Frequency Domain Adaptive Interference Cancellation Method and System Based on Interference Power Detection

Abstract

This application discloses a frequency-domain adaptive interference cancellation method and system based on interference power detection, relating to the field of wireless communication signal processing. The method includes: performing a frequency-domain transformation on the time-domain baseband received signal to obtain the received frequency-domain coefficients of each subcarrier, and estimating the channel frequency response using pilot information; extracting residual power based on the difference between the total received power and the estimated useful signal power, and constructing a dynamic detection threshold by combining the frequency-domain statistical characteristics of the residual power in the subcarrier's neighborhood to identify and estimate the interference power of each subcarrier; subsequently generating initial mask weights based on the interference power, and performing frequency-domain smoothing through the continuity constraint of adjacent subcarrier weights to obtain a smoothed weight mask; using this mask to perform flexible weighted attenuation on the interfered subcarriers, and reconstructing the time-domain waveform by combining channel equalization and inverse frequency-domain transformation. Thus, while suppressing narrowband or local frequency-domain interference, it reduces false positives on useful signals, improving received signal quality and communication reliability.

Description

Technical Field

[0001] This application relates to the field of wireless communication signal processing technology, and in particular to a frequency domain adaptive interference cancellation method and system based on interference power detection. Background Technology

[0002] In recent years, as wireless communication networks have developed towards higher spectral efficiency, massive connectivity, and multi-service multiplexing, interference problems in wireless communication environments have become increasingly prominent. In complex propagation environments, co-channel interference, adjacent-channel interference, and external electromagnetic interference between communication nodes can superimpose on the useful signal, exhibiting trends of randomness, transientity, and broadband. Such interference degrades the signal quality at the receiver, affects demodulation reliability, and may further weaken the stability of the communication link and the accuracy of system resource scheduling.

[0003] In existing wireless communication systems, interference is typically detected and suppressed using methods such as time-domain processing, frequency-domain processing, adaptive processing, or multi-channel processing. These methods can achieve a certain level of interference suppression when the interference characteristics are relatively stable, the noise environment is well-defined, or the system has sufficient prior conditions. However, in real-world wireless environments, the sources, intensity, and frequency distribution of interference often change over time, making it difficult for the receiver to consistently obtain stable and accurate interference status information. This limits the adaptability of traditional processing methods.

[0004] Furthermore, some advanced interference suppression methods are highly dependent on hardware configuration, reference information, computing resources, or signal environment, making it difficult to fully meet their implementation requirements in single-antenna terminals, mobile communication devices, and resource-constrained equipment. In particular, when dynamic interference and useful signals coexist, if the suppression strategy is too coarse, it can easily affect the quality of the useful signal; if the processing is too complex, it will increase the difficulty of real-time implementation. Summary of the Invention

[0005] This application provides a frequency-domain adaptive interference cancellation method, system, storage medium, computer program product, and electronic device based on interference power detection, which at least solves the problem that the signal quality of wireless communication receivers is difficult to guarantee under dynamic interference environments in the prior art.

[0006] In a first aspect, embodiments of this application provide a frequency-domain adaptive interference cancellation method based on interference power detection. The method includes: performing a frequency-domain transformation operation on a time-domain baseband received signal to obtain received frequency-domain coefficients on each subcarrier, and performing channel estimation processing based on pilot information in the received frequency-domain coefficients to obtain channel frequency response estimates corresponding to each subcarrier; determining the total received power corresponding to each subcarrier based on the received frequency-domain coefficients on each subcarrier, and determining the useful signal power estimate corresponding to each subcarrier based on the channel frequency response estimate; removing the corresponding useful signal power estimate from the total received power corresponding to each subcarrier to obtain residual power on each subcarrier; determining a dynamic detection threshold corresponding to each subcarrier based on the frequency-domain statistical characteristics of the residual power in the neighborhood of each subcarrier, and comparing the residual power on each subcarrier with the corresponding dynamic detection threshold to determine the interference power estimate corresponding to each subcarrier; and so on. The interference power estimate corresponding to each subcarrier determines the initial mask weight for each subcarrier. Based on the weight continuity constraint between adjacent subcarriers, the initial mask weights are smoothed in the frequency domain to obtain a smoothed weighted mask composed of the mask weights corresponding to each subcarrier. The mask weights are located within a preset weight interval, and under the condition of satisfying the weight continuity constraint, the mask weight corresponding to the subcarrier with the larger interference power estimate is closer to the lower limit of the preset weight interval. Based on the mask weights corresponding to each subcarrier in the smoothed weighted mask, the received frequency domain coefficients on each subcarrier are weighted and attenuated to obtain the received frequency domain coefficients after flexible interference attenuation. Based on the channel frequency response estimate, the received frequency domain coefficients after flexible interference attenuation are subjected to channel equalization to obtain the frequency domain signal after interference cancellation. An inverse frequency domain transform is performed on the frequency domain signal after interference cancellation to reconstruct the time domain waveform after interference cancellation.

[0007] Secondly, embodiments of this application provide a frequency-domain adaptive interference cancellation system based on interference power detection, deployed at a wireless communication receiver. The system includes: a frequency-domain transformation estimation unit, configured to perform a frequency-domain transformation operation on the time-domain baseband received signal to obtain received frequency-domain coefficients on each subcarrier, and perform channel estimation processing based on pilot information in the received frequency-domain coefficients to obtain channel frequency response estimates corresponding to each subcarrier; a residual power calculation unit, configured to determine the total received power corresponding to each subcarrier based on the received frequency-domain coefficients on each subcarrier, and determine the useful signal power estimate corresponding to each subcarrier based on the channel frequency response estimate, and remove the corresponding useful signal power estimate from the total received power corresponding to each subcarrier to obtain the residual power on each subcarrier; and an interference power detection unit, configured to determine a dynamic detection threshold corresponding to each subcarrier based on the frequency-domain statistical characteristics of the residual power in the neighborhood of each subcarrier, and compare the residual power on each subcarrier with the corresponding dynamic detection threshold to determine the interference power corresponding to each subcarrier. The system comprises: an estimated value; a weighted mask generation unit, configured to determine the initial mask weights corresponding to each subcarrier based on the estimated interference power value corresponding to each subcarrier, and to perform frequency domain smoothing processing on the initial mask weights based on the weight continuity constraint between adjacent subcarriers, so as to obtain a frequency domain smoothed weighted mask composed of the mask weights corresponding to each subcarrier; wherein the mask weights are located within a preset weight interval, and under the condition of satisfying the weight continuity constraint, the mask weights corresponding to the subcarriers with larger estimated interference power values ​​are closer to the lower limit of the preset weight interval; a weighted equalization processing unit, configured to perform weighted attenuation on the received frequency domain coefficients on each subcarrier based on the mask weights corresponding to each subcarrier in the frequency domain smoothed weighted mask, so as to obtain the received frequency domain coefficients after flexible interference attenuation, and to perform channel equalization processing on the received frequency domain coefficients after flexible interference attenuation based on the channel frequency response estimate, so as to obtain the frequency domain signal after interference cancellation; and a waveform inverse transformation unit, configured to perform inverse frequency domain transformation on the frequency domain signal after interference cancellation, so as to reconstruct the time domain waveform after interference cancellation.

[0008] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the frequency domain adaptive interference cancellation method based on interference power detection according to any embodiment of the present application.

[0009] Fourthly, embodiments of this application provide a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the frequency domain adaptive interference cancellation method based on interference power detection according to any embodiment of this application.

[0010] Fifthly, embodiments of this application provide a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the frequency domain adaptive interference cancellation method based on interference power detection according to any embodiment of this application.

[0011] The frequency-domain adaptive interference cancellation method and system based on interference power detection provided in this application can achieve at least the following technical effects: By estimating and removing the useful signal power contribution from the received frequency domain coefficients, subsequent interference judgment no longer directly relies on the total received power of each subcarrier, but instead shifts to the residual power reflecting anomalous energy components. Based on this, a dynamic detection threshold is determined by combining the frequency domain statistical characteristics of the residual power in the neighborhood of each subcarrier, enabling the interference power estimation to adaptively adjust with changes in the local frequency domain environment. This reduces the impact of useful signal amplitude fluctuations, channel frequency response differences, and changes in local noise floor on interference judgment, making the interference power estimation results more accurately correspond to the actual anomalous energy distribution on each subcarrier, thereby improving the detection reliability in dynamic frequency domain interference scenarios.

[0012] Furthermore, corresponding mask weights are generated based on the estimated interference power of each subcarrier, and frequency domain smoothing is performed by constraining the continuity of weights between adjacent subcarriers. This allows the interference suppression strength to continuously change with the magnitude of the interference power, rather than abruptly cutting off or uniformly attenuating subcarriers identified as interference. Since the obtained frequency domain smoothing weight mask takes into account both the difference in interference strength and the spectral continuity between adjacent subcarriers, it can achieve effective attenuation in strong interference regions, create a smooth transition in interference edge regions, and retain useful signal components as much as possible on weak interference or normal subcarriers. This reduces signal distortion caused by frequency domain suppression operations and improves the fidelity of the frequency domain signal after interference cancellation and the stability of equalization processing.

[0013] This technical solution uses residual power as the basis for characterizing the interference state and a frequency-domain smoothing weighted mask as a flexible attenuation control method. This allows the interference detection results to form a continuous correspondence with the subsequent interference cancellation strength, thus unifying interference identification and interference cancellation into a subcarrier-level adaptive frequency domain processing process. Consequently, the receiver can dynamically adjust its suppression strategy according to the actual frequency domain interference distribution, enhancing its adaptability to time-varying, localized, and broadband interference while avoiding unnecessary damage to useful signals from excessive suppression. This improves the anti-interference performance and link stability of the wireless communication receiver with lower hardware dependence and more controllable computational complexity. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A flowchart illustrating an example of a frequency-domain adaptive interference cancellation method based on interference power detection according to an embodiment of this application is shown. Figure 2 A flowchart illustrating an example of determining the interference power estimate corresponding to each subcarrier in a method according to an embodiment of this application is shown. Figure 3 A flowchart illustrating an example of determining a frequency domain smoothing weight mask composed of mask weights corresponding to each subcarrier in a method according to an embodiment of this application is shown. Figure 4 A schematic diagram illustrating the system operation mechanism of an example frequency-domain adaptive interference cancellation method based on interference power detection according to an embodiment of this application is shown. Figure 5 A schematic diagram showing the experimental simulation effect comparison between the frequency domain adaptive interference cancellation method based on interference power detection according to an embodiment of this application and a comparison baseline; Figure 6 The diagram shows the experimental simulation results comparing the frequency domain mask response and error vector magnitude of different methods under the sudden broadband interference scenario; Figure 7 A simulation effect diagram is shown as an example of boundary quantification assessment of engineering implementation under extreme constraints according to the method of embodiments of this application; Figure 8 A structural block diagram of an example of a frequency-domain adaptive interference cancellation system based on interference power detection according to an embodiment of this application is shown. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] It should be noted that in frequency domain receiving systems such as OFDM (Orthogonal Frequency Division Multiplexing) and SC-FDE (Single-Carrier Frequency Domain Equalization), current technologies often reduce the impact of narrowband interference on the received signal by identifying the frequency of interference and performing zeroing or hole-filling on the corresponding DFT (Discrete Fourier Transform) coefficients. This method is structurally simple, but typically relies on the matching relationship between the interference frequency and the DFT frequency grid. When the interference frequency falls between grid lines, the interference energy may diffuse to adjacent subcarriers. Expanding the zeroing range may weaken useful signal components; if only single-point processing is performed, some interference may remain. To address this issue, some studies have proposed SC-FDE fine frequency shift processing, which involves shifting the interference frequency slightly to bring it closer to the target frequency grid before suppression. However, this approach typically requires accurate interference frequency estimation and may introduce additional phase changes and implementation complexity. Furthermore, its processing stability is still limited when the interference frequency changes rapidly or the interference power approaches the useful signal power.

[0018] In the field of adaptive filtering and signal decomposition, common techniques include LMS (Least Mean Square), RLS (Recursive Least Squares), LS (Least Squares), and Kalman Filtering, used to estimate interference components in the time or subband domains. These methods are applicable when the interference statistics are relatively stable or the model assumptions are sufficient. However, they may encounter problems such as decreased convergence speed, unstable parameter estimation, or increased computational cost when dealing with broadband non-stationary interference, uneven input spectra, or signal and interference power being close. Other studies employ mode decomposition methods such as ASEI-VMD (Adaptive Singular Envelope Iterative Variational Mode Decomposition) to extract interference modes from multi-channel signals; however, these methods typically rely on channel amplitude-phase consistency and mode separation effectiveness. Their applicability is limited when low-frequency interference overlaps with useful signal components or when the terminal lacks stable multi-channel reception capabilities.

