A Joint Estimation Method for Wireless Communication Delay and Frequency Offset Based on Hill Climbing Fine Search

Abstract

This application relates to a joint estimation method for wireless communication delay and frequency offset based on hill-climbing fine search. The method includes: setting a time offset search range, a frequency offset search range, a coarse search step size, and a fine search step size based on prior information; generating a two-dimensional grid point set within the search range according to the coarse search step size; calculating the cost value of each grid point based on a cost function; taking the time offset and frequency offset corresponding to the maximum value as the coarse estimate; using the coarse estimate as the initial point, performing fine search iterations using hill-climbing; calculating the cost value of multiple candidate points in the neighborhood in each iteration, comparing and determining the update direction, and updating the current time offset and / or frequency offset; repeating the iteration until a preset exit condition is met, and outputting the final estimate. This application effectively solves the parameter coupling problem of traditional step-by-step estimation in NTN scenarios, while balancing search range and computational complexity, facilitating engineering implementation, and being applicable to various NTN systems and terminal forms.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a joint estimation method for wireless communication delay and frequency offset based on hill-climbing fine search. Background Technology

[0002] Non-terrestrial networks (NTNs) typically consist of low-Earth orbit (LEO), medium-Earth orbit (MEO), and high-Earth orbit (HEO) satellites, relay platforms, and ground terminals. In this scenario, the long link distance and propagation path between the terminal and the satellite result in a significantly greater round-trip time (RTT) than terrestrial cellular systems. Simultaneously, due to the high-speed motion and orbital changes of the satellites, as well as the movement of the terminals, there is a significant Doppler shift in the link, with frequency offset dynamically changing over time. Furthermore, NTN links are characterized by large coverage areas, tight link budgets, limited terminal power, and wide synchronization acquisition windows, forcing receivers to confront greater time and frequency offset uncertainties during the initial access and tracking phases.

[0003] Traditional methods often estimate the time offset first and then the frequency offset, or vice versa. This approach is susceptible to parameter coupling in NTNs: frequency offset can cause correlation peak broadening, shifting, or amplitude attenuation, thus interfering with time offset detection; time offset errors can introduce phase rotation and observation window mismatch, affecting the stability of frequency offset estimation. In actual signal models, these two factors are not independent; step-by-step estimation leads to error propagation and amplification, resulting in prolonged initial synchronization time, increased acquisition failure rate, increased estimation variance, and degraded subsequent channel estimation and demodulation performance. Especially when large frequency offsets and large time delays coexist, single-parameter separate estimation is more prone to getting trapped in local optima or exhibiting convergence instability. Summary of the Invention

[0004] Therefore, it is necessary to provide a joint estimation method for wireless communication delay and frequency offset based on hill climbing fine search, which can improve synchronization robustness, avoid error propagation, and balance search range and computational complexity in scenarios where large time delay and large frequency offset coexist.

[0005] A joint estimation method for wireless communication delay and frequency offset based on hill-climbing fine search, the method comprising: Step 1: Acquire the received signal and the local reference signal; set the time offset search range, frequency offset search range, coarse search step size, and fine search step size based on prior information; Step 2: Within the time offset search range and frequency offset search range, generate a two-dimensional grid point set according to the coarse search step size, calculate the cost value corresponding to each grid point based on the cost function, and obtain the cost value matrix. Replace the time offset and frequency offset corresponding to the maximum value as the coarse estimate. Step 3: Using the coarse estimate as the initial point, perform fine search iterations using the hill climbing method. Each iteration includes calculating the cost value of multiple candidate points in the neighborhood based on the current time offset and current frequency offset, according to the fine search step size; comparing the cost value of multiple candidate points with the current cost value to determine the update direction of the time offset and / or frequency offset; updating the current time offset and / or current frequency offset according to the update direction to obtain the updated time offset and frequency offset. Step 4: Repeat step 3 until the preset exit condition is met, and output the time offset and frequency offset after the last update as the final estimated value.

