A low-complexity OTFS channel estimation method based on windowing processing, a storage medium and a computer

Abstract

A low-complexity OTFS channel estimation method based on windowing processing, a storage medium and a computer are provided, which solve the problems of high complexity and large pilot overhead of the cross-correlation channel estimation method. The method comprises the following steps: constructing a Doppler axis function in a time-delay-Doppler domain equivalent channel processed by a window function, and performing interception to obtain a matching function; setting a threshold coefficient, a Doppler axis resolution, an initial path cycle number and a pilot area; obtaining an estimated time-delay-Doppler domain impulse response matrix according to the pilot area position relationship; performing a cross-correlation operation on the matching function and the impulse response matrix according to the Doppler axis resolution to obtain a maximum cross-correlation value, a Doppler grid point value, a time-delay grid point value and a cross-correlation value thereof; estimating corresponding channel parameters; updating all values on the lth time-delay grid point of the impulse response matrix; setting an iteration stopping condition, and outputting the estimated value when the stopping condition is met. The method is applied to the field of high-speed mobile communication.

Description

Technical Field

[0001] This invention relates to the field of wireless communication, and in particular to a low-computation OTFS channel estimation method based on windowing processing. Background Technology

[0002] High-speed mobile communication is one of the key application scenarios for future 6G. Currently, OFDM technology in 5G systems faces significant challenges in high-speed mobile scenarios. For example, high Doppler shift can disrupt the orthogonality of OFDM subcarriers, thus degrading communication quality. To address the limitations of traditional OFDM modulation in high-speed mobile scenarios, orthogonal time-frequency spatial modulation technology has emerged.

[0003] Orthogonal Time Frequency Space (OTFS) modulation modulates data in the time-delay-Doppler domain, ensuring communication quality in high-speed mobile communication scenarios. However, OTFS suffers from the fractional Doppler problem. Due to the complexity and variability of real-world channels, they typically do not meet the integer Doppler frequency offset condition, posing a significant challenge to channel estimation in OTFS systems.

[0004] For pilot-assisted time-delay-Doppler (OTFS) channel estimation, a commonly used and effective method is the threshold decision method based on embedded pilots. This method is simple and has low pilot overhead, but in the fractional Doppler case, the pilot region is easily affected by interference from surrounding data, leading to false alarms and a sharp drop in channel estimation accuracy. Another effective method is the cross-correlation-based channel estimation method, which has high accuracy but requires cross-correlation operations across the entire Doppler axis, resulting in high computational cost. Furthermore, it occupies an entire frame as a pilot frame, resulting in significant pilot overhead and severely impacting communication efficiency. Summary of the Invention

[0005] This invention addresses the problem that cross-correlation channel estimation methods suffer from high computational complexity and large pilot overhead, severely impacting communication efficiency. It proposes a low-computation OTFS channel estimation method based on windowing processing, the method comprising:

[0006] Step 1: Construct the time-delay-Doppler domain equivalent channel function and set the threshold coefficient Γ. a ,Γ b Doppler axis resolution δ, initial path loop count i, and pilot region;

[0007] Step 2: Add a window function to the time delay-Doppler axis equivalent channel function, and extract the added Doppler axis equivalent channel function to generate a matching function;

[0008] Step 3: Based on the positional relationship of the pilot regions, obtain the estimated time-delay-Doppler domain impulse response matrix.

[0009] Step 4: Update the path loop count;

[0010] Step 5: Perform cross-correlation operation within the channel Doppler extension range based on the time-delay Doppler domain impulse response matrix and the matching function;

[0011] Step 6: Obtain the maximum cross-correlation value according to the cross-correlation operation, and obtain the Doppler grid point value, time delay grid point value and cross-correlation value corresponding to the maximum cross-correlation value;

[0012] Step 7: Estimate the corresponding channel parameters based on the Doppler grid point values, time delay grid point values, and their cross-correlation values;

[0013] Step 8: Update all values ​​at the l-th delay grid point of the delay-Doppler domain impulse response matrix according to the channel parameters;

[0014] Step 9: Set the iteration stop condition. When the stop condition is met, output the estimated value; otherwise, return to step 4 and repeat the above steps.

