2. Calculate the single reflection channel and multiple reflection channel between the transmitter and the receiver respectively:

[0040] The channel calculation is performed on the single reflection channel between the transmitter and the receiver at each time according to the single reflection modeling result;

[0041] For the multiple reflection channels between the transmitter and the receiver, the channel modeling is converted into a recursive form, and the number of recursions is reduced by reusing the results of the previous moment for calculation. At the same time, a low-complexity approximation method and channel correction are combined, and a search function is defined to search and complete the calculation of the multiple reflection channels.

[0042] Will The calculation is converted into a recursive form, and the number of recursions is reduced by reusing the calculation results of the previous moment for calculation. At the same time, the complexity of the overall calculation is reduced by combining low-complexity approximation methods;

[0043] In formula (5) is the channel in each bin, and its size grows exponentially with the number of reflections. Therefore, the complexity of the time-domain radiance algorithm also grows exponentially with the number of reflections.

[0044] The present invention proposes an improved channel calculation method. The calculation is converted into a recursive form, and the number of recursions is reduced by reusing the calculation results of the previous moment. At the same time, combined with the low-complexity approximation method, the complexity of the overall calculation is effectively reduced.

[0045] In the traditional time domain radiometry algorithm, at t i At this moment, t is calculated by performing m-1 reflections through the recursive formula. i The direct channel in each bin at the moment and the reflected channel with a reflection number less than m-1

[0046]

[0047] in is the direct channel in each bin and the reflected channel with a reflection number less than m-2;

[0048] In t i+1 The time also needs to be calculated multiple times recursively And will Replace with t i Calculation results at time When the transmitter is stationary, it can be assumed that the channel at each moment has not changed. When the transmitter is moving, the channel at each moment will change. Therefore, it is necessary to introduce a channel correction function to solve this problem; it can reduce the number of recursions and increase the calculation speed.

[0049] Add a channel correction function and calculate the correction value to correct the channel;

[0050] Considering the direct channel at adjacent times and The deviation of will cause the reflection channel deviation, such as Figure 2 As shown, we need to add the channel modification function modify:

[0051]

[0052] The channel correction value is calculated based on the data of the current moment and the previous moment, so the correction of the channel is real-time and dynamic. Figure 2 As shown in the figure, the time interval between adjacent moments is very short (millisecond level), the facets in the scene will hardly change, and the channels between facets will not change. The channel changes in the facets mainly come from the changes in the direct channels of the transmitter to the facets caused by the movement of the transmitter. Therefore, the modify function completes the correction by obtaining the change values ​​of the delay and attenuation of the direct channel between the transmitter and the facet at different times. For example: Compare t iThe shape factor of the transmitter direct beam element k at time ( Figure 2 The red dashed line in the middle) is related to the current t i+1 The shape factor corresponding to the moment ( Figure 2 The red solid line in the middle) calculates the delay correction value and attenuation correction Finally, the channel is corrected according to the correction value:

[0053]

[0054] Define the search function to search and complete the electromagnetic channel calculation of the reflection channel.

[0055] To reduce the computational complexity, define the search function sort_max L-1 , search for the L-1 reflection channels with the largest strength among the reflection channels of each facet, where L is the channel buffer length, and its value is determined by the ratio of the searched reflection channel strength to the total reflection channel strength.

[0056]

[0057] in t i The L-1 reflection channels with the strongest intensity at each bin at the moment are obtained, and then the searched channels are modified using the modify function:

[0058]

[0059] is the corrected reflection channel of each bin, plus the current t i+1 After the direct channel at time, it is defined as the channel buffer Therefore, t i+1 time The calculation process is expressed as:

[0060]

[0061] Finally, t i+1 Time channel cache The channel in is reflected to the receiver, and the channel between the transceiver is obtained. Complete i+1 The system block diagram of the iterative channel calculation method is as follows: Figure 3 shown.

