A high mobility channel AFDM chirp parameter selection method and system based on DAFT domain strip energy evaluation

Abstract

The application discloses a high-mobility channel AFDM chirp parameter selection method and system based on a DAFT domain strip-shaped energy evaluation, is suitable for a high-mobility wireless communication scene and belongs to the technical field of wireless communication physical layers.The method comprises the following steps: obtaining or generating a linear time-varying channel matrix of a high-mobility channel; a DAFT domain equivalent channel matrix is constructed for a candidate; the cyclic in-band energy proportion, the minimum equivalent cyclic half-bandwidth required to reach a preset energy threshold and the corresponding matrix cyclic out-of-band leakage proportion are calculated, and the strip-shaped truncation loss under a fixed receiving half-bandwidth can be evaluated; according to the above evaluation indexes, the equivalent channel energy is adapted to a strip-shaped MMSE type low-complexity reception, so that the strip-shaped truncation loss and the receiving complexity are reduced.The application can adaptively select parameters more suitable for low-complexity strip-shaped reception according to the current linear time-varying channel matrix.

Description

Technical Field

[0001] This invention relates to the field of wireless communication physical layer technology, and in particular to a method and system for selecting AFDM chirp parameters for high mobility channels based on DAFT domain strip energy evaluation. Background Technology

[0002] High-mobility wireless communication scenarios, such as millimeter-wave links for aircraft, high-speed air-to-ground mobile links, UAV links, high-speed vehicle-mounted links, and aerospace platform communication links, typically exhibit characteristics such as high node movement speed, rapid time-varying channels, significant Doppler frequency shift, and sparse or quasi-sparse propagation paths. Millimeter-wave links, in particular, also exhibit characteristics such as high carrier frequency, narrow beamwidth, and sensitivity to beam misalignment.

[0003] Traditional OFDM systems can achieve low complexity in low-speed or near-time-invariant channels using frequency-domain single-tap equalization; however, in high-mobility dual-select channels, time-varying channels can disrupt subcarrier orthogonality, introducing significant inter-carrier interference. If the receiver only uses frequency-domain diagonal MMSE detection, it is difficult to handle interference caused by strong Doppler and off-diagonal coupling. OTFS can organize the signal in the time-delay-Doppler domain, but in practical low-complexity reception scenarios, the receiver window size, strip approximation method, and transform-domain leakage still affect the trade-off between complexity and performance.

[0004] AFDM constructs modulation and demodulation processes based on Discrete Affine Fourier Transform (DAFT), and its chirp parameters can change the equivalent support position and support shape of the path in the DAFT domain. For high-mobility dual-selection channels, appropriate AFDM chirp parameters can concentrate the equivalent channel energy in the cyclic strip region, thus facilitating processing by low-complexity receivers such as strip LMMSE. Existing fixed-parameter AFDM schemes typically do not fully utilize the current linear time-varying channel matrix and are difficult to adapt to changes in DAFT domain path overlap and leakage under different motion states, Doppler grid offsets, and beam or path variations.

[0005] Therefore, a method is needed that can acquire or generate linear time-varying channel matrices, construct block equivalent channel matrices through CPP boundary processing, and select AFDM chirp parameters based on DAFT domain strip energy evaluation, especially for... Parameters are selected to improve the cyclic band energy concentration of the DAFT domain equivalent channel matrix and reduce leakage and truncation losses under the band receiving window. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for selecting AFDMchirp parameters for high-mobility channels based on DAFT domain strip energy evaluation. This method acquires or generates a linear time-varying channel matrix, forms a block equivalent channel matrix through CPP insertion, linear time-varying channel action, and prefix removal, and constructs a DAFT domain equivalent channel matrix. It then evaluates different [channel types / functions] in the AFDM / DAFT domain. The parameters include the cyclic band energy concentration, the minimum equivalent cyclic half-bandwidth required to reach the preset energy threshold, and the corresponding matrix cyclic out-of-band leakage ratio, thus allowing selection of a suitable option for the current link state. .

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] 1. A method for selecting AFDM chirp parameters for high mobility channels based on DAFT domain strip energy evaluation, characterized by comprising the following steps:

[0009] S1, acquire or generate the linear time-varying channel matrix of the high mobility channel;

[0010] S2, Construct AFDM chirp parameter candidates, which include at least different candidates. ;

[0011] S3, targeting candidates Construct the corresponding DAFT domain equivalent channel matrix based on the linear time-varying channel matrix;

[0012] S4. Calculate the DAFT domain strip energy evaluation index based on the DAFT domain equivalent channel matrix. The DAFT domain strip energy evaluation index is used to characterize the concentration of matrix energy in the cyclic strip region and / or the leakage degree outside the strip receiving window. The DAFT domain strip energy evaluation index includes the minimum equivalent cyclic half bandwidth required to reach the preset energy threshold and the matrix cyclic out-of-band leakage ratio.

[0013] S5, Select AFDMchirp parameters from the AFDM chirp parameter candidates based on the DAFT domain band energy evaluation index. ;

[0014] S6, The transmitter and receiver of the AFDM system use the parameters selected in step S5. Perform AFDM modulation, transmission, DAFT demodulation, and equalization detection.