[0015] Furthermore, a preferred embodiment is provided, wherein the step of adding a window function to the time-delay-Doppler axis equivalent channel function and truncating the added Doppler axis equivalent channel function to generate a matching function is specifically as follows:

[0016] P(x) = Q N ((xk γ -k max ) N ), 0 < x < 2k γ +2k max +1,

[0017] Where P(x) is the matching function, Q N Let x be the Doppler axis equivalent channel function after windowing, and k be the function variable of this function. γ k is the protection interval selected based on the window function. max Here, represents the Doppler extension value, and N represents the number of Doppler axis grid points. N This indicates a cyclic shift.

[0018] Furthermore, a preferred embodiment is provided, wherein the update path loop count is:

[0019] i = i + 1.

[0020] Furthermore, a preferred embodiment is also provided, wherein the low-computation cross-correlation operation based on the time-delay Doppler domain impulse response matrix and the matching function within the channel Doppler extension range specifically comprises:

[0021]

[0022] Among them, among them, For k Δ and k δ Cross-correlation values, For cross-correlation function, k Δ k is an integer Doppler value. δ P is the fractional Doppler value. * This is the conjugate form of the matching function. To represent the obtained time-delay Doppler domain impulse response matrix The value at the l-th delay grid point.

[0023] Furthermore, a preferred method is also provided, wherein estimating the corresponding channel parameters based on the maximum cross-correlation value, Doppler grid point value, time delay grid point value, and cross-correlation value specifically involves:

[0024] Estimated integer and fractional Doppler values ​​for the i-th path:

[0025]

[0026] in, For the estimated integer Doppler values, The estimated fractional Doppler value;

[0027] The estimated delay value of the i-th path

[0028]

[0029] The estimated gain of the i-th path

[0030]

[0031] The estimated phase value of the i-th path:

[0032]

[0033]

[0034] in, M is the phase parameter related to the time delay, and M is the number of time delay axis grid points.

[0035] Furthermore, a preferred embodiment is also provided, wherein updating all values ​​at the l-th time delay grid point of the time delay Doppler domain impulse response matrix according to the channel parameters specifically comprises:

[0036]

[0037] in, Let be all the values ​​at the i-th time-delay grid point of the time-delay Doppler domain impulse response matrix. This refers to the phase parameter related to the time delay.

[0038] Furthermore, a preferred embodiment is provided, wherein the iteration stopping condition is:

[0039]

[0040] Where σ is the noise variance.

[0041] Furthermore, a preferred embodiment is also provided, wherein the setting of the threshold coefficient Γ a ,Γ b The Doppler axis resolution δ is set to 0.1, and the initial path loop count i is set to 0.

[0042] Based on the same inventive concept, the present invention also provides a computer-readable storage medium for storing a computer program that executes a low-computational-load OTFS channel estimation method based on windowing processing as described in any of the preceding claims.

[0043] Based on the same inventive concept, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes a low computational complexity OTFS channel estimation method based on windowing processing as described in any one of the preceding claims.

[0044] The advantages of this invention are:

[0045] This invention solves the problem that cross-correlation channel estimation methods have high computational complexity, large pilot overhead, and seriously affect communication efficiency.

[0046] This invention proposes a low-computation OTFS channel estimation method based on windowing processing. The pilot signal is embedded in the data, and a windowing operation is performed on the time-frequency domain data at the transmitter. The windowing operation makes the channel's equivalent representation in the time-delay-Doppler domain more sparse, which can reduce the interference of surrounding data on the pilot region. In addition, the low-computation OTFS channel estimation method based on windowing processing described in this invention performs cross-correlation operation on a few Doppler grid points, which greatly reduces the computational load and pilot overhead while ensuring the accuracy of channel estimation.