[0062] In an embodiment of the present application, a method for calculating an electromagnetic channel for a large-scale scene specifically includes the following steps:

[0063] A1. At the first moment, iterative modeling is performed according to the multiple reflection channels between the transmitter and the receiver, and all the multiple reflection channels at the first moment are calculated;

[0064] Starting from the second moment, repeating A2 to A5 at each moment can complete the calculation of the electromagnetic channel. It should be noted that in steps A2 to A5, t i+1 The moment represents the current moment, t i The moment indicates the previous moment;

[0065] A2. Starting from the second moment, at each moment, the channel calculation result of the previous moment is used, combined with a low-complexity approximation method to reduce the number of recursions, to complete the calculation of the multiple reflection channel;

[0066] A3. Starting from the second moment, at each moment, use the search function sort_max for each face element L-1 Search for the L-1 reflection channels with the largest intensity in the bin And use the channel correction function to correct it to get

[0067] A4. Starting from the second moment, at each moment, for each face element Add the current t i+1 Direct channel at time Then, define it as t i+1 Time channel cache

[0068]

[0069] A5. Starting from the second moment, at each moment, set t i+1 Time channel cache The channel in is reflected to the receiver, and the channel between the transceiver is obtained. Complete i+1 Channel calculation at time ; where, Indicates t i+1 The direct channel matrix L from each bin to the receiver at the moment R Diagonalization of .

[0070] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the above embodiments, it is still possible for a person skilled in the art to modify the methods described in the above embodiments, such as changing the names of the methods, etc. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for calculating electromagnetic channels for large-scale scenes, characterized by: The following steps are involved: S1. Convert the reflectors in the scene into discrete surface elements, and then model the reflection channel to complete the iterative transformation of the modeling of the single reflection channel and the modeling of the multiple reflection channel of the receiver; S2. Calculate the single reflection channel and multiple reflection channel between the transmitter and the receiver respectively: The channel calculation is performed on the single reflection channel between the transmitter and the receiver at each time according to the single reflection modeling result; For the multiple reflection channels between the transmitter and the receiver, the channel modeling is converted into a recursive form, and the number of recursions is reduced by reusing the results of the previous moment for calculation. At the same time, a low-complexity approximation method and channel correction are combined, and a search function is defined to search and complete the calculation of the multiple reflection channels.

2. The electromagnetic channel calculation method for large-scale scenes according to claim 1, characterized in that: The step S1 comprises: S101. Convert the reflectors in the scene into discrete surface elements, referred to as surface elements, and the reflection channel between the transmitter and the receiver Considered as the signal received by each surface element in the scene, after reflection, it finally converges at the receiver to form: For any facet i, it receives the direct signal from the transmitter and the signal reflected from other facets, and the sum of the channels formed is Surface element i then reflects the received signal to the receiver, forming a reflection channel between the transmitter and the receiver; S102. Define a function F to characterize the channel characteristics. This function is also called a shape factor. For the shape factor F between any bin k and any bin i, ki By the attenuation component D ki , delay component T ki and the visual function V ki composition: in is the propagation delay, d ki is the distance between the facets, c is the speed of light, ρ k is the reflection coefficient, θ i and θ k is the angle between the surface element normal vector and the line connecting the two, A i is the receiving area, visible function The shape factor of the transmitter and the bin i is denoted as F Ti , the shape factor of the bin i and the receiver is expressed as F iR ; S103. Modeling the reflection channel, including single reflection channel modeling and multi-reflection channel modeling between the transmitter and the receiver.

3. The electromagnetic channel calculation method for large-scale scenes according to claim 2, characterized in that: The modeling of a single reflection channel between the transmitter and the receiver includes: For a single reflection channel, the direct channel from the transmitter to each facet in the scene is expressed as The direct channel from each bin to the receiver is represented by L R =[F 1R … F NR ], where N is the total number of facets; Each bin reflects the direct signal from the transmitter to the receiver, forming all the primary reflection channels between the transmitter and the receiver: Define the calculation diagonalization function diag([c 1,1 …c 1,N ])=[c 1,1 …c 1,N ] diag , where c 1,1 to c 1,N is the 1st to Nth submatrix in the block matrix, is the direct channel matrix L from each bin to the receiver R Diagonalization of ; Defining calculations in [C] i,j is the element in the i-th row and j-th column of matrix C, [A] i,k is the element in the i-th row and k-th column of matrix A, [B] k,j is the element in the kth row and jth column of matrix B.