[0015] Further, in step S1, the high-mobility channel includes at least one of the following: aircraft millimeter-wave link, air-to-ground high-speed mobile link, UAV communication link, high-speed vehicle-mounted communication link, and satellite or aerospace platform communication link; the channel information used to construct the linear time-varying channel matrix is ​​obtained by at least one of the following methods: pilot detection, channel estimation, beam training, receiver feedback, historical link performance statistics, and position and velocity-assisted prediction, and is formed into the linear time-varying channel matrix through time-varying channel reconstruction or discrete channel assembly. Further, in step S2, the candidate... Constructed according to at least one of the following methods: uniform grid, non-uniform grid, local search interval, and historical feedback adaptive update interval; Candidates form a finite candidate set And can fix AFDM baseline parameters The actual evaluation set is obtained by merging the candidate sets. ,in, In a finite set of candidates The number of candidate parameters; This means removing duplicate elements from the set, so that the candidate parameters in the resulting actual evaluation set are not repeated.

[0016] Furthermore, in step S3, the construction of the DAFT domain equivalent channel matrix includes: in the matching parameters Construct the DAFT matrix under fixed or preset conditions. Its matrix elements satisfy

[0017]

[0018] in, for 3D DAFT matrix, The chirp parameters of the candidate AFDM to be evaluated. These are fixed or preset matching chirp parameters; Represents the DAFT matrix of the first generation. line, number Column matrix elements; and These are the matrix row index and column index, respectively. ; The AFDM frame length and the dimension of the DAFT matrix; The imaginary unit, It is an exponential function. Pi. In AFDM block transmission scenarios using the chirp periodic prefix (CPP), the block equivalent channel matrix with CPP effect is derived from the linear time-varying channel matrix. CPP insertion matrix and prefix removal matrix Construct and calculate

[0019]

[0020] as well as

[0021] This is the block equivalent channel matrix formed after CPP insertion, linear time-varying channel action, and prefix removal; as a candidate The corresponding DAFT domain equivalent channel matrix; for The conjugate transpose of, where the superscript This indicates the conjugate transpose.

[0022] Furthermore, in step S4, the definition of the DAFT domain band energy evaluation index includes: for candidate The corresponding DAFT domain equivalent channel matrix Define matrix elements Cycle distance For non-negative integer cyclic half-bandwidth Define the cyclic strip support area And calculate the given cyclic half bandwidth. Energy percentage within the lower circulation zone

[0023]

[0024] in, To prevent extremely small positive constants with a denominator of zero;

[0025] Given an energy threshold Define candidates Minimum equivalent cyclic half-bandwidth required to reach the energy threshold

[0026]

[0027] And define a given cyclic half bandwidth The matrix loop out-of-band leakage scaling function

[0028]

[0029] At the minimum equivalent cyclic half-bandwidth, candidate The corresponding matrix loop leakage ratio is

[0030]

[0031] Given a preset receive half bandwidth Define the out-of-band leakage ratio under a fixed receive half-bandwidth.

[0032]

[0033] in, satisfy ; Used to evaluate a given cyclic half bandwidth The energy concentration of the equivalent channel matrix in the lower DAFT domain within the cyclic band support region; Used to evaluate candidates The minimum equivalent cyclic half-bandwidth of the DAFT domain required to achieve the predetermined energy concentration level; Used to evaluate candidates The proportion of remaining matrix energy outside the cyclic band region corresponding to the minimum equivalent cyclic half-bandwidth; and These are used to constitute the evaluation index or selection criteria for the DAFT domain band energy; Used to evaluate the strip cutoff loss under a given reception window.

[0034] Furthermore, in step S5, the AFDM chirp parameter The selection method includes at least one of the following: lexicographical selection criterion, weighted scoring, threshold screening, ranking selection, or joint evaluation of multiple indicators;

[0035] Among them, the lexicographical selection criterion is adopted.

[0036]

[0037] That is, prioritize the selection of the minimum equivalent cyclic half-bandwidth. For smaller candidates, when multiple candidates have the same minimum equivalent cyclic half-bandwidth, further select the matrix cyclic out-of-band leakage ratio. If there are still multiple candidates, choose the one with the fixed AFDM baseline parameters. Choose the candidate with the smaller distance; if they are still the same, select the one with the smaller distance. Candidates with smaller numerical values.

[0038] Furthermore, in step S6, the sending end and the receiving end transmit information through the feedback link, control channel, pre-configuration table, frame header control field, or link control signaling. Candidate index; when the number of candidates is At that time, the candidate index may use no less than Bit representation.

[0039] Furthermore, the method also includes a receive half-bandwidth adaptive configuration step and / or a parameter selection effect evaluation step:

[0040] The receive half-bandwidth adaptive configuration step is as follows: according to the selected... The corresponding minimum equivalent cyclic half bandwidth Determine or update the received half-bandwidth of the strip equalization detection. In strip LMMSE, local window LMMSE, or strip MMSE detection, only the corresponding cyclic strip region of the DAFT domain equivalent channel matrix is ​​retained or preferentially processed.

[0041] The parameter selection effect evaluation step is used to evaluate the effectiveness of AFDM chirp parameters during offline design, link simulation, system testing, or online operation feedback stages. The performance of the selected AFDM scheme was evaluated to verify its validity. The link performance benefits of the parameters; wherein, the performance evaluation includes adopting the AFDM chirp parameters. The AFDM scheme is compared with at least one of fixed-parameter AFDM, OFDM or OTFS, and the evaluation metrics include at least one of BER, out-of-band leakage ratio under fixed receive half bandwidth, minimum equivalent cyclic half bandwidth and detection complexity determined by receive half bandwidth.