[0047] This invention proposes a low-computation OTFS channel estimation method based on windowing processing. Compared with the traditional OTFS system, the windowed OTFS system can concentrate the Doppler axis energy in the equivalent channel in the main lobe region, thereby enhancing the sparsity of the equivalent impulse response matrix.

[0048] This invention is applied to the field of high-speed mobile communication. Attached Figure Description

[0049] Figure 1 This is a flowchart of a low-computational-load OTFS channel estimation method based on windowing processing, as described in Implementation Method 1.

[0050] Figure 2 This is a schematic diagram of the pilot placement as described in Embodiment 1;

[0051] Figure 3 This is a schematic diagram of the OTFS bit error rate curve described in Implementation Method Nine, where the horizontal axis represents the signal-to-noise ratio and the vertical axis represents the bit error rate. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0053] Implementation Method 1, see [link] Figure 1 and Figure 2 This embodiment describes a low-computational-load OTFS channel estimation method based on windowing processing, the method comprising:

[0054] Step 1: Construct the Doppler axis function in the time-delay-Doppler domain equivalent channel and set the threshold coefficient Γ. a ,Γ b Doppler axis resolution δ, initial path loop count i, and pilot region;

[0055] Step 2: Add a window function to the time delay-Doppler axis equivalent channel function, and extract the added Doppler axis equivalent channel function to generate a matching function;

[0056] Step 3: Based on the positional relationship of the pilot regions, obtain the estimated time-delay Doppler domain impulse response matrix.

[0057] Step 4: Update the path loop count;

[0058] Step 5: Perform low-computation cross-correlation operation within the channel Doppler extension range based on the time-delay Doppler domain impulse response matrix and the matching function;

[0059] Step 6: Obtain the maximum cross-correlation value according to the cross-correlation operation, and obtain the Doppler grid point value, time delay grid point value and cross-correlation value corresponding to the maximum cross-correlation value;

[0060] Step 7: Estimate the corresponding channel parameters based on the Doppler grid point values, time delay grid point values, and their cross-correlation values;

[0061] Step 8: Update all values ​​at the l-th delay grid point of the delay-Doppler domain impulse response matrix according to the channel parameters;

[0062] Step 9: Set the iteration stop condition. When the stop condition is met, output the estimated value; otherwise, return to step 4 and repeat the above steps.

[0063] In practical applications, this implementation embeds the pilot region into the data center location and sets a certain pilot guard interval. On the Doppler axis, the transmitter performs time-frequency domain windowing processing, minimizing interference from surrounding data and facilitating channel estimation. Furthermore, this embedded pilot placement saves pilot resources and better reflects the true channel impact of the current frame. In the channel estimation method, this implementation truncates the matched template function based on the main lobe width of the window function and the Doppler spread value. The truncated function serves as the new matching function P(x). This operation reduces redundancy in traditional cross-correlation channel estimation, performing cross-correlation only within a few Doppler grid points, significantly reducing computational load.

[0064] Implementation Method Two: This implementation method further defines the low-computational-load OTFS channel estimation method based on windowing processing described in Implementation Method One. Specifically, the step of adding a window function to the time-delay-Doppler axis equivalent channel function and truncating the added Doppler axis equivalent channel function to generate a matching function is as follows:

[0065] P(x) = Q N ((xk γ -k max ) N ), 0 < x < 2k γ +2k max +1,

[0066] Where P(x) is the matching function, Q N The Doppler axis equivalent channel function after windowing processing, where x represents the function variable and k is the function variable. γ k is the protection interval selected based on the window function. max Here, represents the Doppler extension value, and N represents the number of Doppler axis grid points.N This indicates a cyclic shift.