[0019] In terms of interference detection, current technologies include Energy Detection (ED), Matched Filter (MF), Cyclostationary Feature Detection (CFD), Eigenvalue Detection (EVD), Fourth-Order Cyclic Cumulant (FOCC), and Constant False Alarm Rate (CFAR). ED has a simple structure but is sensitive to background noise power estimation; MF relies on prior knowledge of the waveform; CFD and EVD typically require long observation sequences or high matrix operation overhead; FOCC can detect co-channel or adjacent-channel interference using high-order cyclic statistical features, but its computational complexity is relatively high; CFAR can adjust the detection threshold using neighborhood statistical information, but it usually focuses more on the detection decision itself, and subsequent interference suppression still needs to be further implemented in conjunction with the specific receiving and processing link.

[0020] Furthermore, AIC (Adaptive Interference Cancellation) architectures typically rely on independent noise reference channels or interference reference signals, generating cancellation components by estimating the relationship between the reference channel and the main receiving channel. While this approach is effective in audio noise reduction, multi-sensor acquisition, or scenarios with reference antennas, it is often difficult for the receiver to obtain an independent reference source that is highly correlated with the target interference and contains no useful signal in cellular communications, ad hoc wireless networks, UAV communications, and lightweight terminal devices. Additionally, additional reference channels incur overhead in hardware synchronization, channel calibration, and storage / computation. Therefore, in dynamic wireless propagation environments, current technologies still have room for improvement in interference detection, suppression, and engineering deployment adaptability.

[0021] It should be understood that the above description of the relevant technologies is intended only to help the public better understand the inventive spirit and motivation of this application, and is not intended to limit this application. Furthermore, the technical solutions described in the above-mentioned relevant technologies are not prior art, and may also be undisclosed technical solutions, such as those under research or in the laboratory stage.

[0022] The technical solutions in this application, including the collection, storage, use, processing, transmission, provision, and disclosure of users' personal information, comply with relevant laws and regulations and do not violate public order and good morals.

[0023] Figure 1 A flowchart illustrating an example of a frequency-domain adaptive interference cancellation method based on interference power detection according to an embodiment of this application is shown.

[0024] Regarding the execution subject of the method in the embodiments of this application, it can be any controller or processor with computing or processing capabilities, such as the baseband signal processing controller in a wireless communication receiver. It implements the method of the embodiments of this application by running programs or instructions stored in a storage medium. In some examples, it can be integrated into an electronic device or terminal through software, hardware, or a combination of both, and the type of terminal or electronic device can be diverse.

[0025] like Figure 1As shown, this embodiment provides a frequency-domain adaptive interference cancellation method based on interference power detection, which is applied to a wireless communication receiver. The wireless communication receiver can be a base station receiver, a terminal device, a drone communication receiver, an industrial IoT node, or other devices with wireless baseband reception and processing capabilities. After obtaining the time-domain baseband received signal, the receiver converts the signal to the frequency domain and analyzes the channel response, received power, residual power, and interference power at the subcarrier level. Then, it generates a frequency-domain smoothing weight mask based on the interference power to perform flexible interference attenuation on the received frequency-domain signal.

[0026] Compared to directly suppressing the received signal as a whole in the time domain, frequency domain processing can decompose different frequency components of the broadband signal onto corresponding subcarriers, allowing the receiver to perform differentiated processing based on the interference level of each subcarrier. Therefore, this embodiment can complete interference detection and suppression processing based on the frequency domain power distribution and channel estimation results of the received signal itself, without relying on an independent external interference reference channel, thereby improving reception stability and signal fidelity in complex wireless environments.

[0027] In step S110, a frequency domain transformation operation is performed on the time-domain baseband received signal to obtain the received frequency domain coefficients on each subcarrier, and channel estimation processing is performed based on the pilot information in the received frequency domain coefficients to obtain the channel frequency response estimate corresponding to each subcarrier.

[0028] Specifically, after completing radio frequency reception, down-conversion, filtering, analog-to-digital conversion, synchronization, and necessary baseband preprocessing, the wireless communication receiver obtains a discrete time-domain baseband received signal. This time-domain baseband received signal typically contains the useful communication signal, background noise, and interference components from other wireless nodes or the external electromagnetic environment. After performing a frequency domain transformation operation on the time-domain baseband received signal, the receiver obtains received frequency domain coefficients on multiple subcarriers. Each received frequency domain coefficient characterizes the complex reception state on the corresponding subcarrier; its amplitude is related to the received energy on that subcarrier, and its phase is related to the phase response on that subcarrier.

[0029] After obtaining the received frequency domain coefficients, the receiver further identifies pilot information from them. Pilot information is typically a known reference symbol or sequence pre-agreed upon by the transmitter and receiver. Based on the correspondence between the received pilot components and the locally known pilot information, the receiver can estimate the amplitude and phase changes of the wireless channel at the corresponding frequency points and obtain the channel frequency response estimate for each subcarrier.

[0030] Here, the receiver obtains the received frequency domain coefficients on each subcarrier and the channel frequency response estimate corresponding to each subcarrier, converts the time-domain mixed received signal into frequency domain data suitable for subcarrier-level analysis, and obtains frequency domain response information that can characterize the channel impact, thereby improving the clarity and distinguishability of the subsequent frequency domain processing objects.

[0031] In step S120, the total received power corresponding to each subcarrier is determined according to the received frequency domain coefficients on each subcarrier, and the estimated useful signal power corresponding to each subcarrier is determined based on the channel frequency response estimate. The estimated useful signal power is then removed from the total received power corresponding to each subcarrier to obtain the residual power on each subcarrier.

[0032] Specifically, for any subcarrier, the receiver can determine the corresponding total received power based on the received frequency domain coefficients on that subcarrier. This total received power reflects the overall received energy on that subcarrier, which may include the energy of the useful signal after propagation through the channel, as well as background noise, adjacent channel leakage, co-channel interference, or other abnormal frequency domain energy. If the presence of interference is determined solely based on the total received power, inaccurate detection results are likely to occur when the channel gain is high, the useful signal power is strong, or the background energy in a local frequency band is elevated.

[0033] In this embodiment, the useful signal power estimate for each subcarrier is determined based on the channel frequency response estimate. This useful signal power estimate characterizes the power contribution that the useful signal may make on the corresponding subcarrier under the current channel conditions. Subsequently, the receiver removes the corresponding useful signal power estimate from the total received power corresponding to that subcarrier to obtain the residual power on that subcarrier, which characterizes the remaining energy components other than the useful signal power contribution.

[0034] This avoids simply misinterpreting the increased energy of the useful signal itself as interference energy. This step decomposes the received frequency domain signal from the perspective of power composition, changing the detection basis from the total received power to the residual power after removing the contribution of the useful signal. This reduces the impact of normal channel fluctuations and changes in the power of the useful signal on the judgment of power anomalies, and improves the pertinence of residual energy characterization.

[0035] In step S130, based on the frequency domain statistical characteristics of the residual power in the neighborhood of each subcarrier, the dynamic detection threshold corresponding to each subcarrier is determined, and the residual power on each subcarrier is compared with the corresponding dynamic detection threshold to determine the estimated interference power value corresponding to each subcarrier.

[0036] Specifically, in real-world wireless communication environments, the background noise levels and residual power distributions across different frequency bands may not be consistent. For example, in industrial scenarios, dense wireless access scenarios, or scenarios with multiple concurrent devices, certain frequency bands may experience localized background elevation or short-term abnormal fluctuations. If a uniform fixed threshold is applied to all subcarriers, false detections may occur in frequency bands with high background noise, while missed detections may occur in frequency bands with low background noise but weak interference. Therefore, this embodiment determines the corresponding dynamic detection threshold for each subcarrier by utilizing the frequency domain statistical characteristics of the residual power within its neighborhood, allowing the detection threshold to be adjusted according to changes in the local frequency domain environment.

[0037] Here, frequency domain statistical features can be used to characterize the background power level, local power fluctuation state, or abnormal energy distribution trend in the neighborhood of the target subcarrier. After determining the dynamic detection threshold based on these statistical features, the receiver compares the residual power on the target subcarrier with the corresponding dynamic detection threshold. When the residual power exceeds the corresponding dynamic detection threshold, it can be determined that the subcarrier is subject to interference, and the interference power estimate is determined based on the degree of abnormality of the residual power; when the residual power does not exceed the corresponding dynamic detection threshold, it can be determined that the subcarrier is not subject to significant interference, and the corresponding interference power estimate is set to zero or a low value.

[0038] Therefore, interference power estimation results can be obtained at the subcarrier level. Compared with the fixed threshold detection method, the dynamic detection threshold can better adapt to local noise floor fluctuations and frequency domain power distribution differences; compared with the method that only outputs the presence or absence of interference, this embodiment can further characterize the interference strength on different subcarriers and improve the adaptability of interference detection results to changes in the local frequency domain environment.

[0039] In step S140, initial mask weights for each subcarrier are determined based on the estimated interference power values ​​for each subcarrier. Then, the initial mask weights are smoothed in the frequency domain based on the weight continuity constraint between adjacent subcarriers to obtain a frequency-smoothed weighted mask composed of the mask weights for each subcarrier. The mask weights are located within a preset weight interval, and under the condition of satisfying the weight continuity constraint, the mask weights corresponding to subcarriers with larger estimated interference power values ​​are closer to the lower limit of the preset weight interval.

[0040] Specifically, after obtaining the interference power estimate for each subcarrier, the receiver determines the initial mask weight for each subcarrier based on the interference power estimate. This initial mask weight represents the degree of preservation or attenuation of the corresponding subcarrier in frequency domain processing. Generally, when the interference power estimate for a subcarrier is low, the initial mask weight for that subcarrier is high to preserve useful signals as much as possible; when the interference power estimate for a subcarrier is high, the initial mask weight for that subcarrier is low to enhance the suppression of anomalous energy. Therefore, the interference suppression process no longer uses a simple preservation or zeroing method, but rather continuously adjusts according to the interference intensity.

[0041] Meanwhile, considering that interference energy may have a certain degree of diffusion in the frequency domain, a sharp change in the weights on adjacent subcarriers may cause frequency domain discontinuity and adversely affect waveform recovery. Therefore, this embodiment further performs frequency domain smoothing on the initial mask weights based on the weight continuity constraint between adjacent subcarriers, making the weight changes between adjacent or neighboring subcarriers more gradual.

[0042] After frequency domain smoothing, the mask weights corresponding to each subcarrier collectively constitute a frequency domain smoothing weight mask. These mask weights lie within a preset weight range to limit the range of weighted attenuation and avoid unconstrained amplification or excessive suppression of the frequency domain coefficients. Under the condition of satisfying the weight continuity constraint, the subcarrier with the larger the estimated interference power value, the closer its corresponding mask weight is to the lower limit of the preset weight range. Therefore, while suppressing interference energy, it is possible to reduce the useful signal loss, frequency domain abrupt changes, and waveform distortion that may be caused by traditional hard nulling.

[0043] In step S150, the received frequency domain coefficients on each subcarrier are weighted and attenuated based on the mask weights corresponding to each subcarrier in the frequency domain smoothing weight mask to obtain the received frequency domain coefficients after flexible interference attenuation. Then, based on the channel frequency response estimate, the received frequency domain coefficients after flexible interference attenuation are processed by channel equalization to obtain the frequency domain signal after interference cancellation.

[0044] Specifically, the receiver applies a frequency-domain smoothing weight mask to the received frequency-domain coefficients. For subcarriers with higher interference power estimates, the corresponding mask weight is lower, resulting in stronger attenuation of the received frequency-domain coefficients on those subcarriers. Conversely, for subcarriers with lower interference power estimates or no detected significant interference, the corresponding mask weight is higher, allowing for better preservation of the received frequency-domain coefficients on those subcarriers. Through this weighted attenuation process, the receiver can apply different levels of suppression to subcarriers with varying degrees of interference in the frequency domain, without needing to apply a uniform hard truncation method to all subcarriers within a frequency band.

[0045] After obtaining the received frequency domain coefficients after flexible interference attenuation, the receiver performs channel equalization based on the channel frequency response estimate. This equalization process compensates for the amplitude attenuation and phase rotation caused by the wireless propagation channel to the useful signal, making the frequency domain signal after interference attenuation closer to the expected received symbol state. Because the interference attenuation process reduces some abnormal frequency domain energy, the channel equalization process can be performed on a more stable frequency domain input.

[0046] In this embodiment, coordinated processing between frequency domain interference attenuation and channel compensation is implemented. A frequency domain smoothing weight mask is used to flexibly attenuate the received frequency domain coefficients according to the degree of subcarrier interference, while channel equalization is used to compensate for transmission distortion caused by the wireless channel. Thus, while reducing abnormal frequency domain energy, the equalization and recovery capability of the useful signal is maintained, thereby improving the quality of the frequency domain signal after interference cancellation.