[0006] The aforementioned joint estimation method for wireless communication delay and frequency offset based on hill-climbing fine search replaces the traditional step-by-step estimation method by adopting an innovative approach of two-dimensional joint estimation of delay and frequency offset. This eliminates the mutual interference caused by the coupling of time offset and frequency offset parameters at the root of the signal model, completely avoiding the error propagation and amplification effects of step-by-step estimation, and effectively reducing estimation bias and variance. Based on this, a two-level iterative search architecture of grid-based coarse search and hill-climbing fine search is constructed. First, the grid method performs global point-by-point cost calculation and peak capture within a wide search range preset based on prior information such as system design, satellite orbit parameters, terminal velocity, and local oscillator error, quickly locating the approximate range of the true parameters. This solves the core problems of large delay caused by long link distance, large dynamic frequency offset caused by high satellite speed, and wide synchronization capture window in NTN scenarios, significantly improving the initial synchronization robustness in extreme scenarios with both large delay and large frequency offset. To improve robustness and acquisition success rate, this algorithm uses the globally optimal approximation point obtained from a coarse search as the initial starting point. It then employs a hill-climbing method to perform a high-precision neighborhood iterative search within a small local area. In each round, it calculates the cost value of only four neighboring points in the time offset and frequency offset dimensions and updates the parameters in the direction of increasing cost value. Simultaneously, it sets search boundaries to prevent parameters from going out of bounds and becoming invalid. This approach ensures the final estimation accuracy while completely avoiding the computational explosion problem caused by a full-range fine search, perfectly balancing search range and computational complexity. The algorithm has clear logic, controllable computational load, and is easy to implement in engineering. Furthermore, this application reserves various extended optimization interfaces, such as step size layering or adaptive adjustment, dynamic updating of the frequency offset search range combined with real-time prior orbit information, and the introduction of multi-frame accumulation or weighted cost functions to enhance low signal-to-noise ratio performance. These interfaces can flexibly adapt to different NTN systems such as low-Earth orbit / medium-Earth orbit / high-Earth orbit satellites and the differentiated needs of various ground terminals, demonstrating strong scalability and scenario adaptability. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating a method for jointly estimating wireless communication delay and frequency offset based on hill-climbing fine search in one embodiment. Figure 2 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0008] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0009] In one embodiment, such as Figure 1 As shown, a joint estimation method for wireless communication delay and frequency offset based on hill-climbing fine search is provided, including the following steps: Step 1: Acquire the received signal and the local reference signal; set the time offset search range, frequency offset search range, coarse search step size, and fine search step size based on prior information.

[0010] The received signal is the baseband signal received by the terminal through the antenna and processed through down-conversion, analog-to-digital conversion, etc.; the local reference signal is a known reference signal pre-transmitted by the transmitter, such as a demodulation reference signal or a detection reference signal. Based on the system design, satellite orbit parameters, and prior information on terminal velocity and local oscillator error, the time offset search interval is set. Frequency offset search interval And set the corresponding coarse search step size. , With fine search step size , Prior constraints can reduce the invalid search space, eliminate impossible parameter combinations, and significantly improve real-time synchronization.

[0011] Step 2: Within the time-biased search range and the frequency-biased search range, generate a two-dimensional grid point set according to the coarse search step size. Calculate the cost value corresponding to each grid point based on the cost function to obtain the cost value matrix. Use the time-biased and frequency-biased values ​​corresponding to the maximum value as the coarse estimate.

[0012] The two-dimensional grid point set consists of all candidate temporal biases and candidate frequency offset The combination of , where by For step length in Internal value, by For step length in The candidate time offset and frequency offset of each grid point are substituted into the cost function to calculate the corresponding cost value, forming a two-dimensional cost value matrix. Since a larger cost value indicates a higher degree of matching between the candidate parameters and the true parameters, the value corresponding to the maximum cost value is selected. , This serves as a rough estimate. The above process is robust and can quickly locate the approximate range of the true parameters within a large range of parameter uncertainty, thus completing the initial acquisition.

[0013] Step 3: Using the coarse estimate as the initial point, perform fine search iterations using the hill climbing method. Each iteration includes calculating the cost value of multiple candidate points in the neighborhood based on the current time offset and current frequency offset, according to the fine search step size; comparing the cost value of multiple candidate points with the current cost value to determine the update direction of the time offset and / or frequency offset; updating the current time offset and / or current frequency offset according to the update direction to obtain the updated time offset and frequency offset.