[0067] In practice, this implementation uses a window function to process the Doppler axis:

[0068]

[0069] Where, γ ν (n) is the window function added to the time axis in the time-frequency domain data;

[0070] On the Doppler axis, the window function applied at the transmitter is γ. ν (n), Q N (x) is cyclically shifted, shifted by k. γ With Doppler extension value k max The distance between the sums of these two numbers will be 0 < x < 2k. γ +2k max Q within the range of +1 N (x) is the pre-defined cross-correlation function, i.e.:

[0071] P(x) = Q N ((xk γ -k max ) N ), 0 < x < 2k γ +2k max +1.

[0072] This implementation provides a cross-correlation function, which greatly reduces the computational load and pilot overhead in the OTFS channel estimation method.

[0073] Implementation Method 3: This implementation method further defines the low-computational-load OTFS channel estimation method based on windowing processing described in Implementation Method 1. The number of iterations in the update path is:

[0074] i = i + 1.

[0075] This implementation method updates the path count by incrementing it by one each time.

[0076] Implementation Method Four: This implementation method further defines the low-computation OTFS channel estimation method based on windowing processing described in Implementation Method One. The time-delay Doppler domain impulse response matrix and matching function are within the channel Doppler extension range [-k]. max ,k max The low-computation cross-correlation operation is performed within the [internal area], specifically:

[0077]

[0078] in, For k Δ and kδ Cross-correlation values, For matching function P and The cross-correlation function between them, k Δ k is an integer Doppler value. δ N is the fractional Doppler value. z =2k max +2k γ +1, N z The time-delay-Doppler domain impulse response matrix Doppler point P * This is the conjugate form of the matching function. To represent the obtained time-delay Doppler domain impulse response matrix The value at the l-th delay grid point.

[0079] In practical applications, the cross-correlation operation described in this embodiment is specifically as follows:

[0080] Traversing the time-delayed Doppler domain impulse response matrix All time delay grid points, with fractional resolution δ in [-k max ,k max The search is performed within the range of integer parts k of all discrete Doppler frequency shifts. Δ and fractional part k δ .

[0081] This implementation performs cross-correlation operations only on a few Doppler grid points, which greatly reduces the amount of computation.

[0082] Implementation Method 5: This implementation method further defines the low-computational-load OTFS channel estimation method based on windowing processing described in Implementation Method 1. Specifically, the estimation of the corresponding channel parameters based on the Doppler grid values, delay grid values, and their cross-correlation values ​​is as follows:

[0083] Estimated integer and fractional Doppler values ​​for the i-th path:

[0084]

[0085] in, For the estimated integer Doppler values, The estimated fractional Doppler value;

[0086] The estimated delay value of the i-th path

[0087]

[0088] The estimated gain of the i-th path

[0089]

[0090] The estimated phase value of the i-th path:

[0091]

[0092]

[0093] in, Here, M represents the phase parameter related to the time delay, and M is the number of grid points on the time delay axis. This is the phase value.

[0094] Implementation Method Six: This implementation method further defines the low-computation OTFS channel estimation method based on windowing processing described in Implementation Method One. Specifically, updating all values ​​at the l-th delay grid point of the delay-Doppler domain impulse response matrix according to the channel parameters is as follows:

[0095]

[0096] in, Let be all the values ​​at the i-th time-delay grid point of the time-delay Doppler domain impulse response matrix. This refers to the phase parameter related to the time delay.

[0097] Implementation Method Seven: This implementation method further defines the low-computational-load OTFS channel estimation method based on windowing processing described in Implementation Method One. The iteration stopping condition is:

[0098]

[0099] Where σ is the noise variance.

[0100] Implementation Method Eight: This implementation method provides a specific embodiment of the low-computational-load OTFS channel estimation method based on windowing processing described in Implementation Method One, and also serves to explain Implementation Methods Two through Seven. Specifically:

[0101] 1. In the time-delay-Doppler domain, the transmitted data is converted from serial to parallel, the data signals are arranged and pilots are inserted, the pilot area range is set, and zero-filling is performed within the range.