[0047] In step S160, an inverse frequency domain transformation is performed on the frequency domain signal after interference cancellation to reconstruct the time domain waveform after interference cancellation.

[0048] Specifically, after weighted attenuation and channel equalization, the frequency domain signal obtained by the receiver after interference cancellation is still represented in the subcarrier domain. In order to obtain a signal representation consistent with the time-domain baseband receiving link, the receiver performs an inverse frequency domain transformation on the frequency domain signal, converting it back into a time-domain representation. For example, this inverse frequency domain transformation can be implemented using the inverse transformation method corresponding to the frequency domain transformation in step S110, thereby obtaining the time-domain waveform after interference cancellation.

[0049] It should be noted that, since the aforementioned processing has applied continuous weighted attenuation to different subcarriers in the frequency domain and combined with channel frequency response estimation to complete channel equalization, the reconstructed time-domain waveform has lower interference residue and better waveform continuity compared to the original time-domain baseband received signal. Thus, a complete signal processing procedure is completed from time-domain input, frequency-domain detection and suppression to time-domain output, enabling the receiver to obtain the time-domain baseband waveform after frequency-domain adaptive interference cancellation.

[0050] Regarding the implementation details of obtaining the channel frequency response estimate through channel estimation processing in the method of this application embodiment, in some examples of this application embodiment, the pilot subcarrier coefficients located at the preset pilot position set in the received frequency domain coefficients are extracted, and the initial channel response at each pilot position is obtained by combining the locally known pilot symbols.

[0051] Specifically, after completing the frequency domain transformation, the receiver can extract the pilot subcarrier coefficients from the received frequency domain coefficients on each subcarrier according to the pilot positions pre-configured in the communication protocol. Since the pilot symbols are known to the receiver, the receiver can estimate the channel response at the pilot position based on the correspondence between the received frequency domain coefficients on the pilot subcarriers and the locally known pilot symbols. This initial channel response reflects the amplitude attenuation and phase rotation of the wireless channel on the useful signal at the pilot position, and is the basis for subsequent full-band channel frequency response estimation.

[0052] In some examples, the preset pilot position set is represented as , No. The received frequency domain coefficients on each pilot subcarrier are expressed as follows: The locally known pilot symbol is represented as The initial channel response at the pilot position can be expressed as: Equation (1) In the formula, Indicates the first Initial channel response at each pilot position Indicates the first The received frequency domain coefficients at each pilot position Indicates the first Locally known pilot symbols at each pilot location This represents a preset set of pilot locations. Through the above processing, channel observation results at the pilot locations can be obtained without the need for an additional reference channel.

[0053] Then, multiple interpolation filtering windows with different frequency domain span parameters are configured, and frequency domain reconstruction of the initial channel response is performed using each interpolation filtering window to generate candidate channel frequency response sequences at multiple different interpolation resolution scales.

[0054] Specifically, since pilots are typically deployed only at some subcarrier locations and cannot directly cover all data subcarriers, it is necessary to reconstruct the channel frequency response across the entire frequency band based on the initial channel response at the pilot locations. Interpolation filtering windows with different frequency domain span parameters correspond to different reconstruction scales: interpolation filtering windows with smaller frequency domain spans are more sensitive to local channel variations and can better characterize local fluctuations when frequency-selective fading is strong; interpolation filtering windows with larger frequency domain spans have stronger smoothing capabilities and can reduce the impact of pilot noise or local anomalies on the channel estimation results.

[0055] In some examples, the first The interpolation reconstruction operation at each interpolation resolution scale is represented as follows: The corresponding candidate channel frequency response sequence can be expressed as: Equation (2) In the formula, Indicates the first The subcarrier at the ... The corresponding values ​​in the candidate channel frequency response sequence at each interpolation resolution scale. Indicates the first Frequency domain reconstruction operators corresponding to various interpolation resolution scales. By generating candidate channel frequency response sequences at multiple different interpolation resolution scales, fixed biases can be avoided in highly frequency-selective channels or scenarios with strong noise under a single interpolation scale.

[0056] Subsequently, the corresponding scale evaluation error is determined based on the pilot reconstruction error at the preset pilot position set for each interpolation resolution scale, and the adaptive combination weight corresponding to each interpolation resolution scale is calculated based on the scale evaluation error; among them, the smaller the scale evaluation error of the interpolation resolution scale, the larger the corresponding adaptive combination weight.

[0057] Specifically, the receiver can remap the candidate channel frequency response sequences generated at various interpolation resolution scales to pilot positions and compare the reconstructed response with the initial channel response to evaluate the reconstruction capability of the current interpolation resolution scale for the pilot channel state. To avoid the interpolation reconstruction simply replicating the initial channel response at the pilot position, resulting in a lack of discriminative power in the pilot reconstruction error, the pilot reconstruction error can be determined based on the smoothed reconstruction response, the leave-one-pilot verification result, or the residual evaluation result after adding a stabilization term. In other words, when calculating the reconstruction error at a certain pilot position, the response at that pilot position can be indirectly obtained from neighboring pilot positions or the smoothed reconstruction process, rather than directly using the initial channel response of the pilot position itself for complete replacement.

[0058] In some examples, the first The scale evaluation error corresponding to each interpolation resolution scale can be expressed as: Equation (3) In the formula, Indicates the first The scale evaluation error corresponding to each interpolation resolution scale. This indicates the number of pilots in the preset pilot location set. Indicates the first At the nth interpolation resolution scale, the th The reconstructed response value formed at each pilot position This represents a stable term used to avoid zero scale evaluation error. The smaller the scale evaluation error, the higher the consistency between the candidate channel frequency response sequence and the pilot observation results corresponding to the interpolation resolution scale, and therefore a larger adaptive combination weight can be assigned.

[0059] In some examples, adaptive combined weights can be determined based on the inverse normalized result of the scaling evaluation error: Equation (4) In the formula, Indicates the first Adaptive combined weights corresponding to each interpolation resolution scale. Indicates the first The scale evaluation error corresponding to each interpolation resolution scale. This represents the total number of interpolation resolution scales. Through this process, candidate channel frequency response sequences with smaller scale evaluation errors have a higher weight in the fusion result, while the impact of candidate channel frequency response sequences with larger scale evaluation errors on the final result is correspondingly reduced.

[0060] Then, the corresponding values ​​of the same subcarrier in the candidate channel frequency response sequences at various interpolation resolution scales are normalized and weighted according to the corresponding assigned adaptive combination weights to generate the final channel frequency response estimate for each subcarrier.

[0061] Specifically, for any subcarrier, the receiver can obtain the candidate channel frequency response values ​​for that subcarrier at multiple interpolation resolution scales and fuse them according to the adaptive combination weights corresponding to each interpolation resolution scale. Since the adaptive combination weights already reflect the reconstruction reliability of each interpolation resolution scale, the final channel frequency response estimate can take into account both the ability to characterize local channel variations and the ability to smooth noise. This fusion method does not select a fixed interpolation scale, but rather adaptively combines multiple candidate results based on the pilot reconstruction error, thereby improving the stability of the channel frequency response estimate.

[0062] In some examples, the channel frequency response estimate for a subcarrier can be expressed by the following formula: Equation (5) In the formula, Indicates the first The channel frequency response estimates corresponding to each subcarrier, and the adaptive combination weights satisfy... and Equation (5) indicates that, for the same subcarrier, the final channel frequency response estimate is obtained by fusing candidate channel frequency response values ​​under multiple interpolation resolution scales according to normalized weights. The larger the weight of the candidate channel frequency response value, the higher its contribution to the final result.

[0063] In this embodiment, by obtaining the initial channel response at the pilot location and forming a multi-scale candidate channel frequency response sequence using multiple interpolation filtering windows with different frequency domain span parameters, and then adaptively allocating combined weights based on the pilot reconstruction error, a fusion estimation of the channel frequency response of each subcarrier is achieved. This reduces the limitation of a single interpolation scale on the channel estimation results, enabling the channel frequency response estimate to maintain good stability and adaptability under conditions of both frequency-selective fading and pilot observation disturbances, thereby improving the accuracy of the wireless communication receiver's characterization of the subcarrier-level channel state.

[0064] Regarding the implementation details of calculating the residual power on the subcarrier in the method of this application embodiment, in some examples of this application embodiment, firstly, the channel estimation error variance is obtained based on the pilot reconstruction residual in the channel estimation process, and the dynamic error compensation coefficient corresponding to each subcarrier is calculated in combination with the reliability of the channel frequency response estimate; wherein, the smaller the channel estimation error, the higher the degree of preservation of the preliminary useful signal estimated power by the dynamic error compensation coefficient, and the larger the channel estimation error, the higher the degree of reduction of the preliminary useful signal estimated power by the dynamic error compensation coefficient.

[0065] Specifically, the channel frequency response estimate obtained from discrete pilot estimation may be affected by thermal noise, pilot observation errors, and frequency domain reconstruction errors. If this channel frequency response estimate is directly used to estimate the useful signal power, on subcarriers with large channel estimation errors, the estimated useful signal power may deviate from the actual reception state, thus affecting the accuracy of the residual power. Therefore, this embodiment uses the pilot reconstruction residual to characterize the uncertainty of the channel estimation and obtains the channel estimation error variance corresponding to each subcarrier accordingly. This variance can be obtained by smoothing or interpolating the reconstruction error at the pilot position to each subcarrier in the frequency domain, or by statistically analyzing the pilot residual energy at the receiver within a preset observation window.

[0066] In some examples, if the difference between the initial channel response and the reconstructed channel response at the pilot location is denoted as the pilot reconstruction residual, the channel estimation error variance can be determined based on the local statistics of the pilot reconstruction residual. For non-pilot subcarriers, the error variance at the pilot location can be extended to the corresponding subcarrier based on its frequency domain distance to adjacent pilot locations or on the reconstruction weights used in the channel frequency response estimation process. In this way, each subcarrier can obtain an error evaluation quantity that matches its channel estimation reliability.

[0067] Then, based on the channel frequency response estimate corresponding to each subcarrier and the transmit symbol characterization value available at the receiver, the preliminary estimated power of the useful signal corresponding to each subcarrier is calculated.

[0068] Specifically, the transmitted symbol representation value can be derived from different sources depending on the stage of the receiving process. At the pilot subcarrier location, the transmitted symbol representation value can be a pilot symbol known locally at the receiver; at the data subcarrier location, the transmitted symbol representation value can be a symbol estimate obtained from decision feedback, or it can be the average symbol power representation value corresponding to the current modulation scheme. Based on the channel frequency response estimate and the transmitted symbol representation value corresponding to the subcarrier, the receiver can estimate the received power contribution of the useful signal on that subcarrier after passing through the wireless channel, and use this as the preliminary estimated power of the useful signal.

[0069] Therefore, the preliminary useful signal power estimate reflects the expected power level of the useful signal on the corresponding subcarrier under the current channel frequency response estimation result. This power estimate may still be affected by the channel estimation error. Therefore, this embodiment does not directly use it as the final useful signal power to be eliminated, but further corrects it by combining the dynamic error compensation coefficient.

[0070] Subsequently, the dynamic error compensation coefficient is used to perform deviation scaling correction on the preliminary useful signal power estimate to obtain the useful signal power estimate for each subcarrier.

[0071] Specifically, the dynamic error compensation coefficient characterizes the reliability of the channel frequency response estimate on the current subcarrier relative to the channel estimation error. When the channel estimation error is small and the channel frequency response estimate is reliable, the dynamic error compensation coefficient tends to be higher, allowing the initial useful signal estimation power to be more fully preserved. When the channel estimation error is large, the dynamic error compensation coefficient decreases, appropriately reducing the initial useful signal estimation power. This deviation scaling correction can reduce the impact of channel estimation deviation on the useful signal power elimination process.

[0072] In this embodiment, instead of simply removing useful signal power at a fixed ratio, the removal intensity is adaptively adjusted based on the channel estimation reliability on each subcarrier. Therefore, at frequencies with more reliable channel estimation, the contribution of useful signal power can be more fully deducted; at frequencies with higher channel estimation uncertainty, excessive removal due to estimation bias can be reduced, resulting in a useful signal power estimate that better reflects the current reception status.

[0073] Then, the useful signal power estimate is removed from the corresponding total received power, and non-negative truncation processing is performed on the removal result to avoid abnormally negative values ​​of residual power. The remaining energy after truncation processing is determined as the residual power on each subcarrier.

[0074] Specifically, the total received power is determined by the received frequency domain coefficients on the corresponding subcarrier, representing the overall received energy on that subcarrier. After removing the estimated useful signal power, corrected by the dynamic error compensation coefficient, from the total received power, the remaining amount is used to characterize the residual energy other than the useful signal contribution. Due to factors such as received noise fluctuations, channel estimation errors, or symbol representation deviations, the removal result may have a value less than zero on individual subcarriers. To ensure that the residual power conforms to the non-negative property of power, this embodiment performs non-negative truncation processing on the removal result and uses the truncated result as the residual power on that subcarrier.