[0042] A high-mobility channel AFDM chirp parameter selection system based on DAFT domain strip energy evaluation includes:

[0043] The linear time-varying channel matrix acquisition module is used to acquire or generate the linear time-varying channel matrix of high mobility channels;

[0044] DAFT domain equivalent channel construction module, used for candidate Construct the corresponding DAFT domain equivalent channel matrix based on the linear time-varying channel matrix;

[0045] A strip energy evaluation module is used to calculate a DAFT domain strip energy evaluation index based on the DAFT domain equivalent channel matrix. The DAFT domain strip energy evaluation index is used to characterize the degree of concentration of matrix energy in the cyclic strip region and / or the degree of leakage outside the strip receiving window.

[0046] The AFDM chirp parameter selection module is used to select AFDM chirp parameters based on the DAFT domain band energy evaluation index. ;

[0047] The parameter configuration and transmission / reception module is used to configure the AFDM chirp parameters. Configured at both the sending and receiving ends, and based on the AFDM chirp parameters. Perform AFDM modulation, DAFT demodulation, and equalization detection.

[0048] Optionally, the system further includes one or two of the following modules:

[0049] The receive half-bandwidth configuration module is used to configure the AFDM chirp parameters. The corresponding minimum equivalent cyclic half-bandwidth is determined or the receive half-bandwidth is updated.

[0050] The link performance evaluation module is used to evaluate the performance of links using the AFDM chirp parameters. The performance of the AFDM scheme was evaluated.

[0051] The present invention also provides an electronic device, including a processor, a memory, and a program stored in the memory and executable by the processor, wherein the program, when executed by the processor, implements the method described herein.

[0052] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described thereon.

[0053] Beneficial effects:

[0054] Compared with the prior art, the present invention has at least the following beneficial effects:

[0055] 1. This invention selects AFDM chirp parameters based on the cyclic banded energy distribution of the DAFT domain equivalent channel matrix. This transforms the parameter selection problem into an evaluation problem of minimum equivalent cyclic half-bandwidth and out-of-band leakage ratio. Compared with fixed-parameter AFDM, it can adaptively select a more suitable low-complexity strip reception method based on the current linear time-varying channel matrix and the equivalent channel of the CPP block. parameter.

[0056] 2. This invention calculates a given cyclic half-bandwidth. Energy percentage within the lower circulation zone Candidates Corresponding internal energy percentage Matrix loop leakage ratio and minimum equivalent cyclic half bandwidth It can directly evaluate different Energy concentration of the equivalent channel matrix in the DAFT domain under candidate conditions; out-of-band leakage ratio under fixed receive half-bandwidth. It can be further used to evaluate the truncation loss under a given striped receiving window.

[0057] 3. This invention can employ a finite candidate set search and lexicographical selection criterion to preferentially reduce... , then reduce This eliminates the need to use BER as the primary selection target, reducing computational overhead during the parameter selection phase.

[0058] 4. This invention can improve link performance under low-complexity reception while reducing the complexity of striped reception. This is achieved by selecting a channel that makes the energy of the DAFT domain equivalent channel more concentrated. This invention reduces the minimum equivalent cyclic half-bandwidth required to reach a preset energy threshold, enabling the receiver to use a smaller receiving half-bandwidth while meeting predetermined strip cutoff loss requirements. The complexity of solving the strip matrix approximately increases with... With the growth and the receiver configured with the minimum required equivalent cyclic half-bandwidth, a potential reduction in complexity can be achieved.

[0059] In the example simulation, the average minimum equivalent cyclic half-bandwidth was reduced from 38.0 for the fixed AFDM to the selected... The AFDM has a complexity ratio of 22.5, corresponding to a complexity ratio of approximately The potential complexity is reduced by approximately 65.0%; the minimum equivalent cyclic half-bandwidth corresponding to fixed AFDM parameters in a representative short frame is 41, using the selected... The minimum equivalent cyclic half-bandwidth corresponding to AFDM is 28, and the corresponding complexity ratio is approximately The potential complexity is reduced by approximately 53.4%; in the receiving half-bandwidth... When the SNR is 12 dB, the BER is determined by the fixed AFDM. Reduce to This indicates that the Parameter selection can simultaneously improve DAFT domain strip energy concentration, strip reception complexity, and link error rate performance. Attached Figure Description

[0060] Figure 1 This is a schematic diagram of the overall process of the method of the present invention;

[0061] Figure 2 for Search algorithm flowchart;

[0062] Figure 3 For candidate set based Parameter search and Select a simulation diagram;

[0063] Figure 4 For different DAFT domain equivalent channel matrix amplitude heatmap under parameters;

[0064] Figure 5 This is a comparison chart of BER-SNR performance under low-complexity receivers. Detailed Implementation

[0065] The invention will now be further explained with reference to the accompanying drawings.