[0014] Starting with a coarse estimate, a local iterative search is performed using a smaller fine search step size. Each iteration calculates only the cost value of the four neighborhoods of the current point. By comparing the magnitudes of the neighborhood cost values, the algorithm moves in the direction of increasing cost value, gradually approaching the global optimum. This method performs calculations only in local regions, significantly reducing computational load while maintaining accuracy, thus balancing estimation precision and real-time performance. Furthermore, the hill-climbing fine search also sets search boundaries. When the updated time offset exceeds the time offset search range or the updated frequency offset exceeds the frequency offset search range, the updated time offset or frequency offset is restricted to the corresponding search boundary to prevent parameter out-of-bounds errors that could lead to search failure.

[0015] Step 4: Repeat step 3 until the preset exit condition is met, and output the time offset and frequency offset after the last update as the final estimated value.

[0016] The number of iterations can be pre-configured according to system requirements, or it can adaptively exit based on changes in cost. When the preset exit condition is met, it indicates that the current point has reached a local optimum, and the corresponding time offset and frequency offset are the final estimates. This step ensures that the search process terminates promptly after reaching the required accuracy, avoiding unnecessary computational overhead.

[0017] This application employs a joint estimation architecture—first a coarse search using a grid method, then a fine search using a hill-climbing method—to transform the estimation of two-dimensional parameters for time and frequency offsets into a collaborative optimization process. This fundamentally avoids the interference of initial offset estimation on subsequent offset estimations in traditional methods. The cost function is constructed to include both time and frequency offset variables. During the coarse search phase, the two-dimensional grid points are globally traversed, ensuring that the coarse estimate is already within the joint optimal region near the true parameters. This eliminates the interference of frequency offset-induced correlation peak broadening and shift on time offset detection, as well as the impact of phase rotation introduced by time offset errors on the stability of frequency offset estimation. Subsequently, using the coarse estimate as the initial point, the hill-climbing method is used to simultaneously update the time and frequency offsets in the local neighborhood. Through multiple iterations, the joint maximum value is approximated, avoiding the unidirectional transmission and amplification of errors from earlier to later stages in step-by-step estimation, significantly reducing estimation bias and variance. In NTN scenarios with both large time delays and large frequency offsets, the initial synchronization robustness and acquisition success rate are improved, and the convergence efficiency of the joint estimation is higher. Meanwhile, this application balances a wide search range with low computational complexity through a two-stage coarse and fine step size configuration. The coarse search step size covers a large uncertainty range, ensuring that no true parameters are missed; the fine search step size performs precise local optimization, preventing getting trapped in local optima—making it easier to implement in engineering compared to traditional single-stage search or separate estimation. Furthermore, it allows for flexible configuration of parameters such as step size, search range, and exit conditions, enabling it to adapt to various NTN systems and terminal forms, including low-Earth orbit / medium-Earth orbit / high-Earth orbit satellites, different terminal movement speeds, and local oscillator errors. This provides strong scalability and achieves efficient and reliable synchronization in complex wireless environments.

[0018] In a subsequent embodiment, the cost function is: ; in, In order to receive signals, For local reference signal, for conjugate, It was a time of misfortune. This is frequency offset.

[0019] Specifically, the cost function has a clear physical meaning. The larger the cost value, the higher the matching degree between the candidate combination of time delay and frequency offset and the real channel parameters, and the better the statistical performance.

[0020] In a subsequent embodiment, calculating the cost value corresponding to each grid point includes: Each candidate time bias and candidate frequency offset Substitute into the cost function The cost was calculated. .

[0021] Specifically, for each point in the set of two-dimensional grid points Using these as candidate time delays and frequency offsets, a local reconstruction reference signal is generated. Then, the magnitude of the sum of squares of the differences between the received signal and the locally reconstructed reference signal is calculated, which is the cost value of that grid point. By calculating the cost value of all grid points one by one, the preset parameter search range can be fully covered, ensuring that no area where the real parameters are located is missed.

[0022] In a later embodiment, the coarse search step size includes a time-biased coarse step size. Sum frequency biased coarse step size Fine search step size includes time-biased fine search step size. Sum of frequency offset fine step size .