[0075] In some examples, the residual power on the subcarrier can be expressed by the following formula: Equation (6) Equation (7) In the formula, Indicates the first Residual power on each subcarrier; Indicates the first The received frequency domain coefficients on each subcarrier, their squared amplitude values Characterizes the total received power; Indicates the first Channel frequency response estimates for each subcarrier; Indicates the first The transmitted symbol representation value corresponding to each subcarrier; Characterize the power of the preliminary useful signal; Indicates the first The variance of the channel estimation error corresponding to each subcarrier; Indicates the first Dynamic error compensation coefficients corresponding to each subcarrier; Characterizes the estimated useful signal power after dynamic error compensation; This indicates non-negative truncation.

[0076] In the above mathematical model, the dynamic error compensation coefficient is jointly determined by the power of the channel frequency response estimate and the variance of the channel estimation error. When the channel frequency response estimate is dominant over the variance of the channel estimation error, the dynamic error compensation coefficient is higher, and the useful signal power estimate retains more of the preliminary estimation result; when the variance of the channel estimation error is relatively high, the dynamic error compensation coefficient is lower, and the useful signal power estimate shrinks accordingly. This reflects the process of removing the reliability-corrected useful signal power estimate from the total received power and performing non-negative truncation on the removal result.

[0077] Through the embodiments of this application, the useful signal power estimate can be adaptively corrected according to the reliability of channel estimation, and the non-negative property of residual power can be maintained after power removal. Therefore, the impact of channel estimation error on residual power calculation can be reduced, and the residual power on each subcarrier can more stably reflect the remaining energy beyond the contribution of the useful signal, thereby improving the reliability and physical consistency of the residual power quantization results.

[0078] Figure 2 A flowchart illustrating an example of determining the interference power estimate corresponding to each subcarrier in a method according to an embodiment of this application is shown.

[0079] like Figure 2 As shown, in step S210, for the target subcarrier, a frequency domain neighborhood window is defined based on a preset single-sided radius. When the frequency domain neighborhood window exceeds the effective frequency band boundary of the system, an edge mirror extension strategy is used for compensation, and the residual power outside the target subcarrier within the frequency domain neighborhood window is extracted to construct a reference power sample set.

[0080] Specifically, when determining interference in subcarrier analysis, examining only the residual power of the target subcarrier itself is susceptible to instantaneous noise spikes or single-point abnormal fluctuations. Therefore, this embodiment uses the target subcarrier as the center and selects neighboring subcarriers within a certain range in the frequency domain as reference objects to obtain the local residual power distribution around the target subcarrier. Here, a preset unilateral radius is used to limit the frequency domain range on both sides of the target subcarrier involved in the statistics, so that the reference power sample set can reflect the background power level and fluctuation state of the local frequency band where the target subcarrier is located.

[0081] When a target subcarrier approaches the boundary of the system's effective frequency band, the frequency domain neighborhood window defined by the radius on one side may exceed the effective subcarrier range. In this case, an edge mirroring extension strategy can be used to compensate for the missing reference position at the boundary, ensuring that the boundary subcarriers and non-boundary subcarriers maintain consistency in sample quantity and statistical scale. To avoid the abnormal residual power of the target subcarrier itself affecting the neighborhood background estimation, this embodiment excludes the target subcarrier from the frequency domain neighborhood window and extracts only the residual power corresponding to other reference subcarriers to construct a reference power sample set.

[0082] In some examples, the first The reference power sample set corresponding to each subcarrier can be represented as: Equation (8) In the formula, Indicates the first Reference power sample set corresponding to each subcarrier Indicates the first The residual power corresponding to each reference subcarrier Indicates the first The set of subcarrier indices corresponding to the frequency domain neighborhood window centered on the target subcarrier and after boundary compensation is used to characterize the local power state in the frequency domain region around the target subcarrier.

[0083] In step S220, the reference power sample set is sorted to extract the local sample median and the quantile values ​​of a preset order.

[0084] Specifically, the reference power sample set may simultaneously contain normal background noise samples, locally elevated power samples, and a small number of anomalous spike samples. After sorting the residual power in the reference power sample set, the receiver can extract the local sample median and quantile values ​​of a preset order from the sorted sample distribution. The local sample median is used to characterize the stable central level of the residual power in the neighborhood, and its influence is relatively small due to a small number of extreme samples; the quantile values ​​of the preset order are used to characterize the distribution boundary of higher power samples in the neighborhood, thus reflecting the characteristics of local high power fluctuations.

[0085] By simultaneously extracting the median and quantile values ​​of local samples, this embodiment can take into account both the stable level of the neighborhood background power and the high power fluctuation state. Compared with using only the average value as the statistical basis, this method can reduce the excessive influence of a few outliers on the background estimation results; compared with using only the median value, introducing quantile values ​​can retain the ability to perceive high power changes in the neighborhood, so that the subsequent dynamic detection threshold can better fit the local frequency domain environment of the target subcarrier.

[0086] In step S230, the estimated value of system background noise power is obtained, and the local sample median value adjusted by the first adjustment coefficient and the quantile value adjusted by the second adjustment coefficient are combined with the estimated value of system background noise power to form the dynamic detection threshold corresponding to the target subcarrier; wherein, the local sample median value is used to characterize the stable background power level in the neighborhood, and the quantile value is used to characterize the high power fluctuation characteristics in the neighborhood.

[0087] Specifically, the estimated system background noise power can be obtained by the receiver based on silent bands, guard subcarriers, historical noise observations, or other noise estimation methods, and is used to provide a reference for the basic noise level. Building upon this, this embodiment further introduces the local sample median and quantile values ​​from the reference power sample set, so that the detection threshold not only reflects the system background noise level but also the actual residual power distribution in the neighborhood of the target subcarrier. A first adjustment coefficient is used to adjust the influence of the local sample median on the dynamic detection threshold, and a second adjustment coefficient is used to adjust the influence of the quantile values ​​on the dynamic detection threshold.

[0088] For example, the dynamic detection threshold can be calculated using the following formula: Equation (9) In the formula, Indicates the first The dynamic detection threshold corresponding to each subcarrier. This represents the estimated background noise power. Indicates the radius of one side of the preset neighborhood window; Indicates the first Reference power sample set corresponding to each subcarrier; This represents the function for calculating the median; This indicates the preset quantile order, and ; express Quantile calculation function; This represents the first adjustment coefficient that controls the degree of dependence on the sample median. Indicates control pair The second adjustment coefficient for the degree of dependence of the order quantile values, and , .

[0089] In equation (9) above, the background noise power estimate provides the basic threshold level, the local sample median term is used to adjust the threshold according to the stable background power level of the neighborhood, and the quantile term is used to adjust the threshold according to the high power fluctuation state of the neighborhood. When the residual power of the frequency band where the target subcarrier is located increases as a whole, the dynamic detection threshold can be increased accordingly; when the neighborhood background is relatively stable, the dynamic detection threshold will not be directly increased due to the abnormal power of a single target subcarrier. Through this calculation method, the dynamic detection threshold can be adjusted with the change of local frequency domain statistical state, reducing the risk of misjudgment of the fixed threshold in non-uniform noise environment.

[0090] In step S240, when the residual power on the target subcarrier exceeds the dynamic detection threshold, it is determined that there is interference on the target subcarrier, and the energy of the excess part of the residual power relative to the system background noise power estimate after non-negative truncation is determined as the corresponding interference power estimate; when the residual power does not exceed the dynamic detection threshold, it is determined that there is no interference and the corresponding interference power estimate is set to zero.

[0091] Specifically, the receiver compares the residual power on the target subcarrier with the dynamic detection threshold corresponding to that subcarrier. If the residual power on the target subcarrier exceeds the dynamic detection threshold, it indicates that the remaining energy on that subcarrier is abnormally elevated relative to the local frequency domain background, and it can be determined that the target subcarrier is experiencing interference. In this case, this embodiment does not directly use all the residual power as the interference power, but instead uses the portion of the residual power that exceeds the estimated system background noise power as the basis for estimating the interference energy, thereby reducing the impact of background noise on the estimated interference power.

[0092] For example, in determining the first When interference exists on each subcarrier, the estimated interference power of the subcarrier can be expressed by the following formula: Equation (10) in, Indicates the first Interference power estimates for each subcarrier Indicates the first Residual power on each subcarrier This represents the estimated background noise power. This indicates non-negative truncation.

[0093] When residual power Not greater than the dynamic detection threshold At that time, the judgment of the first Each subcarrier is free from interference, and the corresponding interference power estimate is given. Set to zero. Thus, this embodiment can obtain the interference determination result and the interference power estimate in the same processing process, so that the abnormal energy state of the target subcarrier can be represented in a more explicit power form.

[0094] Through the embodiments of this application, the receiving end can determine a dynamic detection threshold for the target subcarrier based on the statistical characteristics of the residual power in the neighborhood of the target subcarrier, and determine and quantify the residual power on the target subcarrier based on the dynamic detection threshold. This reduces the sensitivity of a fixed threshold to non-uniform background noise, makes the interference power estimate more accurately reflect the abnormal energy level on the target subcarrier, and improves the stability and usability of the subcarrier-level interference detection results.

[0095] Figure 3 A flowchart illustrating an example of determining a frequency-domain smoothing weight mask composed of mask weights corresponding to each subcarrier in a method according to an embodiment of this application is shown.

[0096] like Figure 3 As shown, in step S310, a nonlinear continuous decay model configured with a preset weight protection lower limit is obtained. The nonlinear continuous decay model has a mapping characteristic based on an S-shaped function and is controlled by an interference power threshold used to define the suppression start boundary and a slope coefficient used to control the steepness of the decay curve.

[0097] Specifically, in frequency domain interference suppression, directly zeroing out subcarriers identified as being interfered with can reduce the interference energy at the corresponding frequency, but it may also suppress useful signal components that still exist in the subcarrier, causing frequency domain discontinuities between adjacent frequency points. Therefore, this embodiment employs a nonlinear continuous attenuation model, generating continuously varying initial mask weights based on the estimated interference power for each subcarrier, transforming the interference suppression process from hard-decision zeroing to flexible attenuation that varies according to interference intensity.

[0098] Here, the nonlinear continuous attenuation model possesses mapping characteristics based on a sigmoid function. The interference power threshold characterizes the reference boundary at which enhancement and suppression begin, while the slope coefficient adjusts the transition speed of the mask weights as the interference power changes. A preset weight protection lower limit restricts the minimum value of the mask weights, ensuring that even under strong interference conditions, the signal information of the corresponding subcarrier is not completely zeroed. This reduces the risk of amplitude and phase information loss due to excessive weight attenuation, maintaining good continuity and processability in the frequency domain weighting results.

[0099] In step S320, the interference power estimate corresponding to each subcarrier is input into the nonlinear continuous attenuation model for nonlinear mapping transformation, the initial mask weight corresponding to each subcarrier is calculated, and the signal information of the corresponding subcarrier is completely set to zero by using a preset weight protection lower limit; wherein, when the interference power estimate is lower than the interference power threshold, the initial mask weight approaches the upper limit of the preset weight interval; when the interference power estimate is higher than the interference power threshold, the initial mask weight converges to the preset weight protection lower limit as the interference power increases.

[0100] For example, the initial mask weights can be expressed as follows: Equation (11) In the formula, Indicates the first The initial mask weights corresponding to each subcarrier. Indicates the first Interference power estimates for each subcarrier This indicates the preset interference power threshold. The slope coefficient, which controls the steepness of the decay curve, has the dimension of the reciprocal of power, and , This indicates a preset lower limit for weight protection, and .

[0101] In equation (11) above, when the estimated interference power is below the interference power threshold, the suppression effect corresponding to the exponential term is weak, and the initial mask weight approaches the upper limit of the preset weight interval, allowing more of the received frequency domain coefficients of the corresponding subcarrier to be preserved. When the estimated interference power is above the interference power threshold, the initial mask weight gradually decreases with the increase of interference power and converges towards the lower limit of the preset weight protection. Thus, the initial mask weight can continuously change between the lower limit of the preset weight protection and the upper limit of the preset weight interval, achieving graded attenuation for different interference intensities.

[0102] This establishes a continuous and monotonic mapping between the estimated interference power and the initial mask weights. Compared to hard-zeroing, the continuous attenuation model reduces abrupt changes in frequency domain processing; compared to a fixed attenuation ratio, this model adjusts the attenuation level according to the magnitude of the interference power, thereby improving the adaptability of frequency domain weighted processing.