[0066] See Figure 1 The overall process of the method of the present invention includes obtaining or generating a linear time-varying channel matrix; constructing... Candidate set; construct different The equivalent channel matrix in the DAFT domain is calculated; the energy percentage within the cyclic band, the leakage ratio outside the cyclic band, and the minimum equivalent cyclic half-bandwidth are calculated; and the DAFT domain strip energy evaluation index is selected. The steps include parameter configuration, AFDM modulation, DAFT demodulation, and equalization detection. Details are as follows:

[0067] First, a linear time-varying channel matrix for high-mobility channels is acquired or generated, and used as the channel input for the search algorithm. This linear time-varying channel matrix can be formed by reconstructing time-varying channel information obtained from pilot detection or channel estimation, or it can be formed by generating discrete dual-selection channel taps from path parameters and assembling them according to a linear convolution input-output relationship.

[0068] Secondly, construction The candidate parameter set is then used to form the actual evaluation set. Let the candidate set be... Fixed AFDM baseline parameters Merge into the candidate set to obtain

[0069]

[0070] Then, for each Construct the DAFT-domain equivalent channel matrix. AFDM uses the DAFT matrix. Modulation and demodulation are performed, among which The chirp parameters selected for this invention, These are fixed or preset parameters. Matrix elements can be represented as...

[0071]

[0072] For AFDM block transmission scenarios using CPP, a CPP insertion matrix is ​​introduced. and prefix removal matrix The CPP insertion matrix is ​​used to map several samples at the end of the AFDM time-domain block to the prefix position, and according to... Phase compensation is performed using relevant secondary chirp phase factors to match the block equivalent channel after prefix removal with the DAFT domain cyclic structure. CPP differs from ordinary cyclic prefixes in that its boundary phase is... This is relevant. At this point, the block equivalent channel matrix with CPP effect is first obtained.

[0073]

[0074]

[0075] The linear time-varying channel matrix is ​​assembled from discrete time-varying channel taps according to the linear convolution input-output relationship; the block equivalent channel matrix with CPP is formed by CPP insertion and prefix removal of the linear time-varying channel matrix, and serves as the block channel input for constructing the DAFT domain equivalent channel matrix.

[0076] Secondly, for each candidate Calculate the zonal energy evaluation index in the DAFT domain. Define the cycle distance. and in a given cyclic half bandwidth Define the cyclic strip support area .based on calculate

[0077]

[0078] in, To prevent extremely small positive constants with a denominator of zero, a given energy threshold is used. Calculate the minimum equivalent cyclic half bandwidth.

[0079]

[0080] And calculate the out-of-band leakage ratio of the matrix loop.

[0081]

[0082] Fixed receive half bandwidth The proportion of external leakage below It can be used to evaluate the strip truncation loss under a given reception window, but it is not a necessary ranking item as a lexicographical selection criterion.

[0083] Finally, based on each candidate corresponding and choose In a preferred approach, a lexicographical selection criterion is used:

[0084]

[0085] That is, priority selection Smaller candidates; if multiple candidates have the same Then choose The smaller candidate; if still the same, choose the one that is the same. Candidates with smaller distances; if still the same, then... The numerical value is determined. The sending and receiving ends can share this information via a feedback link or control signaling. The candidate indexes are used to ensure consistent use of AFDM chirp parameters in subsequent data frames.

[0086] The present invention will be further described below with reference to specific embodiments.

[0087] Example 1: Method for Obtaining Linear Time-Varying Channel Matrix

[0088] In one embodiment, channel information used to construct a linear time-varying channel matrix can be obtained through pilot detection, beam training, sparse path estimation, previous frame channel estimation, receiver feedback, or position-velocity-assisted prediction. Alternatively, link state parameters can be obtained first, and then discrete dual-selection channel taps can be generated by combining discretized configurations such as sampling period, frame length, and delay interpolation kernel, and assembled into a linear time-varying channel matrix according to linear convolution input-output relationships. The link state parameters may include path delay, Doppler frequency shift, and path complex gain to characterize the dual-selection channel, and may further include beam gain or fractional Doppler offset; the path complex gain can be synthesized from the propagation path gain and the beam gains of the transmitter and receiver, or it can be obtained from channel estimation. Link BER feedback can serve as information for performance evaluation or candidate set adjustment. In a typical millimeter-wave link implementation for aircraft, a geometry and beam sensing acquisition method can be used, considering a single millimeter-wave link for a high-speed aircraft, where the transmitter and receiver have given initial positions, velocity vectors, and selectable acceleration vectors, and there is non-zero relative motion. The system carrier frequency can be set to 28 GHz, the sampling rate to 7.68 MHz, and the AFDM frame length to [missing information]. This corresponds to a Doppler grid spacing of 30 kHz. In the example configuration, the transmitter velocity can be 1481.76 km / h, or approximately 411.6 m / s, and the receiver velocity can be 1852.20 km / h, or approximately 514.5 m / s. The relative radial velocities of the two in different path directions are determined by the projection of the velocity vector onto the corresponding path direction, rather than being preset as fixed constants. The path set may include one LoS path, one ground specular reflection path, and several candidate single-hop scattering paths.

[0089] For a Loss-of-Stake (LoS) path, the propagation distance can be... The latency can be For ground-based specular reflection paths, the receiver can be positioned relative to the ground plane. Mirror the image to obtain the mirror point. The equivalent reflection path delay is calculated from the distance from the transmitter to the mirror point. For sparse scattering paths, the path length can be the sum of the distance from the transmitter to the scattering point and the distance from the scattering point to the receiver.