[0102]

[0103] Where, x p Represents pilot frequency, x d Representative data, pilot region filled with 0s. Set pilot guard interval k. b The guard interval is used to prevent interference from surrounding data to the pilot area to a certain extent, and its value range is k.b ≥k max +k γ k γ This represents the guard interval selected based on the Doppler axis time-frequency domain window function.

[0104] 2. The Inverse Symplectic Finite Fourier Transform (ISFFT) converts time-delay-Doppler domain data into time-frequency domain data.

[0105]

[0106] In equation (2), x[k,l] represents the time-delay-Doppler domain data, and X[n,m] represents the time-frequency domain data, where n is the time axis index of the time-frequency domain data and m is the frequency axis index of the time-frequency domain data.

[0107] 3. Perform windowing operation on time-frequency domain data:

[0108]

[0109] Represents the data after windowing in the time and frequency domain, γ v,τ The matrix representing the joint action of the time-delay window function and the Doppler window function, with γ v,τ [n,m] is decomposed into the Doppler window function γ v [n] and the time delay axis window function γ τ [m] that is

[0110] γ v,τ [n,m]=γ v [n]γ τ [m](4)

[0111] 4. Perform a Heisenberg transform to convert the windowed time-frequency domain data into a time-domain signal, which is represented as follows:

[0112]

[0113] In equation (5), s(t) is the time-domain signal transmitted into the channel, and w tx (t-nT) represents the transmitted pulse, Δf represents the subcarrier spacing of the system, and T represents the time length of one OTFS symbol.

[0114] 5. The time-domain signal reaches the receiver after passing through a multipath channel. The received time-domain signal r(t) is represented as:

[0115] r(t)=∫∫h(τ,ν)e j2πν(t-τ) s(t-τ)dτdν+z(t) (6)

[0116] In equation (6), z(t) represents the time-domain form of the noise. After Wigner transform, the received time-domain signal is converted into time-frequency domain data as shown in the following equation:

[0117]

[0118] Y[n,m] represents the time-frequency domain data received at the receiver, w rx (t-nT) represents the received pulse, which should satisfy the biorthogonality characteristic with the transmitted pulse.

[0119] 6. The Symptotic Finite Fourier Transform (SFFT) operation converts time-frequency domain data into time-delay-Doppler domain data y[k,l], and its representation is shown in the following equation:

[0120]

[0121] 7. Channel estimation scheme:

[0122] Step 1: Initialization operation, generate a function of the Doppler axis in the time-delay-Doppler domain equivalent channel, and set the threshold coefficient Γ. a ,Γ b Set the Doppler axis resolution δ, set the initial path loop count i, and add a window function:

[0123]

[0124] On the Doppler axis, the window function applied at the transmitter is γ. ν (n), Q N (x) is cyclically shifted, shifted by k. γ With Doppler extension value k max The distance between the sums of these two numbers will be 0 < x < 2k. γ +2k max Q within the range of +1 N (x) is the preset matching function, i.e.:

[0125] P(x) = Q N ((xk γ -k max ) N ), 0 < x < 2k γ +2k max +1 (10)

[0126] Step 2: Extract the estimated time-delay-Doppler domain impulse response matrix based on the positional relationship of the pilot regions. Its coordinate range in the received delay Doppler domain data frame is k. p -k γ -kmax <k<k p +k γ +k max , l p <l<l p +l max ;

[0127] Step 3: Update the loop count i = i + 1;

[0128] Step 4: Perform cross-correlation operations. Represents the time-delayed Doppler domain impulse response matrix Iterate through all values ​​at the l-th time delay grid point. All time delay grid points, with fractional resolution δ in [-k max ,k max The search is performed within the range of integer parts k of all discrete Doppler frequency shifts. Δ and fractional part k δ :

[0129]

[0130] Step 5: Find the cross-correlation function The maximum value, the integer Doppler lattice value k corresponding to the maximum cross-correlation value. Δ Fractional Doppler lattice point value k δ Time delay grid value l, cross-correlation value