[0023] Specifically, the coarse search step size is set relatively large to cover the entire preset parameter search range in a short time, quickly narrowing down the possible ranges of parameters; the fine search step size is set relatively small to perform a high-precision search within the local area obtained by the coarse search, ensuring the final estimation accuracy. By setting two levels of step sizes, the relationship between search range, estimation accuracy, and computational complexity can be effectively balanced, avoiding the computational explosion caused by a single fine step size and the insufficient accuracy caused by a single coarse step size.

[0024] In a subsequent embodiment, calculating the cost value of multiple candidate points within the neighborhood includes: Based on the current time offset and current frequency offset Calculate separately: ; in, Decrease the current time bias by a small step. The value calculated in time, Add a small step size to the current time bias. The value calculated in time, Reduce the current frequency offset by a small step. The value calculated in time, Add a fine step to the current frequency offset The cost value calculated at that time.

[0025] Specifically, each iteration only calculates the cost of the current point's four neighboring points in the time and frequency offset dimensions, eliminating the need to traverse the entire local region and significantly reducing computational load. By comparing the cost of these four neighboring points with the cost of the current point, the optimization direction of the parameters can be quickly determined.

[0026] In a subsequent embodiment, determining the update direction of the time offset and / or frequency offset; updating the current time offset and / or current frequency offset according to the update direction includes: Compare and Size: If Then the update is biased towards ;like Then the update is biased towards If they are equal, then the time bias remains unchanged. Compare and Size: If The update frequency offset is then... ;like The update frequency offset is then... If they are equal, the frequency offset remains unchanged.

[0027] Specifically, the update rule is based on the principle of moving in the direction of increasing generation value. That is, if decreasing the time bias by one step results in a higher generation value, the update moves in the direction of decreasing time bias; if increasing the time bias by one step results in a higher generation value, the update moves in the direction of increasing time bias; if both are equal, it indicates that the current time bias has reached the optimal value for that dimension and remains unchanged. The update rule for frequency bias is the same as for time bias. This rule allows for iterative approximation of the peak value of the current generation value, i.e., the location of the true parameters.

[0028] In a later embodiment, the preset exit conditions include: The new generation value obtained in the current iteration At the same time greater than , , , ,and It is greater than the cost of the previous iteration and greater than the cost of the next iteration.

[0029] Specifically, this exit condition ensures that the current point is the peak value point within the local region, meaning the optimal parameter combination has been found. A new value simultaneously greater than the values ​​of its four neighboring regions indicates that the current point has reached optimality in both time and frequency offset dimensions. Values ​​greater than those of the previous and next iterations further verify that the current point is the peak value, preventing premature exit due to value fluctuations. Furthermore, a maximum number of iterations can be preset; when the maximum number of iterations is reached, the system will force an exit to prevent infinite loops.

[0030] In a later embodiment, the hill-climbing fine search also sets a search boundary. When the updated time offset exceeds the time offset search range or the updated frequency offset exceeds the frequency offset search range, the updated time offset or frequency offset is restricted to the corresponding search boundary.

[0031] Specifically, since the coarse search results may be close to the boundary of the search range, parameter updates may exceed the preset search range during the hill-climbing iteration process. By setting search boundaries, parameters that exceed the limits are restricted to boundary values, ensuring that the search process always proceeds within the valid parameter range and avoiding meaningless calculations.

[0032] In a later embodiment, the time offset search range is The frequency offset search range is The time offset search range, frequency offset search range, coarse search step size, and fine search step size are set according to the system design, satellite orbit parameters, terminal velocity, and prior information on local oscillator error.

[0033] Specifically, system design parameters include subcarrier spacing and symbol length; satellite orbit parameters include satellite altitude, orbital inclination, and orbital speed; terminal speed includes the terminal's moving speed; and local oscillator error includes the local oscillator frequency deviation between the transmitter and receiver. By combining this prior information, the search range and step size can be reasonably set, minimizing the search range and improving search efficiency while ensuring that no real parameters are missed.

[0034] In a subsequent embodiment, the fine search step size employs an adaptive step size strategy, gradually decreasing the finer step size during the hill-climbing iteration process. Sum of frequency offset fine step size .