[0090] The relative radial velocity between the transmitter and receiver can be obtained from the transmitter's velocity vector, the receiver's velocity vector, and the path direction projection, and is used to calculate the Doppler frequency shift for each path. For the th The Doppler frequency shift of a path can be determined by the rate of change of path length or the projection of the transceiver velocity onto the path direction, i.e.

[0091]

[0092] At millimeter-wave carrier frequencies, relative radial velocities on the order of hundreds of meters per second can produce significant Doppler frequency shifts. Due to the high speed of the aircraft, the path Doppler typically does not fall precisely on a discrete Doppler grid. , ,when At this time, the path has a fractional Doppler shift and will produce energy leakage in discrete frames of finite length.

[0093] Furthermore, for millimeter-wave narrow-beam links, the path propagation gain is combined with the beam gain. The transmitter and receiver... The beam gains in the path directions are respectively and The effective path complex gain is When beam pointing error exists, the effective gain changes with angle drift, thus affecting the subsequent DAFT domain matrix structure and link BER.

[0094] After obtaining the path delay, Doppler shift, and effective path complex gain, a discrete dual-selection channel tap can be further formed. The continuous-time equivalent impulse response can be expressed as...

[0095]

[0096] After sampling, fractional time delay interpolation kernel is used. Obtain discrete dual-selection channel taps

[0097]

[0098] in, For time sampling index, For discrete time delay tap index, For the first The set of active paths at each sampling time. The discrete dual-selection channel taps can be assembled into a linear time-varying channel matrix according to the linear convolution input-output relationship. And after combining CPP insertion and prefix removal, a block equivalent channel matrix is ​​formed for DAFT domain strip energy evaluation.

[0099] The motion conditions, path types, and path parameters described above are only used to generate or verify linear time-varying channel matrices and to provide input for the construction of subsequent CPP block equivalent channel matrices and DAFT domain equivalent channel matrices. They do not limit the AFDMchirp parameter selection method described in this invention to a specific channel model.

[0100] Example 2: DAFT Domain Feature Analysis

[0101] In one embodiment, the AFDM transmit symbol vector is The modulated signal is The receiver obtains the result after DAFT demodulation. ,in Keep it fixed or preset. Construct the DAFT domain equivalent channel matrix based on the block equivalent channel matrix formed by CPP insertion and prefix removal of the linear time-varying channel matrix. Construct the DAFT domain equivalent channel matrix .

[0102] Because the DAFT matrix contains The second-order chirp phase term, This will change the equivalent location of the path in the DAFT domain. In one implementation, for the first... A path, which can be based on an approximate cyclic offset. Describe its cyclic offset trend, where For normalized Doppler, To normalize the time delay and analyze... The coupling relationship between path delay and Doppler parameters and its impact on path support location. Path support, path overlap, and fractional Doppler leakage are used to explain the variation of the equivalent channel matrix structure in the DAFT domain. The reasons for the change; when When chosen appropriately, the energy of the equivalent channel matrix can be more concentrated in a narrower cyclic band region.

[0103] In this embodiment, the core evaluation metric is the given cyclic half-bandwidth. Energy percentage within the lower circulation zone The minimum equivalent cyclic half-bandwidth required to reach the preset energy threshold and the proportion of matrix cyclic out-of-band leakage at the minimum equivalent cyclic half bandwidth Out-of-band leakage ratio under fixed receive half bandwidth It can be used to evaluate the strip truncation loss under a given receiving window. One-dimensional projection leakage, maximum energy grid ratio, path separability, path overlap, or fractional Doppler leakage can serve as auxiliary indicators to explain changes in the equivalent channel matrix in the DAFT domain. The above core indicators are directly related to the truncation loss and complexity of strip LMMSE receivers, and BER is not required as the basis for selecting the main parameter.

[0104] Figure 4 Showing different Heatmap of the equivalent channel matrix amplitude in the DAFT domain under fixed AFDM parameters. This graph includes a fixed AFDM parameter and representative candidate... Parameters and the selected The amplitude distribution of the DAFT domain equivalent channel matrix under the same normalization method; the color represents the matrix amplitude under the same normalization method, used to display different... Differences in lower cyclic strip support structure, off-diagonal leakage, and minimum equivalent cyclic half-bandwidth. This figure is used to illustrate... The effect of parameters on the DAFT domain matrix structure does not directly represent BER performance.

[0105] Example 3: Candidate Search and Selection

[0106] The candidate search and selection process in this embodiment can be found in [reference]. Figure 2 The figure illustrates the input linear time-varying channel matrix, the construction of the actual evaluation set, and the traversal of candidate evaluations. Construct the DAFT matrix and the DAFT-domain equivalent channel matrix, and calculate... and Select according to lexicographical order. And the process of configuring the selected parameters on the sending and receiving ends.

[0107] In one embodiment, the DAFT matching parameters are fixed or preset, and... Restricted to a finite set of candidates

[0108]

[0109] For example, the number of candidates can be 32, and the candidate range can be... Fixed baseline parameters Add to the candidate set to obtain the actual evaluation set.

[0110]

[0111] For each Perform the following processing:

[0112] 1. Construct the DAFT matrix ;

[0113] 2. Preferably, in the AFDM block transmission scenario employing CPP, the block equivalent channel matrix with CPP effect is derived from the linear time-varying channel matrix. CPP insertion matrix and prefix removal matrix Construction; where the CPP insertion matrix maps several samples at the tail of the AFDM time-domain block to the prefix position, and according to the... Phase compensation is performed using relevant secondary chirp phase factors to match the block equivalent channel after prefix removal with the DAFT domain cyclic structure. CPP differs from ordinary CP in that its boundary phase is... Relevant. First calculate.