[0131] Step Six: Based on the location and value of the maximum cross-correlation value obtained in Step Five, estimate the following channel parameters:

[0132] Estimated integer and fractional Doppler values ​​for the i-th path:

[0133]

[0134] The estimated delay value of the i-th path:

[0135]

[0136] The estimated gain of the i-th path:

[0137]

[0138] The estimated phase value of the i-th path:

[0139]

[0140]

[0141] Step 7: Update

[0142]

[0143] Step 8: Determine whether to stop iteration. The condition for stopping iteration is:

[0144]

[0145] If the condition for stopping iteration is not met, return to step three to update the loop count and continue iterating. If iteration stops, proceed to step nine to output the estimated value.

[0146] Step 9: Output all integer and fractional Doppler values, delay values, gain, and initial phase values ​​for P paths to complete channel estimation.

[0147] 8. Channel equalization and pilot removal

[0148] Based on the channel parameter values ​​estimated in the above steps, an equalization matrix is ​​obtained. Channel equalization is performed in the time-delay Doppler domain, and pilots are removed according to their positions to obtain the final received data.

[0149] This embodiment also compares the pilot overhead of the low computational complexity OTFS channel estimation method based on windowing processing provided by the present invention with that of existing cross-correlation channel estimation methods under different system conditions, as shown in Table 1:

[0150] Table 1 Comparison of Pilot Overhead

[0151]

[0152] As can be clearly seen from Table 1, the low computational complexity OTFS channel estimation method provided by this invention has low pilot overhead and high communication efficiency.

[0153] Similarly, this embodiment also compares the number of complex multiplications between the low-computation OTFS channel estimation method based on windowing processing provided by the present invention and the existing cross-correlation channel estimation method, as shown in Table 2:

[0154] Table 2 Comparison of the number of complex number multiplications

[0155]

[0156]

[0157] As can be clearly seen from Table 2, the low computational complexity OTFS channel estimation method provided by the invention is far less computationally intensive than the existing cross-correlation channel estimation method.

[0158] Implementation Method Nine, see below Figure 3This embodiment describes a further limitation on the low-computational-load OTFS channel estimation method based on windowing processing described in Embodiment 1. The setting of the threshold coefficient Γ... a ,Γ b The Doppler axis resolution δ is set to 0.1, and the initial path loop count i is set to 0.

[0159] In practical applications, to demonstrate the effectiveness of the low-computational-load OTFS channel estimation method based on windowing processing described in Implementation Method 1, a specific embodiment is provided in conjunction with Implementation Method 8, and the specific implementation steps are as follows:

[0160] 1. Initialization operation: Simulate a single-transmitter single-receiver system, set the number of grid points in the delay-Doppler domain, set M=16, N=16, select the carrier spacing as Δf=15kHz, and select the carrier as f=2GHz.

[0161] A random data stream is generated and QPSK modulated. After modulation, serial-to-parallel conversion is performed, and then pilot signals are inserted, with the pilot guard interval set to k. b =2, the set window main lobe protection interval is k γ =2, pilot placement position is (k p ,l p ) = (8,8).

[0162] 2. The transmitting end performs an ISFFT operation to convert the time-delay-Doppler domain data into time-frequency domain data;

[0163] 3. Apply a windowing operation to the time-frequency domain data by multiplying the time axis of the time-frequency domain data by a window function. For example, a Chebyshev window is selected as the window function.

[0164] 4. The time-frequency domain data is converted into a time-domain signal through the Heisenberg transform;

[0165] 5. The signal passes through a multipath channel;

[0166] 6. After receiving the time-domain signal, the receiving end performs Wigner transform and SFFT operation to convert the data into time-delay-Doppler domain data;

[0167] 7. The low computational complexity OTFS channel estimation method of this invention is adopted:

[0168] Step 1: Obtain the cross-correlation function of the Chebyshev window according to equations (1) and (2), and set the threshold Γ. a =0.1、Γ b =0.1 fractional Doppler resolution is δ=0.1.