[0103] In step S330, a frequency domain smoothing window sequence that satisfies the normalization condition and is centrally symmetric is introduced. Combined with the boundary symmetric extension strategy, a one-dimensional convolutional filtering operation is performed on the sequence composed of the initial mask weights corresponding to each subcarrier to eliminate abrupt changes in frequency domain weights. The result after convolutional filtering is determined as the mask weights that satisfy the weight continuity constraint, and the mask weights constitute the frequency domain smoothing weight mask.

[0104] For example, mask weights can be expressed as follows: Equation (12) In the formula, Indicates the first The final mask weights corresponding to each subcarrier after frequency domain smoothing Indicates the index in the frequency domain smoothing window. The window coefficient, and satisfying as well as , Indicates the first The initial mask weights corresponding to each subcarrier. The smoothing radius is set to a positive integer. A centrally symmetric frequency domain smoothing window can satisfy... This allows the neighborhood weights on both sides of the target subcarrier to participate in the smoothing process symmetrically.

[0105] Specifically, the initial mask weights for each subcarrier are calculated independently based on the interference power estimates for that subcarrier. When the interference power estimates on adjacent subcarriers differ significantly, the initial mask weight sequence may exhibit local jumps. If this initial mask weight sequence is directly used to process the received frequency domain coefficients, it may cause discontinuities in the frequency domain weighting results and adversely affect the time domain waveform recovery. Therefore, this embodiment employs a centrally symmetric frequency domain smoothing window to perform one-dimensional convolution filtering on the initial mask weight sequence, making the mask weight changes between adjacent subcarriers smoother.

[0106] When performing a one-dimensional convolutional filtering operation, if the sliding window exceeds the beginning or end of the effective subcarrier frequency band of the system, a boundary-symmetric extension strategy can be used to compensate for the initial mask weight sequence. For example, this boundary-symmetric extension operation can be represented as follows:

[0107] Equation (13) In the formula, This represents the total number of effective subcarriers in the system. This represents the index offset at which boundary extension occurs. This represents the smoothing radius setting. Through this process, subcarriers located at the frequency band boundary can also obtain smoothing at the same scale as internal subcarriers, reducing weight abrupt changes at the boundary caused by insufficient samples.

[0108] Since the window coefficients of the frequency domain smoothing window satisfy the non-negativity and normalization conditions, the smoothed mask weights can be regarded as a weighted combination of the initial mask weights in the neighborhood. When all the initial mask weights are within a preset weight interval, the smoothed mask weights can still remain within that preset weight interval. Thus, frequency domain smoothing can reduce local weight spikes without destroying the original weight value range, making the obtained frequency domain smoothed weighted mask both continuous and constrained stable.

[0109] Through the embodiments of this application, the receiver can generate continuously varying initial mask weights based on the interference power estimate, and smooth the initial mask weights using a centrally symmetric frequency domain smoothing window. This reduces the frequency domain discontinuities caused by hard zeroing and abrupt changes in local weights, enabling the frequency-domain smoothed weight mask to maintain suppression capabilities while possessing smoother frequency domain variation characteristics, thereby improving the stability and waveform consistency of the frequency-domain weighted attenuation results.

[0110] Regarding the implementation details of obtaining the received frequency domain coefficients on the subcarrier through frequency domain transformation operation in the method of this application embodiment, in some examples of this application embodiment, multiple time truncation windows with different length parameters, including a preset main time window, are used to perform frequency domain transformation operation on the time domain baseband received signal respectively, generating multiple sets of initial frequency domain coefficient sequences at different frequency resolutions.

[0111] Specifically, time truncation windows of different lengths correspond to different time observation scales and frequency resolutions. Longer time truncation windows provide higher frequency resolution, which is beneficial for distinguishing interference components with similar frequency positions; shorter time truncation windows provide better temporal locality, which is beneficial for reflecting the energy changes of transient or sudden interference within a short period of time. By performing frequency domain transformation on the same time-domain baseband received signal based on multiple time truncation windows at the receiver, multiple sets of initial frequency domain coefficient sequences with different frequency resolutions can be obtained.

[0112] In some examples, the first Each time cutoff window is represented as The initial frequency domain coefficient sequence obtained after processing by this time truncation window is represented as follows: ,in, , This indicates the total number of time-truncation windows. Indicates the first Each time truncation window corresponds to a frequency index under the frequency grid. Through the above multi-window frequency domain transformation, the same received signal can be observed at multiple frequency resolutions, thereby enabling the interference energy distribution at different time scales to be characterized separately.

[0113] Then, the initial frequency domain coefficient sequence corresponding to the preset main time window is determined as the received frequency domain coefficient sequence, and the received frequency domain coefficient sequence is used as the alignment reference sequence corresponding to the preset main time window; frequency domain raster resampling and interpolation alignment operations are performed on the initial frequency domain coefficient sequences generated by other time truncation windows to generate multiple sets of aligned frequency domain coefficient sequences that match the frequency raster of the alignment reference sequence; wherein, the alignment reference sequence and the multiple sets of aligned frequency domain coefficient sequences together serve as the aligned frequency domain coefficient sequences corresponding to each time truncation window.

[0114] Specifically, since time truncation windows with different length parameters correspond to different frequency resolutions, the frequency grids containing their initial frequency domain coefficient sequences may not be consistent. To enable the frequency domain results from multiple time truncation windows to be compared and fused at the same subcarrier position, this embodiment selects the initial frequency domain coefficient sequence corresponding to a preset main time window as the alignment reference sequence. The initial frequency domain coefficient sequences corresponding to other time truncation windows can be mapped onto the frequency grid of the alignment reference sequence through frequency domain grid resampling, interpolation mapping, or projection onto nearby frequency points.

[0115] After completing the alignment, the first... The first time truncation window The aligned frequency domain coefficients corresponding to each subcarrier are denoted as follows: , among which, when the first When the time truncation window is the preset main time window That is, the received frequency domain coefficient sequence is at the 1st... The corresponding values ​​at each subcarrier are obtained. Through the above processing, the frequency domain observation results under each time truncation window are unified to the same subcarrier grid, making the multi-scale detection results at the same subcarrier position comparable.

[0116] Furthermore, regarding the implementation details of determining the interference power estimate corresponding to the subcarrier in the method of the embodiments of this application, in some examples, the single-scale received power under each time truncation window is determined based on the aligned frequency domain coefficient sequence corresponding to each time truncation window, and the useful signal power estimate of the corresponding scale is independently removed from the single-scale received power to obtain the single-scale residual power under each time truncation window.

[0117] Specifically, for any time truncation window, the receiver can determine the single-scale received power on each subcarrier based on the aligned frequency domain coefficients corresponding to that window. Since different time truncation windows have different characterization capabilities for the temporal locality and frequency resolution of interference, the single-scale received power may differ between time truncation windows. To make the detection target at each scale more focused on anomalous energy components, this embodiment independently removes the estimated useful signal power at the corresponding scale from the single-scale received power to obtain the single-scale residual power under that time truncation window.

[0118] In some examples, the first The first time truncation window The single-scale residual power corresponding to each subcarrier can be expressed as: Equation (14) In the formula, Indicates the first The first time truncation window Single-scale residual power corresponding to each subcarrier Indicates the first The first time truncation window Aligned frequency domain coefficients corresponding to each subcarrier Indicates the first The first time truncation window The estimated useful signal power for each subcarrier. This indicates non-negative truncation. This formula is only used to illustrate one possible way to calculate the single-scale residual power; the specific useful signal power estimate can be determined in a manner consistent with other embodiments of this application.

[0119] Subsequently, the single-scale residual power is compared with the single-scale dynamic detection threshold that matches the current frequency resolution to generate a single-scale interference detection flag. When the single-scale residual power exceeds the corresponding single-scale dynamic detection threshold, the single-scale interference detection flag indicates the presence of interference, and the portion exceeding the estimated background noise power at the corresponding scale is identified as the single-scale interference power. Simultaneously, the corresponding single-scale confidence level is calculated based on the magnitude of the single-scale residual power overflowing the single-scale dynamic detection threshold. The single-scale confidence level corresponding to the time truncation window where no interference is detected is set to zero.

[0120] Specifically, for each time truncation window, the receiver can determine a matching single-scale dynamic detection threshold based on the frequency resolution corresponding to that time truncation window. If the single-scale residual power of the target subcarrier exceeds the corresponding single-scale dynamic detection threshold within a certain time truncation window, the single-scale interference detection flag for that time truncation window indicates the presence of interference; if it does not exceed the corresponding single-scale dynamic detection threshold, the single-scale interference detection flag indicates that no interference has been detected. This allows for obtaining detection results for the same subcarrier across multiple time truncation windows.

[0121] For example, in the first The first time truncation window When all subcarriers are determined to be subject to interference, the single-scale interference power under the time truncation window can be expressed by the following formula: Equation (15) In the formula, Indicates the first Under the time truncation window, targeting the first Single-scale interference power extracted from each subcarrier Indicates the first The first time truncation window Single-scale residual power corresponding to each subcarrier Indicates the relationship with the first Background noise power estimates corresponding to each time truncation window This indicates non-negative truncation; it should be noted that when The corresponding single-scale dynamic detection threshold was not exceeded. At that time, the corresponding single-scale interference detection flag indicates that no interference was detected, and the corresponding [indicator / signature] can be [activated / displayed]. Set to zero or not participate in subsequent fusion.

[0122] To characterize the reliability of detection results under different time cutoff windows, this embodiment further calculates the single-scale confidence level based on the overflow amplitude of the single-scale residual power exceeding the single-scale dynamic detection threshold. For example, the single-scale confidence level can be expressed as:

[0123] Equation (16) In the formula, Indicates the first The subcarrier at the ... Single-scale confidence level under a time truncation window Represents single-scale residual power. This represents the single-scale dynamic detection threshold. This represents a positively stable term used to avoid a denominator of zero. The single-scale confidence increases with the overflow of the single-scale residual power relative to the single-scale dynamic detection threshold; when no interference is detected, the corresponding single-scale confidence is zero.

[0124] Furthermore, a scale fusion decision based on logical OR is performed on each single-scale interference detection identifier at the same subcarrier location. When it is confirmed that the current subcarrier is interfered with, the single-scale interference power under each time truncation window is normalized and weighted according to the corresponding single-scale confidence level to obtain the final output interference power estimate. When it is not confirmed that the current subcarrier is interfered with, the final output interference power estimate is set to zero.

[0125] Specifically, for the same subcarrier location, if at least one time-truncation window indicates the presence of interference using a single-scale interference detection flag, then the subcarrier can be confirmed as an interfered subcarrier in the multi-scale detection results. Subsequently, the receiver normalizes and weights the single-scale interference power obtained from each time-truncation window according to the corresponding single-scale confidence level. Time-truncation windows with higher confidence levels contribute more to the final interference power estimate, while time-truncation windows with zero confidence levels do not participate in effective fusion.

[0126] In some examples, the estimated interference power for a subcarrier can be expressed by the following formula: Equation (17) In the formula, Indicates the first The final estimated interference power value for each subcarrier This indicates the total number of time-truncation windows. Indicates the first The subcarrier at the ... Single-scale confidence level under a time truncation window Indicates the first Under the time truncation window, targeting the first The formula is for the single-scale interference power extracted from each subcarrier. In situations where it is not confirmed that the current subcarrier is being interfered with, or At that time, set directly .

[0127] In Equation (17), the final interference power estimate is determined by the single-scale interference power under multiple time-truncation windows. Since the fusion weight is given by the single-scale confidence, the detection scale with a larger overflow amplitude can obtain a higher contribution ratio, while the scale that did not detect interference will not affect the final result because its confidence is zero. In this way, the frequency domain observation results of different time-truncation windows can be integrated at the same subcarrier position, making the interference power estimation more adaptable to the scenario where transient interference and narrowband interference coexist.

[0128] Through the embodiments of this application, the receiver can obtain frequency domain observation results with different frequency resolutions under multiple time truncation windows, and unify them onto the same subcarrier grid for single-scale detection and fusion. This allows for a balance between the frequency resolution capability of long-term windows and the temporal local sensing capability of short-term windows, resulting in a more stable reflection of the abnormal energy state at the same subcarrier location by the final interference power estimate, thus improving the reliability and adaptability of multi-scale frequency domain detection results.

[0129] In some examples of embodiments of this application, after reconstructing the time-domain waveform after interference cancellation, a closed-loop parameter adaptive adjustment step based on demodulation feedback can also be performed.

[0130] First, the distribution of subcarrier interference within the preset time sliding window is statistically analyzed and smoothed to obtain the smoothed interference detection rate, and the real-time bit error rate is obtained by combining the demodulation and decoding results.