[0114]

[0115] Recalculate

[0116]

[0117] in, This is the block equivalent channel matrix that already includes block boundary processing or equivalent estimation results;

[0118] 3. Based on Calculate the DAFT domain band energy evaluation index. For matrix elements... Define the cycle distance For non-negative integer cyclic half-bandwidth Define the cyclic strip support area and calculate

[0119]

[0120] Given an energy threshold ,calculate

[0121]

[0122] and define

[0123]

[0124] Thus, candidates are obtained. Out-of-band leakage ratio of the matrix at the minimum equivalent cyclic half bandwidth

[0125]

[0126] in, To prevent extremely small positive constants with a denominator of zero, Out-of-band leakage ratio under fixed receive half bandwidth It can be calculated separately to evaluate the strip cutoff loss under a given reception window;

[0127] 4. Preferably, the selection criterion is lexicographical order.

[0128]

[0129] Select The selection criterion is based on the minimum equivalent cyclic half-bandwidth. As a priority target, in order to provide a basis for configuring the strip receiving window; in Under the same conditions, compare the out-of-band leakage ratio of the matrix at the minimum equivalent cyclic half-bandwidth. If the parameters are still the same, select the candidate that is closer to the fixed baseline to reduce parameter jumps; if they are still the same, then proceed as follows: Numerical values ​​are determined. Path overlap, out-of-band leakage ratio under a fixed receive half-bandwidth, or BER can be used as supplementary recording, offline verification, or candidate set adjustment information, but do not replace the above-mentioned data based on... and The primary selection criterion.

[0130] Figure 3 For candidate set based Parameter search and Select a simulation graph. This graph shows different... Minimum equivalent cyclic half bandwidth corresponding to the candidate Energy percentage within the circulation zone and the proportion of matrix loop leakage outside the band This is used to illustrate the selection of zonal energy evaluation based on DAFT domain. The process.

[0131] Example 4: Parameter Sharing and Link Reception

[0132] AFDM modulation and demodulation require the transmitter and receiver to use consistent chirp parameters.

[0133] because From a finite candidate set, candidate indices can be transmitted in a practical system rather than consecutive real numbers. The sender and receiver can communicate via feedback links, control channels, pre-configuration tables, frame header control fields, or link control signaling. Candidate index; when the number of candidates is At that time, the candidate index may use no less than This is indicated by the indexing method. For example, when the number of candidates is 32, the indexing overhead is 5 bits. This method is suitable for short-term, stable, high-mobility links, where the duration of a single frame is short and the link state changes little within one or several frames.

[0134] During data transmission, the sending end uses The receiver performs AFDM modulation using fixed or preset parameters, performs DAFT demodulation using consistent parameters, and employs striped LMMSE, local windowed LMMSE, striped MMSE, or other low-complexity equalization detection algorithms. The receiver can also adjust the selected parameters accordingly. The corresponding minimum equivalent cyclic half bandwidth Determine or update the receive half bandwidth It retains or prioritizes processing only the corresponding cyclic band region in the DAFT domain equivalent channel matrix. Because this invention concentrates the energy of the DAFT domain equivalent channel matrix in a narrower cyclic band region, it reduces the receiver's dependence on the large half-bandwidth matrix.

[0135] Example 5: Link Performance Evaluation

[0136] The following uses an aircraft millimeter-wave link as a typical example of a high-mobility channel to verify the effectiveness of the parameter selection in this invention. In one embodiment, during offline design, link simulation, system testing, or online operation feedback phases, the same or comparable aircraft millimeter-wave link states, the same QPSK bitstream, the same random seed, and the same SNR settings are used to evaluate OFDM, OTFS, fixed AFDM, and the selected [parameter selection method]. Link-level comparison is performed on AFDM. OFDM can use CP and frequency domain diagonal MMSE detection; OTFS can use strip LMMSE detection; fixed AFDM uses fixed ; Use the selected The AFDM scheme adopted in this invention is selected based on the DAFT domain band energy evaluation. This embodiment is used to verify the selected... The link performance benefits of the parameters are not a necessary step in the main parameter selection process.

[0137] Example simulation conditions may include: system carrier frequency of 28 GHz, sampling rate of 7.68 MHz, and AFDM frame length of [missing information]. Both the OFDM CP length and the AFDM CPP length are set to 32 sampling points, and the modulation scheme is QPSK. Keep it fixed or preset. The number of candidates is set to 32, and the fixed AFDM baseline parameters are set to... Energy threshold is taken The receiver uses striped LMMSE detection and a fixed receive half-bandwidth. The available number of complete frames is 14. All comparison schemes use the same or comparable linear time-varying channel matrix and CPP configuration, the same QPSK bitstream, the same random seed, and the same SNR settings to evaluate different waveforms or different... Relative performance under parameter configuration.

[0138] Evaluation metrics include BER, out-of-band leakage ratio under a fixed receive half-bandwidth, minimum equivalent cyclic half-bandwidth, and the ratio of receive half-bandwidth to bandwidth. At least one of the determined detection complexities. If a striped matrix is ​​used for solving, the complexity can be approximately proportional to... Growth; therefore, when the present invention is selected When the DAFT domain equivalent channel matrix reaches the preset energy threshold within a smaller cyclic half-bandwidth, the receiver can use a smaller receiving half-bandwidth while meeting the predetermined strip cutoff loss requirements, thereby achieving a potential reduction in complexity.