[0169] Step 2: Based on the positional relationship of the pilot regions, extract the time-delay Doppler domain impulse response matrix used for estimation.

[0170] Step 3: Update the loop count i = i + 1

[0171] Step 4: Perform cross-correlation operation in [-k max ,k max Within a range, a search is performed to iterate through all integer parts k of the discrete fractional Doppler frequency shift representation at a resolution δ = 0.1. Δ and fractional part k δ .

[0172]

[0173] Step 5: Search The maximum cross-correlation value, and the integer Doppler grid value k used to store the maximum cross-correlation value. Δ Fractional Doppler lattice point value k δ Time delay grid value l, cross-correlation value

[0174] Step 6: Based on the location and value of the maximum cross-correlation value obtained in Step 5, estimate the corresponding channel parameters using equations (12)-(16).

[0175] Step 7: Update the equation using the parameters obtained in Step 6 according to Equation (17).

[0176] Step 8: Determine whether to stop the iteration according to equation (18). If the iteration stops, proceed to step 9; otherwise, return to step 3 to perform the next path search.

[0177] Step 9: The iteration ends, and the channel parameters for all paths are output.

[0178] 8. Perform MMSE equalization, remove the pilot after equalization, and perform QPSK demodulation to obtain the received data.

[0179] like Figure 3 As shown, it can be clearly seen that the low computational complexity OTFS channel estimation method based on windowing processing provided by this invention achieves bit error rate performance similar to existing cross-correlation channel estimation methods, while reducing computational complexity and pilot overhead without sacrificing bit error rate performance.

[0180] Implementation Method 10: A computer-readable storage medium according to this implementation method, the computer-readable storage medium being used to store a computer program, the computer program executing a low-computational-load OTFS channel estimation method based on windowing processing as described in any one of Implementation Methods 1 to 8.

[0181] Implementation Method 11: A computer device according to this implementation method includes a memory and a processor. The memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes a low-computational-load OTFS channel estimation method based on windowing processing as described in any one of Implementation Methods 1 to 8.

[0182] Although preferred embodiments of this disclosure have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this disclosure.

[0183] Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from its spirit and scope. Therefore, if such modifications and variations fall within the scope of the claims of this disclosure and their equivalents, this disclosure is also intended to include such modifications and variations.