[0131] Specifically, the receiver can count the number of subcarriers identified as having interference within a preset time sliding window, and use the ratio between this number and the total number of subcarriers participating in the detection as the interference detection rate for the current adjustment period. Since interference in the wireless environment has short-term fluctuation characteristics, directly using the detection rate of a single processing period for parameter adjustment is easily affected by instantaneous abnormal detection results. Therefore, this embodiment can perform smoothing filtering on the interference detection rate over multiple processing periods to obtain a smoothed interference detection rate. For example, the smoothed interference detection rate can be expressed by the following formula:

[0132] Equation (18) In the formula, This indicates the smoothing interference detection rate corresponding to the current adjustment period. This represents the smoothing interference detection rate corresponding to the previous adjustment period. This represents the interference detection rate obtained from statistics during the current adjustment period. Represents the smoothing filter coefficients, and This smoothing process reduces the impact of single-detection fluctuations on parameter adjustments, allowing the interference detection rate to more stably represent the current frequency domain interference distribution.

[0133] Simultaneously, the receiving end can demodulate and decode the time-domain waveform after interference cancellation, and obtain the real-time bit error rate (BER) index based on the decoding verification results, pilot-assisted decision results, or known reference information. The BER index characterizes the quality status of the current received signal at the demodulation and decoding level; it can be the bit error rate, symbol decision error rate, or other indicators that reflect demodulation reliability. The smoothing interference detection rate reflects the abnormal energy distribution on the frequency domain detection side, while the BER index reflects the communication quality on the receiving demodulation side; both serve as feedback for adaptive parameter adjustment.

[0134] Then, when the smoothed interference detection rate deviates from the preset detection rate control target, and the real-time bit error rate index does not exceed the preset bit error rate tolerance upper limit, based on the signed detection difference between the smoothed interference detection rate and the detection rate control target, the first adjustment coefficient and the second adjustment coefficient are adaptively updated according to the first preset step size set under the constraint of preset coefficient boundaries, so as to dynamically balance the threshold sensitivity of the system to abnormal power fluctuations; wherein, when the smoothed interference detection rate is higher than the detection rate control target, the dynamic detection threshold in the subsequent processing cycle is increased; when the smoothed interference detection rate is lower than the detection rate control target, the dynamic detection threshold in the subsequent processing cycle is decreased.

[0135] Specifically, if the real-time bit error rate (BER) does not exceed the preset BER tolerance limit, it indicates that the current communication quality is within an acceptable range. In this case, the sensitivity of the dynamic detection threshold can be adjusted primarily based on the interference detection rate. In some examples, the difference between the smoothed interference detection rate and the preset detection rate control target can be defined as the signed detection difference.

[0136] Equation (19) In the formula, This indicates the signed detection difference. This indicates the smoothing interference detection rate corresponding to the current adjustment period. This indicates the preset detection rate control target. When When the current detection rate is higher than the target level, the dynamic detection threshold can be increased by increasing the first and second adjustment coefficients, making the system's response to local power fluctuations more conservative; when When the current detection rate is lower than the target level, the first and second adjustment coefficients can be reduced accordingly to keep the system highly sensitive to abnormal power changes.

[0137] In some examples, the threshold adaptive operation is used to adaptively update the first adjustment coefficient under preset coefficient boundary constraints. Second adjustment coefficient Based on the sensitivity of the dynamic equilibrium system to abnormal power fluctuations, its update calculation model is as follows: Equation (20) Equation (21) In the formula, and These represent the first adjustment coefficients corresponding to the current adjustment period and the next adjustment period, respectively. and These represent the second adjustment coefficients corresponding to the current adjustment period and the next adjustment period, respectively. Indicates the difference in signed detection values; This indicates the preset first adjustment coefficient update step size; This indicates the preset update step size of the second adjustment coefficient; This indicates the preset upper limit of the first adjustment coefficient; This indicates the preset upper limit of the second adjustment coefficient; and These represent the functions for calculating the maximum and minimum values, respectively.

[0138] In the updated model described above, the outer boundary constraint is used to keep the first and second adjustment coefficients within a non-negative range and a preset upper limit, preventing excessive changes in parameters due to continuous feedback accumulation. Since the first and second adjustment coefficients act on local statistical terms in the dynamic detection threshold, the dynamic detection threshold increases when their values ​​increase and decreases when their values ​​decrease. Therefore, the threshold parameter can be bounded according to the deviation direction between the smoothing interference detection rate and the detection rate control target.

[0139] Furthermore, when the real-time bit error rate (BER) deviates from the preset target BER, and the smoothed interference detection rate does not exceed the preset detection rate tolerance limit, based on the signed bit error difference between the real-time BER and the target BER, the interference power threshold and slope coefficient in the nonlinear continuous attenuation model are dynamically updated according to the second preset step size set under preset parameter boundary constraints. This adaptively adjusts the suppression depth and activation boundary of the frequency domain weights in subsequent processing cycles. Specifically, when the real-time BER is higher than the target BER, the interference suppression intensity in subsequent processing cycles is enhanced; when the real-time BER is lower than the target BER, the interference suppression intensity in subsequent processing cycles is weakened.

[0140] Specifically, when the real-time bit error rate (BER) deviates from the target BER, but the smoothed interference detection rate does not exceed the preset detection rate tolerance limit, it can be considered that the current detection count is not significantly high. In this case, the suppression depth and suppression initiation boundary of the frequency domain weights can be changed by adjusting the parameters of the nonlinear continuous decay model. In some examples, the difference between the real-time BER and the target BER can be defined as the signed bit error difference.

[0141] Equation (22) In the formula, This represents the difference in signed bit errors. This indicates the real-time bit error rate metric corresponding to the current adjustment period. This represents the preset target bit error rate. When... When the current bit error rate is higher than the target bit error rate, the continuous attenuation model can be made to enter a stronger suppression state earlier by lowering the interference power threshold and increasing the slope coefficient; when When the current bit error rate is lower than the target bit error rate, the interference power threshold and slope coefficient can be adjusted in reverse within the preset boundary range to reduce unnecessary frequency domain attenuation.

[0142] In some examples, mask suppression adaptive operations utilize signed error differentials. The parameter update of the continuous attenuation model is subject to boundary constraints to dynamically adjust the interference suppression intensity in subsequent processing cycles. The update calculation model is as follows: Equation (23) Equation (24) In the formula, and These represent the interference power thresholds corresponding to the current adjustment period and the next adjustment period, respectively. and These represent the slope coefficients corresponding to the current adjustment period and the next adjustment period, respectively. This represents the difference in signed bit errors; This indicates the preset threshold update step size, which is expressed in units of power. This represents the preset slope update step size, which is the reciprocal of power. and This indicates the preset lower and upper thresholds; and This indicates the preset lower and upper bounds of the slope; the updated interference power threshold and slope coefficients are used in the continuous attenuation model in the next processing cycle.

[0143] In the updated model described above, the interference power threshold influences the power position at which the continuous attenuation model begins to enhance suppression, while the slope coefficient affects the rate at which the mask weights decrease with changing interference power. When the bit error rate (BER) is higher than the target BER, the signed bit error difference is positive, the interference power threshold decreases under boundary constraints, and the slope coefficient increases under boundary constraints, making it easier for the mask weights corresponding to the interference power estimate to change towards stronger suppression. When the BER is lower than the target BER, the signed bit error difference is negative, and the interference power threshold and slope coefficient can be adjusted in opposite directions, making the attenuation of the frequency domain weights more moderate. Preset parameter boundary constraints are used to limit the parameter adjustment range, avoiding excessively strong or weak suppression that could lead to unstable reception.

[0144] Through the embodiments of this application, the receiver can adjust the statistical adjustment parameters in the dynamic detection threshold according to the smoothed interference detection rate, and adjust the threshold and slope parameters in the continuous attenuation model according to the real-time bit error rate index. This allows for dual-loop decoupling and bounded adjustment of detection sensitivity and frequency domain suppression strength as the receiver's physical state and link decoding quality change, completely breaking the mismatch bottleneck of fixed parameters in dynamic broadband electromagnetic environments, and greatly improving the steady-state convergence and engineering adaptability of the adaptive interference cancellation system.

[0145] Figure 4 A schematic diagram illustrating the system operation mechanism of an example of a frequency-domain adaptive interference cancellation method based on interference power detection according to an embodiment of this application is shown.

[0146] like Figure 4 As shown, the system operation mechanism in this embodiment takes the air interface received baseband signal as the forward input starting point. After the receiver acquires the time-domain baseband received signal in a complex interference environment, it first converts the time-domain signal into a frequency-domain subcarrier representation through FFT frequency domain transformation to obtain the received frequency-domain coefficients on each subcarrier. Subsequently, the receiver performs multi-resolution channel estimation in conjunction with pilot information to obtain the channel frequency response estimation results corresponding to each subcarrier, and removes the useful signal power contribution from the total received power based on the channel frequency response estimation results, thereby obtaining the residual power on each subcarrier.

[0147] After obtaining the residual power at the subcarrier level, the system enters the interference power detection stage. This stage determines a dynamic detection threshold based on the frequency domain statistical characteristics of the residual power and determines the estimated interference power value for each subcarrier based on the detection result fusion process. Thus, the receiver can obtain the location of the interfered subcarrier and its corresponding interference power information in the frequency domain, providing a power quantization basis for subsequent frequency domain adaptive cancellation.

[0148] Furthermore, the system generates a continuous weighted mask based on the interference power estimate for each subcarrier, and forms a frequency-domain smoothed weighted mask through frequency-domain smoothing. The frequency-domain adaptive cancellation module uses this frequency-domain smoothed weighted mask to weightedly attenuate the received frequency-domain coefficients on each subcarrier to reduce the impact of abnormal interference energy on the received signal. The frequency-domain coefficients after weighted attenuation are further subjected to channel equalization and IFFT (Inverse Fast Fourier Transform) reconstruction to obtain the time-domain signal after interference cancellation. At the same time, the system can also output interference power spectrum, bit error rate index, and signal fidelity evaluation results to characterize the interference status and signal quality of the current receiving link.

[0149] also, Figure 4 The feedback adjustment link in this embodiment is also illustrated. The system can adaptively adjust parameters such as the dynamic detection threshold, interference power threshold, and slope coefficient based on demodulation and decoding results, interference detection distribution, and signal quality evaluation results. Specifically, dynamic threshold adjustment improves the adaptability of the interference detection stage to local power fluctuations; threshold and slope updates adjust the suppression initiation boundary and attenuation variation during the continuous weight mask generation process; and closed-loop parameter control coordinates the above parameter update process, enabling the system to adjust the frequency domain detection and cancellation strength according to changes in the received state.

[0150] pass Figure 4 The system operation mechanism shown in this embodiment integrates frequency domain transformation, channel estimation, residual power calculation, interference power detection, continuous weight mask generation, frequency domain adaptive cancellation, and channel equalization reconstruction into a complete receiver interference suppression process. This process can achieve adaptive interference cancellation based on the frequency domain power state of the received signal and link feedback information without relying on an independent external interference reference channel, thereby improving the stability and fidelity of the received signal in complex wireless environments.

[0151] To verify the engineering effectiveness and robustness of the frequency-domain adaptive interference cancellation method based on interference power detection proposed in this invention, this embodiment constructs an OFDM link-level simulation platform containing a complete receive baseband processing link based on the 3GPP 5G NR physical layer standard architecture. Regarding system parameters and channel environment configuration, the total number of effective subcarriers in the system is set. The subcarrier spacing is 1024. The frequency response is 15 kHz, and it supports dynamic switching between 16-QAM (Quadrature Amplitude Modulation) and 64-QAM modulation orders. To realistically reproduce the complex dynamic propagation patterns of wireless communication, the simulation adopts an Extended Typical Urban (ETU) multipath time-varying fading channel model, sets the maximum Doppler frequency shift to 70 Hz to simulate low-to-medium speed mobile terminal scenarios, and globally superimposes Additive White Gaussian Noise (AWGN) to construct the basic physical noise floor.

[0152] To address the interference evolution characteristics under complex electromagnetic environments, this embodiment custom-injects two high-dynamic interference models into the simulation platform. The first is cross-grid burst narrowband interference, whose generation center frequency is a non-integer multiple of the receiver's Fast Fourier Transform frequency domain grid. The first type is frequency offset, and the center frequency dynamically changes over time. This is specifically designed to verify the accuracy of the multi-scale grid resampling alignment mechanism and residual power extraction in this scheme. The second type is time-varying broadband frequency sweep interference, which is a sudden broadband interference in which the injected power spectral density evolves over time among multiple subcarriers. This is mainly used to examine the sensitivity of the system in tracking non-stationary background noise with dynamic detection threshold, as well as the flexible suppression capability of continuous S-shaped weighted masks.