[0139] Figure 5 : BER-SNR performance comparison chart under low-complexity receivers. This chart shows the OFDM frequency domain diagonal MMSE, OTFS strip LMMSE, fixed AFDM strip LMMSE, and the performance of receivers using the selected... BER performance of AFDM strip LMMSE, AFDM complete MMSE reference lower bound, and QPSK AWGN reference lower bound as a function of SNR; where fixed AFDM and using the selected AFDM can provide different receive half-bandwidths. The following is a strip-shaped LMMSE result. This figure is used to verify the invention described herein. The link performance gains of parameter selection methods under low-complexity reception conditions.

[0140] Numerical results at representative SNR points are shown in Table 1. Table 1 provides the numerical results. Figure 5 Numerical summaries at representative SNR points, where the average minimum equivalent cyclic half-bandwidth is used to characterize the cyclic band energy concentration of the DAFT domain equivalent channel matrix, and BER is used to verify the selected Link performance gains under low-complexity strip reception conditions.

[0141] Table 1. Comparison of structural half-bandwidth and BER performance at representative SNR points

[0142]

[0143] In the example simulation, the minimum equivalent cyclic half-bandwidth corresponding to the fixed AFDM parameters in a representative short frame is 41, using the selected... The minimum equivalent cyclic half-bandwidth corresponding to AFDM is 28; in cross-frame statistics, the average minimum equivalent cyclic half-bandwidth of fixed AFDM is 38.0, using the selected... The average minimum equivalent cyclic half-bandwidth of AFDM is 22.5. When the SNR is 12 dB, the BER of the fixed AFDM is , using the selected The BER of AFDM is The OFDM frequency domain diagonal MMSE maintains a BER of approximately 0.43 under high-speed dual-selection conditions. These results demonstrate that the present invention… Parameter selection can reduce the DAFT domain strip support range and strip truncation loss, and improve link performance under low-complexity reception conditions.

[0144] The above simulation results are only used to illustrate the present invention. The technical effects of parameter selection methods. In practical applications, carrier frequency, sampling rate, frame length, CP or CPP length, candidate set, energy threshold, receive half-bandwidth, receiver type, and channel acquisition method can all be adjusted according to system design requirements.

[0145] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for selecting AFDM chirp parameters for high mobility channels based on DAFT domain strip energy evaluation, characterized in that, Includes the following steps: S1, acquire or generate the linear time-varying channel matrix of the high mobility channel; S2, Construct AFDM chirp parameter candidates, which include at least different candidates. ; S3, targeting candidates Construct the corresponding DAFT domain equivalent channel matrix based on the linear time-varying channel matrix; S4. Calculate the DAFT domain strip energy evaluation index based on the DAFT domain equivalent channel matrix. The DAFT domain strip energy evaluation index is used to characterize the concentration of matrix energy in the cyclic strip region and / or the leakage degree outside the strip receiving window. The DAFT domain strip energy evaluation index includes the minimum equivalent cyclic half bandwidth required to reach the preset energy threshold and the matrix cyclic out-of-band leakage ratio. S5, Select AFDMchirp parameters from the AFDM chirp parameter candidates based on the DAFT domain band energy evaluation index. ; S6, The transmitter and receiver of the AFDM system use the parameters selected in step S5. Perform AFDM modulation, transmission, DAFT demodulation, and equalization detection.

2. The method according to claim 1, characterized in that, In step S1, the high mobility channel includes at least one of the following: aircraft millimeter-wave link, air-to-ground high-speed mobile link, UAV communication link, high-speed vehicle-mounted communication link, and satellite or aerospace platform communication link; the channel information used to construct the linear time-varying channel matrix is ​​obtained by at least one of the following methods: pilot detection, channel estimation, beam training, receiver feedback, historical link performance statistics, and position and velocity-assisted prediction, and is formed by time-varying channel reconstruction or discrete channel assembly to form the linear time-varying channel matrix.

3. The method according to claim 2, characterized in that, In step S2, the candidate Constructed according to at least one of the following methods: uniform grid, non-uniform grid, local search interval, and historical feedback adaptive update interval; Candidates form a finite candidate set And can fix AFDM baseline parameters The actual evaluation set is obtained by merging the candidate sets. ,in, In a finite set of candidates The number of candidate parameters; This means removing duplicate elements from the set, so that the candidate parameters in the resulting actual evaluation set are not repeated.

4. The method according to claim 3, characterized in that, In step S3, the construction of the DAFT domain equivalent channel matrix includes: in the matching parameters Construct the DAFT matrix under fixed or preset conditions. Its matrix elements satisfy in, for 3D DAFT matrix, The chirp parameters of the candidate AFDM to be evaluated. These are fixed or preset matching chirp parameters; Represents the DAFT matrix of the first generation. line, number Column matrix elements; and These are the matrix row index and column index, respectively. ; The AFDM frame length and the dimension of the DAFT matrix; The imaginary unit, It is an exponential function. Pi. In AFDM block transmission scenarios using the chirp periodic prefix (CPP), the block equivalent channel matrix with CPP effect is derived from the linear time-varying channel matrix. CPP insertion matrix and prefix removal matrix Construct and calculate as well as in, This is the block equivalent channel matrix formed after CPP insertion, linear time-varying channel action, and prefix removal; as a candidate The corresponding DAFT domain equivalent channel matrix; for The conjugate transpose of, where the superscript This indicates the conjugate transpose.