[0184] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0185] The technical solutions provided by the present invention have been described in further detail above with reference to the accompanying drawings in order to highlight their advantages and benefits, and are not intended to limit the present invention. Any modifications, combinations, improvements and equivalent substitutions of the present invention based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A low-computational-load OTFS channel estimation method based on windowing processing, characterized in that, The method includes: Step 1: Construct the time-delay-Doppler domain equivalent channel function and set the threshold coefficient. , Doppler axis resolution Initial path loop count and pilot region; Step 2: Add a window function to the time delay-Doppler axis equivalent channel function, and extract the added time delay-Doppler axis equivalent channel function to generate a matching function; Step 3: Based on the positional relationship of the pilot regions, obtain the estimated time-delay-Doppler domain impulse response matrix. ; Step 4: Update the path loop count; Step 5: Perform cross-correlation operation within the channel Doppler extension range based on the time-delay-Doppler domain impulse response matrix and the matching function; Step 6: Obtain the maximum cross-correlation value according to the cross-correlation operation, and obtain the Doppler grid point value, time delay grid point value and cross-correlation value corresponding to the maximum cross-correlation value; Step 7: Estimate the corresponding channel parameters based on the Doppler grid point values, time delay grid point values, and their cross-correlation values; Step 8: Update the delay-Doppler domain impulse response matrix according to the channel parameters. All values ​​on the time delay grid; Step 9: Set the iteration stopping condition. When the stopping condition is met, output the estimated value; otherwise, return to step 4 and repeat the above steps. The process of adding a window function to the time-delay-Doppler axis equivalent channel function and then extracting the added time-delay-Doppler axis equivalent channel function to generate a matching function is as follows: ， in, For matching functions, This is the Doppler axis equivalent channel function after windowing. For the function variable of this function, The protection interval is selected based on the window function. This is the Doppler extension value. N The number of Doppler axis lattice points, Indicates a circular shift; The specific steps involve performing cross-correlation operations within the channel Doppler extension range based on the time-delay-Doppler domain impulse response matrix and the matched function: Traversal All time delay grid points, at fractional resolution exist Iterate through the range to search for the integer part of all discrete Doppler frequency shifts. and fractional part : ， in, for and Cross-correlation values, It is a cross-correlation function. The integer Doppler value. The fractional Doppler value. This is the conjugate form of the matching function. To represent the obtained time-delay-Doppler domain impulse response matrix The The value at each delay grid point; The method further includes: In the time-delay-Doppler domain, the transmitted data is converted from serial to parallel, the data signals are arranged and pilots are inserted, the pilot region range is set, and zero-padding is performed within the range: in, Represents pilot frequency, Representative data, pilot area filled with 0, pilot guard interval set. Its value range is , This represents the protection interval selected based on the Doppler axis time-frequency domain window function; The transmitting end performs an inverse symplectic Fourier transform to convert the time-delay-Doppler domain data into time-frequency domain data: in Data representing the delay-Doppler domain, Represents time-frequency domain data, where n For time axis indexing of time-frequency domain data, m For time-frequency domain data, use the frequency axis index; Windowing operation is performed on time-frequency domain data: This represents the data after windowing in the time and frequency domain. The matrix representing the joint action of the time-delay window function and the Doppler window function will... Decomposed into Doppler window functions With time delay axis window function Right now Performing the Heisenberg transform converts the windowed time-frequency domain data into a time-domain signal, which is represented as follows: in, For time-domain signals transmitted into the channel, To send pulses, The subcarrier spacing of the system, T The time length of an OTFS symbol; The time-domain signal reaches the receiver after passing through a multipath channel, and the receiver receives the time-domain signal. Represented as: in, The noise in its time-domain form is converted into time-frequency domain data by Wigner transform, as shown in the following equation: This represents the time-frequency domain data received. For a pulse to be received, it must satisfy the biorthogonality property with the transmitted pulse; The Sin-Fourier transform operation converts time-frequency domain data into time-delay-Doppler domain data. Its representation is shown in the following formula: ； The estimation of the corresponding channel parameters based on the Doppler grid values, time delay grid values, and their cross-correlation values ​​specifically involves: The estimated first Integer Doppler values ​​and fractional Doppler values ​​for each path: ， in, For the estimated integer Doppler values, The estimated fractional Doppler value; The estimated first The delay value of the path : ， The estimated first Gain of each path : ， The estimated first Phase value of the path: ， ， in, For phase parameters related to time delay, This represents the number of time delay axis grid points; The first step of updating the delay-Doppler domain impulse response matrix according to the channel parameters All values ​​on the time delay grid are as follows: ， in, Let be all the values ​​at the i-th time-delay grid point of the time-delay-Doppler domain impulse response matrix. The phase parameter is related to the time delay; The iteration stopping condition is: ， in, This represents the noise variance.

2. The low-computational-load OTFS channel estimation method based on windowing processing according to claim 1, characterized in that, The number of iterations for the update path is: 。 3. The low-computational-load OTFS channel estimation method based on windowing processing according to claim 1, characterized in that, Setting threshold coefficient , The Doppler axis resolution is set to 0.

1. The initial path loop count is set to 0.

1. It is 0.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program that executes a low-computational-load OTFS channel estimation method based on windowing processing as described in any one of claims 1-3.

5. A computer device, characterized in that: The system includes a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes a low-computational-load OTFS channel estimation method based on windowing processing as described in any one of claims 1-3.

Images