[0153] In addition, to provide an objective benchmark for evaluating the technical performance, two comparative baseline schemes were set up for simulation testing. Baseline 1 adopts the traditional frequency domain hard-decision zeroing method, which directly forces the amplitude of the corresponding frequency point to zero after detecting interference, in order to highlight the fidelity advantage of the continuous weight mask of this scheme in preventing severe distortion of useful signals and maintaining frequency domain continuity. Baseline 2 adopts an adaptive cancellation method based on an ideal external reference signal, which serves as the ideal performance upper limit, to verify that under the strict constraints of no external reference channel, this scheme can approach or even surpass the interference cancellation limit of the traditional dual-channel algorithm in drastically time-varying scenarios by relying solely on endogenous residual power detection and closed-loop adaptive adjustment.

[0154] Figure 5 A schematic diagram showing the experimental simulation effect comparison between the frequency domain adaptive interference cancellation method based on interference power detection according to an embodiment of this application and a baseline is presented. The simulation result diagram is divided into a joint performance analysis section of two core physical dimensions.

[0155] like Figure 5Part (a) shows the receiver operating characteristic (ROC) curve of the interference detection performance at the receiver and the dynamic threshold tracking principle. It can be seen that under the set false alarm rate constraints (e.g., Under these conditions, compared to baseline 1 which uses traditional frequency domain fixed hard threshold detection, the detection probability of the method in this application embodiment is significantly improved (detection probability difference). This performance gain is intuitively reflected in... Figure 5 In the embedded graph features of (a): the embodiments of this application introduce a dynamic detection threshold (such as a quantile dynamic threshold). It can closely track the time-varying envelope of broadband non-stationary interference power, thereby effectively avoiding serious missed detections or excessive false alarms caused by fixed thresholds.

[0156] The significant advantages of front-end interference detection are further demonstrated in the demodulation performance at the receiver back end. For example... Figure 5 As shown in section (b), in the evolution comparison of bit error rate (BER) and signal-to-interference ratio (SIR), when the system is in a strong interference region (e.g., SIR < 5 dB), the method of this application embodiment achieves a reduction in BER of approximately 1.5 to 2 orders of magnitude compared to baseline 1. Figure 5 (b) The embedded complex plane constellation diagram (taking 16-QAM modulation under SIR=3 dB as an example) further reveals the underlying physical reason for this performance difference: When dealing with cross-grid frequency interference, the multi-point hard nulling operation adopted by baseline 1 leads to severe spectral leakage and useful signal energy collapse, which manifests as severe phase divergence and amplitude distortion at the constellation diagram decision point; In contrast, the embodiments of this application benefit from the high-precision interference power quantization and the flexible weighting mechanism of the continuous attenuation mask, and the interfered complex symbols are more concentrated near the ideal decision point, which greatly improves the signal fidelity after interference cancellation.

[0157] Figure 6 This diagram illustrates the comparative simulation results of different methods in the frequency domain mask response and error vector magnitude under sudden broadband interference scenarios. The simulation results are plotted using the horizontal axis of the three layers sharing the same subcarrier index. The aligned subplots reveal the underlying impact of fine-grained frequency domain processing on signal fidelity.

[0158] like Figure 6 In the burst broadband interference power spectrum of part (a), the system in local frequency bands (e.g., subcarrier index) Sudden broadband interference energy was injected within the interval.

[0159] like Figure 6Part (b) of the document compares the frequency domain weight response distribution of the dynamically smoothed mask generated by the embodiments of this application with that of the traditional hard truncation method. Unlike the step-like abrupt change (i.e., a hard bottoming out that is either 1 or 0) generated by the traditional hard truncation method at the interference boundary, the dynamically smoothed mask generated by the embodiments of this application... Not only does it maintain a preset weight protection lower limit greater than zero in the strong interference region (as shown by the water level of approximately 0.2 in the figure), preventing the signal information of the interfered subcarrier from being completely erased, but it also exhibits a continuous and gentle transition characteristic in the interference edge region. This smooth transition is due to the synergistic effect of the nonlinear continuous attenuation model (S-shaped function) and the one-dimensional convolutional filtering operation in the frequency domain. Its weight change trajectory can reduce the abrupt changes in frequency domain weights at the interference edge, making the weight change smoother.

[0160] The signal fidelity gain brought about by this frequency domain continuity mechanism is Figure 6 The error vector magnitude (EVM) distribution of the compensated subcarrier in part (c) of this paper provides objective evidence for this. Traditional hard filtering truncation, due to the rectangular abrupt change in the frequency domain, induces severe edge oscillations and secondary distortions after inverse frequency domain transformation, resulting in a sharp spike in the EVM of subcarriers near the interference boundary (as high as 10% to 12%), exceeding the system reference threshold. Conversely, after smooth mask compensation according to the embodiments of this application, the local EVM of the interfered frequency band is uniformly and stably suppressed below 8% of the reference threshold, avoiding spectrum leakage caused by hard truncation and meeting the stringent requirements of high-quality communication demodulation.

[0161] Figure 7 The diagram shows a simulation result of an example of boundary quantification evaluation of engineering implementation under extreme constraints according to the method of the present application. The simulation diagram visually quantifies the performance boundary of the method of the present invention under computational overhead and physical limits in the form of a bipartite graph.

[0162] like Figure 7 Part (a) shows the Pareto front trend of computational overhead versus frequency domain smoothing window length, with the horizontal axis representing the frequency domain smoothing window length (corresponding to the aforementioned length of...). The vertical axis represents the increase in single-frame processing latency (the smoothing window). It can be seen that the algorithm's latency overhead is approximately linearly positively correlated with the smoothing window length. As the smoothing window length increases to the right, the surge in underlying floating-point multiply-accumulate operations causes the processing latency to gradually approach and eventually exceed the baseline red line (10%) of the Ultra-Reliable and Low-Latency Communications (URLLC) protocol shown by the dashed line in the figure; for example, when the smoothing window length reaches 13, the increase in single-frame processing latency reaches approximately 12%. This indicates that in micro-communication nodes facing extremely limited computing power or highly latency-sensitive environments, the system must move to the left according to the latency red line, achieving a compromise in processing time and lightweight deployment by appropriately shrinking the dimension of the smoothing window.

[0163] like Figure 7 As shown in section (b), the curves illustrating the false negative rate of weak interference components under extremely low system signal-to-noise ratio (SNR) are presented. The horizontal axis represents the system SNR, and the vertical axis represents the false negative rate of weak interference components. Since the interference detection process in this application involves removing useful signal energy from the total received energy, when the system SNR is below approximately -5 dB, the reliability of useful signal power estimation and residual power characterization is significantly affected by noise disturbances, resulting in a marked increase in the false negative rate of weak interference components. This reflects the performance boundary of the method in this application under extremely low SNR conditions. As can be seen from the curves, the false negative rate exhibits an "avalanche-like" upward trend after crossing the approximately -5 dB critical region, indicating that some weak interference components may be difficult to detect stably under extremely low SNR conditions.

[0164] Compared to traditional hard-decision zeroing methods that rely on strict frequency-domain grid alignment, and complex adaptive cancellation techniques that demand external reference signals or involve massive high-order statistical calculations, the frequency-domain adaptive interference cancellation method based on interference power detection provided in this application exhibits significant technical advantages. This application achieves accurate real-time sensing of time-varying non-stationary interference energy by introducing multi-resolution channel estimation and a dynamic detection threshold based on neighborhood residual power statistics. Simultaneously, it abandons the crude hard truncation mechanism of either 1 or 0, and utilizes a nonlinear continuous attenuation model to construct a frequency-domain smooth weight mask, achieving flexible adaptive suppression of each disturbed subcarrier, thus completely avoiding the risks of useful signal energy collapse and waveform secondary distortion from a physical level.

[0165] Furthermore, this application constructs a closed-loop parameter control mechanism based on bidirectional feedback of smooth interference detection rate and real-time demodulation bit error rate, enabling the system's detection sensitivity and frequency domain suppression depth to adaptively iterate with the evolution of the external dynamic electromagnetic environment. Based on the aforementioned simulation test data, the method of this application, while completely eliminating rigid dependence on external independent interference reference channels, significantly reduces the bit error rate (BER) and error vector magnitude (EVM) at the receiver under poor signal-to-interference ratios, greatly improving signal fidelity and interference robustness.

[0166] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of combined actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Secondly, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application. In the above embodiments, the descriptions of each embodiment have their own emphasis; for parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0167] Figure 8 A structural block diagram of an example of a frequency-domain adaptive interference cancellation system based on interference power detection according to an embodiment of this application is shown, the system being deployed at a wireless communication receiver.

[0168] like Figure 8 As shown, the frequency domain adaptive interference cancellation system 800 based on interference power detection includes a frequency domain transformation estimation unit 810, a residual power calculation unit 820, an interference power detection unit 830, a weight mask generation unit 840, a weighted equalization processing unit 850, and a waveform inverse transformation unit 860.

[0169] The frequency domain transformation estimation unit 810 is used to perform a frequency domain transformation operation on the time domain baseband received signal to obtain the received frequency domain coefficients on each subcarrier, and to perform channel estimation processing based on the pilot information in the received frequency domain coefficients to obtain the channel frequency response estimate corresponding to each subcarrier.

[0170] The residual power calculation unit 820 is used to determine the total received power corresponding to each subcarrier based on the received frequency domain coefficients on each subcarrier, and to determine the useful signal power estimate corresponding to each subcarrier based on the channel frequency response estimate, and to remove the corresponding useful signal power estimate from the total received power corresponding to each subcarrier to obtain the residual power on each subcarrier.

[0171] The interference power detection unit 830 is used to determine the dynamic detection threshold corresponding to each subcarrier based on the frequency domain statistical characteristics of the residual power in the neighborhood of each subcarrier, and to compare the residual power on each subcarrier with the corresponding dynamic detection threshold to determine the interference power estimate corresponding to each subcarrier.

[0172] The weight mask generation unit 840 is used to determine the initial mask weights corresponding to each subcarrier based on the interference power estimate corresponding to each subcarrier, and to perform frequency domain smoothing processing on the initial mask weights based on the weight continuity constraint between adjacent subcarriers, so as to obtain a frequency domain smoothed weight mask composed of the mask weights corresponding to each subcarrier; wherein, the mask weights are located within a preset weight interval, and under the condition of satisfying the weight continuity constraint, the mask weights corresponding to the subcarriers with larger interference power estimates are closer to the lower limit of the preset weight interval.

[0173] The weighted equalization processing unit 850 is used to perform weighted attenuation on the received frequency domain coefficients on each subcarrier based on the mask weights corresponding to each subcarrier in the frequency domain smoothing weight mask, so as to obtain the received frequency domain coefficients after flexible interference attenuation, and to perform channel equalization processing on the received frequency domain coefficients after flexible interference attenuation based on the channel frequency response estimate, so as to obtain the frequency domain signal after interference cancellation.

[0174] The waveform inverse transformation unit 860 is used to perform inverse frequency domain transformation on the frequency domain signal after interference cancellation in order to reconstruct the time domain waveform after interference cancellation.

[0175] In some embodiments, this application provides a non-volatile computer-readable storage medium storing one or more programs including execution instructions. The execution instructions can be read and executed by an electronic device (including but not limited to a computer, server, or network device) to perform the steps of any of the frequency domain adaptive interference cancellation methods based on interference power detection described above.

[0176] In some embodiments, this application also provides a computer program product, the computer program product including a computer program stored on a non-volatile computer-readable storage medium, the computer program including program instructions, which, when executed by a computer, cause the computer to perform the steps of any of the above-described frequency domain adaptive interference cancellation methods based on interference power detection.

[0177] In some embodiments, this application also provides an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of a frequency-domain adaptive interference cancellation method based on interference power detection.

[0178] The above-described product can perform the methods provided in the embodiments of this application, and has the corresponding functional modules and beneficial effects for performing the methods. Technical details not described in detail in this embodiment can be found in the methods provided in the embodiments of this application.

[0179] The electronic devices in this application can exist in various forms, including but not limited to: mobile communication devices, ultra-mobile personal computer devices, portable entertainment devices, or other airborne electronic devices with data interaction functions.