5. The method according to claim 4, characterized in that, In step S4, the definition of the DAFT domain band energy evaluation index includes: for candidate The corresponding DAFT domain equivalent channel matrix Define matrix elements Cycle distance For non-negative integer cyclic half-bandwidth Define the cyclic strip support area And calculate the given cyclic half bandwidth. Energy percentage within the lower circulation zone in, To prevent extremely small positive constants with a denominator of zero; Given an energy threshold Define candidates Minimum equivalent cyclic half-bandwidth required to reach the energy threshold And define a given cyclic half bandwidth The matrix loop out-of-band leakage scaling function At the minimum equivalent cyclic half-bandwidth, candidate The corresponding matrix loop leakage ratio is Given a preset receive half bandwidth Define the out-of-band leakage ratio under a fixed receive half-bandwidth. in, satisfy ; Used to evaluate a given cyclic half bandwidth The energy concentration of the equivalent channel matrix in the lower DAFT domain within the cyclic band support region; Used to evaluate candidates The minimum equivalent cyclic half-bandwidth of the DAFT domain required to achieve the predetermined energy concentration level; Used to evaluate candidates The proportion of remaining matrix energy outside the cyclic band region corresponding to the minimum equivalent cyclic half-bandwidth; and These are used to constitute the evaluation index or selection criteria for the DAFT domain band energy; Used to evaluate the strip cutoff loss under a given reception window.

6. The method according to claim 5, characterized in that, In step S5, the AFDM chirp parameters The selection method includes at least one of the following: lexicographical selection criterion, weighted scoring, threshold screening, ranking selection, or joint evaluation of multiple indicators; Among them, the lexicographical selection criterion is adopted. That is, prioritize the selection of the minimum equivalent cyclic half-bandwidth. For smaller candidates, when multiple candidates have the same minimum equivalent cyclic half-bandwidth, further select the matrix cyclic out-of-band leakage ratio. If there are still multiple candidates, choose the one with the fixed AFDM baseline parameters. Choose the candidate with the smaller distance; if they are still the same, select the one with the smaller distance. Candidates with smaller values; In step S6, the sending end and the receiving end transmit information through the feedback link, control channel, pre-configuration table, frame header control field, or link control signaling. Candidate index; when the number of candidates is At that time, the candidate index may use no less than Bit representation.

7. The method according to claim 1, characterized in that, The method further includes a receive half-bandwidth adaptive configuration step and / or a parameter selection effect evaluation step: The receive half-bandwidth adaptive configuration step is as follows: according to the selected... The corresponding minimum equivalent cyclic half bandwidth Determine or update the received half-bandwidth of the strip equalization detection. In strip LMMSE, local window LMMSE, or strip MMSE detection, only the corresponding cyclic strip region of the DAFT domain equivalent channel matrix is ​​retained or preferentially processed. The parameter selection effect evaluation step is used to evaluate the effectiveness of AFDM chirp parameters during offline design, link simulation, system testing, or online operation feedback stages. The performance of the selected AFDM scheme was evaluated to verify its validity. The link performance benefits of the parameters; wherein, the performance evaluation includes adopting the AFDM chirp parameters. The AFDM scheme is compared with at least one of fixed-parameter AFDM, OFDM or OTFS, and the evaluation metrics include at least one of BER, out-of-band leakage ratio under fixed receive half bandwidth, minimum equivalent cyclic half bandwidth and detection complexity determined by receive half bandwidth.

8. A high-mobility channel AFDM chirp parameter selection system based on DAFT domain strip energy evaluation, characterized in that, include: The linear time-varying channel matrix acquisition module is used to acquire or generate the linear time-varying channel matrix of high mobility channels; DAFT domain equivalent channel construction module, used for candidate Construct the corresponding DAFT domain equivalent channel matrix based on the linear time-varying channel matrix; A strip energy evaluation module is used to calculate a DAFT domain strip energy evaluation index based on the DAFT domain equivalent channel matrix. The DAFT domain strip energy evaluation index is used to characterize the degree of concentration of matrix energy in the cyclic strip region and / or the degree of leakage outside the strip receiving window. The AFDM chirp parameter selection module is used to select AFDM chirp parameters based on the DAFT domain band energy evaluation index. ; The parameter configuration and transmission / reception module is used to configure the AFDM chirp parameters. Configured at both the sending and receiving ends, and based on the AFDM chirp parameters. Perform AFDM modulation, DAFT demodulation, and equalization detection; Optionally, the system further includes one or two of the following modules: The receive half-bandwidth configuration module is used to configure the AFDM chirp parameters. The corresponding minimum equivalent cyclic half-bandwidth is determined or the receive half-bandwidth is updated. The link performance evaluation module is used to evaluate the performance of links using the AFDM chirp parameters. The performance of the AFDM scheme was evaluated.

9. An electronic device, characterized in that, It includes a processor, a memory, and a program stored in the memory and executable by the processor, wherein the program, when executed by the processor, implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1 to 7.

Images