[0035] Specifically, using a larger, finer step size in the early stages of iteration allows for a faster approach to the peak point, improving convergence speed. This larger step size is set based on prior information and practical operational requirements. Gradually decreasing the step size in the later stages of iteration improves the final estimation accuracy. An adaptive step size strategy further balances convergence speed and estimation accuracy, enhancing the overall performance of the algorithm.

[0036] In addition, in other embodiments, the frequency offset search range can be dynamically updated by combining prior orbit information to further reduce the amount of computation; multi-frame cumulative or weighted cost functions can be introduced to enhance the estimation performance under low signal-to-noise ratio.

[0037] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0038] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 2 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When executed by the processor, the computer program implements a joint estimation method for wireless communication delay and frequency offset based on hill-climbing fine-search. The display screen can be an LCD screen or an e-ink display screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0039] Those skilled in the art will understand that Figure 2 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0040] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0041] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0042] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for joint estimation of time delay and frequency offset based on hill-climbing fine search, characterized in that, The method includes: Step 1: Acquire the received signal and the local reference signal; set the time offset search range, frequency offset search range, coarse search step size, and fine search step size based on prior information; Step 2: Within the time offset search range and frequency offset search range, generate a two-dimensional grid point set according to the coarse search step size, calculate the cost value corresponding to each grid point based on the cost function, obtain the cost value matrix, and use the time offset and frequency offset corresponding to the maximum value as the coarse estimate. Step 3: Using the coarse estimate as the initial point, perform fine search iterations using the hill-climbing method. Each iteration includes calculating the cost value of multiple candidate points in the neighborhood based on the current time offset and current frequency offset, according to the fine search step size; comparing the cost value of the multiple candidate points with the current cost value to determine the update direction of the time offset and / or frequency offset; updating the current time offset and / or current frequency offset according to the update direction to obtain the updated time offset and frequency offset. Step 4: Repeat step 3 until the preset exit condition is met, and output the time offset and frequency offset after the last update as the final estimated value.

2. The method of claim 1, wherein, The cost function is: ； wherein is a received signal, is a local reference signal, is a conjugate of is a time offset, is a frequency offset.

3. The method of claim 2, wherein, Calculating the cost value corresponding to each grid point includes: The candidate time offsets and candidate frequency offsets are substituted into the cost function , and the cost values are calculated.

4. The method according to claim 1, characterized in that, The coarse search step size includes the time-biased coarse search step size. Sum frequency biased coarse step size The fine search step size includes a time-biased fine step size. Sum of frequency offset fine step size .

5. The method according to claim 4, characterized in that, The calculation of the cost value of multiple candidate points in the neighborhood includes: Based on the current time offset and current frequency offset Calculate separately: ； in, Decrease the current time bias by a small step. The value calculated in time, Add a small step size to the current time bias. The value calculated in time, Reduce the current frequency offset by a small step. The value calculated in time, Add a fine step to the current frequency offset The cost value calculated at that time.

6. The method according to claim 5, characterized in that, Determine the update direction for time offset and / or frequency offset; update the current time offset and / or current frequency offset according to the update direction, including: Compare and Size: If Then the update is biased. ;like Then the update is biased. If they are equal, then the time bias remains unchanged. Compare and Size: If Then the update frequency offset is ;like Then the update frequency offset is If they are equal, the frequency offset remains unchanged.

7. The method according to claim 6, characterized in that, The preset exit conditions include: The new generation value obtained in the current iteration At the same time greater than , , , ,and It is greater than the cost of the previous iteration and greater than the cost of the next iteration.

8. The method according to claim 1, characterized in that, The hill-climbing fine search also sets a search boundary. When the updated time offset exceeds the time offset search range or the updated frequency offset exceeds the frequency offset search range, the updated time offset or frequency offset is restricted to the corresponding search boundary.

9. The method according to claim 1, characterized in that, The time offset search range is: The frequency offset search range is The time offset search range, the frequency offset search range, the coarse search step size, and the fine search step size are set according to the system design, satellite orbit parameters, terminal velocity, and prior information on local oscillator error.

10. The method according to claim 4, characterized in that, The fine search step size adopts an adaptive step size strategy, gradually decreasing the finer step size during the hill-climbing iteration process. Sum of frequency offset fine step size .

Images