[0180] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0181] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0182] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A frequency-domain adaptive interference cancellation method based on interference power detection, applied to a wireless communication receiver, characterized in that, The method includes: A frequency domain transformation operation is performed on the time-domain baseband received signal to obtain the received frequency domain coefficients on each subcarrier, and channel estimation processing is performed based on the pilot information in the received frequency domain coefficients to obtain the channel frequency response estimate corresponding to each subcarrier. The total received power corresponding to each subcarrier is determined based on the received frequency domain coefficients on each subcarrier, and the estimated useful signal power corresponding to each subcarrier is determined based on the channel frequency response estimate. The estimated useful signal power corresponding to each subcarrier is removed from the total received power corresponding to each subcarrier to obtain the residual power on each subcarrier. Based on the frequency domain statistical characteristics of the residual power in the neighborhood of each subcarrier, the dynamic detection threshold corresponding to each subcarrier is determined, and the residual power on each subcarrier is compared with the corresponding dynamic detection threshold to determine the estimated interference power value corresponding to each subcarrier. The initial mask weights for each subcarrier are determined based on the estimated interference power values ​​for each subcarrier. The initial mask weights are then smoothed in the frequency domain based on the weight continuity constraint between adjacent subcarriers to obtain a frequency-domain smoothed weight mask composed of the mask weights for each subcarrier. The mask weights are located within a preset weight interval, and under the condition of satisfying the weight continuity constraint, the mask weights corresponding to subcarriers with larger estimated interference power values ​​are closer to the lower limit of the preset weight interval. Based on the mask weights corresponding to each subcarrier in the frequency domain smoothing weight mask, the received frequency domain coefficients on each subcarrier are weighted and attenuated to obtain the received frequency domain coefficients after flexible interference attenuation. Based on the channel frequency response estimate, the received frequency domain coefficients after flexible interference attenuation are subjected to channel equalization processing to obtain the frequency domain signal after interference cancellation. An inverse frequency domain transform is performed on the frequency domain signal after interference cancellation to reconstruct the time domain waveform after interference cancellation.

2. The method according to claim 1, characterized in that, The channel estimation process based on the pilot information in the received frequency domain coefficients to obtain the channel frequency response estimate corresponding to each subcarrier includes: Extract the pilot subcarrier coefficients located at the preset pilot position set from the received frequency domain coefficients, and combine them with the locally known pilot symbols to obtain the initial channel response at each pilot position; Multiple interpolation filtering windows with different frequency domain span parameters are configured, and frequency domain reconstruction is performed on the initial channel response using each interpolation filtering window to generate candidate channel frequency response sequences at multiple different interpolation resolution scales; Based on the pilot reconstruction error at the preset pilot position set for each interpolation resolution scale, the corresponding scale evaluation error is determined, and the adaptive combination weight corresponding to each interpolation resolution scale is calculated based on the scale evaluation error; wherein, the smaller the scale evaluation error of the interpolation resolution scale, the larger the corresponding adaptive combination weight. The corresponding values ​​of the candidate channel frequency response sequences of the same subcarrier at each interpolation resolution scale are normalized and weighted according to the corresponding assigned adaptive combination weights to generate the final channel frequency response estimate for each subcarrier.

3. The method according to claim 1, characterized in that, The step of removing the corresponding useful signal power estimate from the total received power corresponding to each subcarrier to obtain the residual power on each subcarrier includes: The variance of the channel estimation error is obtained based on the pilot reconstruction residual in the channel estimation process, and the dynamic error compensation coefficient corresponding to each subcarrier is calculated in combination with the reliability of the channel frequency response estimate. Among them, the smaller the channel estimation error, the higher the degree of preservation of the preliminary useful signal estimation power by the dynamic error compensation coefficient; the larger the channel estimation error, the higher the degree of reduction of the preliminary useful signal estimation power by the dynamic error compensation coefficient. Based on the channel frequency response estimate corresponding to each subcarrier and the transmit symbol characterization value obtainable at the receiver, the preliminary estimated power of the useful signal corresponding to each subcarrier is calculated. The dynamic error compensation coefficient is used to perform deviation scaling correction on the preliminary useful signal power estimate to obtain the useful signal power estimate for each subcarrier; The estimated useful signal power is removed from the corresponding total received power, and a non-negative truncation process is performed on the removal result to avoid abnormally negative values ​​in the residual power. The remaining energy after truncation is determined as the residual power on each subcarrier.

4. The method according to claim 1 or 3, characterized in that, The step of determining the dynamic detection threshold corresponding to each subcarrier based on the frequency domain statistical characteristics of the residual power in the neighborhood of each subcarrier, and comparing the residual power on each subcarrier with the corresponding dynamic detection threshold to determine the estimated interference power value corresponding to each subcarrier includes: For the target subcarrier, a frequency domain neighborhood window is defined based on a preset single-sided radius. When the frequency domain neighborhood window exceeds the effective frequency band boundary of the system, an edge mirror extension strategy is used for compensation. The residual power outside the target subcarrier within the frequency domain neighborhood window is extracted to construct a reference power sample set. The reference power sample set is sorted to extract the local sample median and the quantile value of a preset order; The system background noise power estimate is obtained, and the local sample median adjusted by the first adjustment coefficient and the quantile value adjusted by the second adjustment coefficient are combined with the system background noise power estimate to form the dynamic detection threshold corresponding to the target subcarrier; wherein, the local sample median is used to characterize the stable background power level in the neighborhood, and the quantile value is used to characterize the high power fluctuation characteristics in the neighborhood. When the residual power on the target subcarrier exceeds the dynamic detection threshold, it is determined that there is interference on the target subcarrier, and the energy of the excess portion of the residual power relative to the estimated system background noise power is determined after non-negative truncation processing as the corresponding interference power estimate; when the residual power does not exceed the dynamic detection threshold, it is determined that there is no interference and the corresponding interference power estimate is set to zero.

5. The method according to claim 4, characterized in that, The step of determining the initial mask weights corresponding to each subcarrier based on the estimated interference power values ​​corresponding to each subcarrier, and performing frequency domain smoothing processing on the initial mask weights based on the weight continuity constraint between adjacent subcarriers to obtain a frequency domain smoothed weight mask composed of the mask weights corresponding to each subcarrier, includes: Obtain a nonlinear continuous decay model configured with a preset weight protection lower limit. The nonlinear continuous decay model has a mapping characteristic based on an S-shaped function and is controlled by an interference power threshold used to define the suppression start boundary and a slope coefficient used to control the steepness of the decay curve. The interference power estimate for each subcarrier is input into the nonlinear continuous attenuation model for nonlinear mapping transformation to calculate the initial mask weight for each subcarrier. The preset weight protection lower limit is used to prevent the signal information of the corresponding subcarrier from being completely zeroed. When the interference power estimate is lower than the interference power threshold, the initial mask weight approaches the upper limit of the preset weight range. When the interference power estimate is higher than the interference power threshold, the initial mask weight converges to the preset weight protection lower limit as the interference power increases. A frequency domain smoothing window sequence that satisfies the normalization condition and is centrally symmetric is introduced, and combined with a boundary symmetric extension strategy, a one-dimensional convolutional filtering operation is performed on the sequence composed of the initial mask weights corresponding to each subcarrier to eliminate abrupt changes in frequency domain weights. The result after convolutional filtering is determined as the mask weights that satisfy the weight continuity constraint, and the mask weights constitute the frequency domain smoothing weight mask.

6. The method according to claim 1, characterized in that, The step of performing a frequency domain transformation operation on the time-domain baseband received signal to obtain the received frequency domain coefficients on each subcarrier includes: Multiple time truncation windows with different length parameters, including a preset main time window, are used to perform frequency domain transformation operations on the time-domain baseband received signal to generate multiple sets of initial frequency domain coefficient sequences at different frequency resolutions. The initial frequency domain coefficient sequence corresponding to the preset main time window is determined as the received frequency domain coefficient sequence, and the received frequency domain coefficient sequence is used as the alignment reference sequence corresponding to the preset main time window; frequency domain raster resampling and interpolation alignment operations are performed on the initial frequency domain coefficient sequences generated by other time truncation windows to generate multiple sets of aligned frequency domain coefficient sequences that match the frequency raster of the alignment reference sequence; wherein, the alignment reference sequence and the multiple sets of aligned frequency domain coefficient sequences together serve as the aligned frequency domain coefficient sequences corresponding to each time truncation window; Accordingly, the step of determining the interference power estimate for each subcarrier includes: Based on the aligned frequency domain coefficient sequence corresponding to each time truncation window, the single-scale received power under each time truncation window is determined, and the useful signal power estimate of the corresponding scale is independently removed from the single-scale received power to obtain the single-scale residual power under each time truncation window. The single-scale residual power is compared with a single-scale dynamic detection threshold that matches the current frequency resolution to generate a single-scale interference detection flag. When the single-scale residual power exceeds the corresponding single-scale dynamic detection threshold, the single-scale interference detection flag indicates the presence of interference, and the portion exceeding the estimated background noise power at the corresponding scale is identified as single-scale interference power. Simultaneously, based on the magnitude of the single-scale residual power overflowing the single-scale dynamic detection threshold, the corresponding single-scale confidence level is calculated. The single-scale confidence level corresponding to the time truncation window where no interference is detected is set to zero. For each single-scale interference detection flag at the same subcarrier location, a scale fusion decision based on logical OR is performed. When it is confirmed that the current subcarrier is interfered with, the single-scale interference power under each time truncation window is normalized and weighted according to the corresponding single-scale confidence level to obtain the final output interference power estimate. When it is not confirmed that the current subcarrier is interfered with, the final output interference power estimate is set to zero.

7. The method according to claim 5, characterized in that, After reconstructing the time-domain waveform after interference cancellation, the method further includes an adaptive adjustment step of the closed-loop parameters based on demodulation feedback, specifically including: The distribution of subcarrier interference within a preset time sliding window is statistically analyzed and smoothed to obtain the smoothed interference detection rate, and the real-time bit error rate is obtained by combining the demodulation and decoding results. When the smoothed interference detection rate deviates from the preset detection rate control target, and the real-time bit error rate index does not exceed the preset bit error rate tolerance limit, based on the signed detection difference between the smoothed interference detection rate and the detection rate control target, the first adjustment coefficient and the second adjustment coefficient are adaptively updated according to the first preset step size set under the constraint of preset coefficient boundaries, so as to dynamically balance the threshold sensitivity of the system to abnormal power fluctuations; wherein, when the smoothed interference detection rate is higher than the detection rate control target, the dynamic detection threshold in subsequent processing cycles is increased; when the smoothed interference detection rate is lower than the detection rate control target, the dynamic detection threshold in subsequent processing cycles is decreased. When the real-time bit error rate (BER) deviates from the preset target BER, and the smoothed interference detection rate does not exceed the preset detection rate tolerance limit, based on the signed bit error difference between the real-time BER and the target BER, the interference power threshold and slope coefficient in the nonlinear continuous attenuation model are dynamically updated according to a second preset step size set under preset parameter boundary constraints. This adaptively adjusts the suppression depth and activation boundary of the frequency domain weights in subsequent processing cycles. Specifically, when the real-time BER is higher than the target BER, the interference suppression intensity in subsequent processing cycles is enhanced; when the real-time BER is lower than the target BER, the interference suppression intensity in subsequent processing cycles is weakened.

8. A frequency-domain adaptive interference cancellation system based on interference power detection, deployed at a wireless communication receiver, characterized in that, The system includes: The frequency domain transformation estimation unit is used to perform frequency domain transformation operation on the time domain baseband received signal to obtain the received frequency domain coefficients on each subcarrier, and to perform channel estimation processing based on the pilot information in the received frequency domain coefficients to obtain the channel frequency response estimate value corresponding to each subcarrier. The residual power calculation unit is used to determine the total received power corresponding to each subcarrier based on the received frequency domain coefficients on each subcarrier, and to determine the useful signal power estimate corresponding to each subcarrier based on the channel frequency response estimate, and to remove the corresponding useful signal power estimate from the total received power corresponding to each subcarrier to obtain the residual power on each subcarrier. An interference power detection unit is used to determine the dynamic detection threshold corresponding to each subcarrier based on the frequency domain statistical characteristics of the residual power in the neighborhood of each subcarrier, and to compare the residual power on each subcarrier with the corresponding dynamic detection threshold to determine the interference power estimate corresponding to each subcarrier. A weight mask generation unit is configured to determine the initial mask weights corresponding to each subcarrier based on the estimated interference power value corresponding to each subcarrier, and to perform frequency domain smoothing on the initial mask weights based on the weight continuity constraint between adjacent subcarriers, so as to obtain a frequency domain smoothed weight mask composed of the mask weights corresponding to each subcarrier; wherein, the mask weights are located within a preset weight interval, and under the condition of satisfying the weight continuity constraint, the mask weights corresponding to the subcarriers with larger estimated interference power values ​​are closer to the lower limit of the preset weight interval; The weighted equalization processing unit is used to perform weighted attenuation on the received frequency domain coefficients on each subcarrier based on the mask weights corresponding to each subcarrier in the frequency domain smoothing weight mask, so as to obtain the received frequency domain coefficients after flexible interference attenuation, and to perform channel equalization processing on the received frequency domain coefficients after flexible interference attenuation based on the channel frequency response estimate, so as to obtain the frequency domain signal after interference cancellation. The waveform inverse transformation unit is used to perform inverse frequency domain transformation on the frequency domain signal after interference cancellation in order to reconstruct the time domain waveform after interference cancellation.

9. An electronic device, comprising: At least one processor; as well as A memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the method as described in any one of claims 1-7.

10. A storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1-7.

Images