4. The electromagnetic channel calculation method for large-scale scenes according to claim 2, characterized in that: The multiple reflection channel modeling between the transmitter and the receiver includes: For the modeling of multiple reflection channels, the single reflection channel from bin k to bin i is The first reflection channel from other pixels in pixel i is represented by the vector Therefore, the primary reflection channels in each facet form an N×N matrix; The matrix The element F in row i and column j ij is the shape factor between face element i and face element j, and this matrix is ​​defined as the transfer matrix L H , define the function and and will Recorded as Represents the first reflection channel received by each facet from other facets. Each facet reflects the first reflection signal from other facets to the receiver, forming the sum of the second reflection channels between the transceiver and the receiver. Similarly, the secondary reflection channel from bin j to bin i is The two reflection channels in the bin i from other bins are represented as vectors The secondary reflection channel in each bin constitutes an N 2 ×N matrix in Represents the secondary reflection channel received by each bin from other bins. The secondary reflection signal in each bin is reflected to the receiver, forming the sum of the three-reflection channels between the transceiver and the receiver. Therefore, t i The sum of the reflected channels between the transceivers at the moment It is expressed as: Define matrix concatenation calculation t i The direct channel in each bin at the moment and the reflected channel with a reflection number less than m-1, where m-1 is the upper limit of the reflection number.

5. The electromagnetic channel calculation method for large-scale scenes according to claim 3 is characterized in that: In the step S2, for the primary reflection channel between the transmitter and the receiver, at each time, a model is built according to the primary reflection channel, and all the primary reflection channels between the transmitter and the receiver are calculated.

6. The electromagnetic channel calculation method for large-scale scenes according to claim 4, characterized in that: In step S2, for the multiple reflection channels between the transmitter and the receiver, the channel modeling is converted into a recursive form, and the number of recursions is reduced by reusing the result of the previous moment for calculation. At the same time, a low-complexity approximation method and channel correction are combined to define a search function for searching, and the calculation of the reflection channel is completed, including: A1. At the first moment, iterative modeling is performed according to the multiple reflection channels between the transmitter and the receiver, and all the multiple reflection channels at the first moment are calculated; A2. Starting from the second moment, at each moment, the channel calculation result of the previous moment is used, combined with a low-complexity approximation method to reduce the number of recursions, to complete the calculation of the multiple reflection channel; A3. Starting from the second moment, at each moment, use the search function sort_max for each face element L-1 Search for the L-1 reflection channels with the largest intensity in the bin And use the channel correction function to correct it to get A4. Starting from the second moment, at each moment, for each face element Add the current t i+1 Direct channel at time Then, define it as t i+1 Time channel cache A5. Starting from the second moment, at each moment, set t i+1 Time channel cache The channel in is reflected to the receiver, and the channel between the transceiver is obtained. Complete i+1 Channel calculation at time ; where, Indicates t i+1 The direct channel matrix L from each bin to the receiver at the moment R Diagonalization of .

7. The electromagnetic channel calculation method for large-scale scenes according to claim 6, characterized in that: The step A2 comprises: In the traditional time domain radiometry algorithm, at t i At this moment, t is calculated by performing m-1 reflections through the recursive formula. i The direct channel in each bin at the moment and the reflected channel with a reflection number less than m-1 in is the direct channel in each bin and the reflected channel with a reflection number less than m-2; In t i+1 The time also needs to be calculated multiple times recursively And will Replace with t i Calculation results at time Reducing the number of recursions increases calculation speed.

8. The electromagnetic channel calculation method for large-scale scenes according to claim 6, characterized in that: The step A3 comprises: Define the search function sort_max L-1 , search for the L-1 reflection channels with the largest intensity among the reflection channels of each facet, where L is the channel buffer length, and its value is determined by the ratio of the intensity of the reflection channels searched out to the total intensity of the reflection channels; in t i The L-1 reflection channels with the strongest intensity in each bin at the moment; Use the modify function to modify the searched channel, that is, Each reflection channel in is corrected to obtain 9. The electromagnetic channel calculation method for large-scale scenes according to claim 6, characterized in that: In step A3, the method of determining the modify function includes: The time interval between adjacent moments is very short. Assuming that the facets in the scene will not change, the channels between facets will not change. The channel changes in the facets are mainly caused by the change of the direct channel of the transmitter to the facets due to the movement of the transmitter. Therefore, for each channel, according to t i The shape factor of the transmitter direct beam element k at time With the current t i+1 The shape factor corresponding to the moment Calculate the delay correction value and attenuation correction The modify function will be D mod T mod With t i The channel at time t is convolved, and the result is i+1 Channel calculation result at time.

Images