A method of millimeter wave communication interference mitigation

Abstract

This invention discloses a millimeter-wave communication interference suppression method, relating to the field of millimeter-wave communication transmission technology. First, it acquires real-time channel state information across the entire frequency band of the target millimeter-wave communication link, pilot parameters at the transmitting end, the original received signal at the receiving end, and electromagnetic environment scanning data around the link. After completing basic link feature calibration, it performs multi-dimensional spatiotemporal interference feature extraction, filters and calibrates real interference sources, generates a dynamic suppression threshold to complete initial interference suppression, adjusts the phased array antenna weighting coefficients for secondary spatial beamforming suppression, and reconstructs the communication signal after residual interference cancellation. Simultaneously, it incrementally updates the interference feature library. This invention can cover multiple types of multi-domain mixed interference, has high interference identification accuracy, does not damage the effective communication signal, requires no modification to the existing millimeter-wave equipment hardware architecture, and can adapt to the reliability requirements of millimeter-wave communication in various scenarios under complex electromagnetic environments.

Description

Technical Field

[0001] This invention relates to the field of millimeter-wave communication transmission technology, and in particular to a method for suppressing millimeter-wave communication interference. Background Technology

[0002] Millimeter-wave communication, with its high bandwidth and high speed, is widely used in scenarios such as 5G access network backhaul, autonomous driving vehicle-to-everything (V2X) communication, and high-speed data transmission for the Industrial Internet of Things (IIoT). However, millimeter-wave communication operates at relatively high frequencies, resulting in significant propagation path loss and low communication link margins. Therefore, it is highly susceptible to communication interruptions and soaring bit error rates when subjected to deliberate interference or spurious interference from adjacent channels. Interference suppression is one of the core supporting technologies for the practical application of millimeter-wave communication, and the industry continues to demand higher levels of adaptability and accuracy from interference suppression technologies.

[0003] Currently, mainstream millimeter-wave communication interference suppression technologies fall into two categories. The first is frequency-domain fixed-threshold filtering technology. Its working principle involves pre-setting a fixed interference power threshold and directly filtering out frequency subbands in the received signal whose power exceeds the threshold. This technology has simple implementation logic, low hardware computing power requirements, and is widely used in consumer-grade millimeter-wave communication equipment. However, this technology can only handle strong interference with a defined frequency domain; it cannot handle sudden time-domain interference or co-channel interference from the spatial direction. Furthermore, setting the fixed threshold too high can lead to interference leakage, while setting it too low can damage the effective communication signal, making it unsuitable for complex electromagnetic environments with dynamically changing interference power. The second category is adaptive beamforming interference suppression technology. Its working principle involves identifying the spatial direction of the interference through angle-of-arrival estimation and adjusting the phased array antenna weights to generate nulls in the interference direction. This technology can suppress spatial interference without affecting the effective frequency domain signal and is often used in high-reliability industrial millimeter-wave communication scenarios. This type of technology only addresses spatial interference and has no effect on suppressing multi-type mixed interference in the same direction or adjacent channel frequency domain interference. Furthermore, it does not screen for false targets during the interference identification process, which can easily lead to the misjudgment of multipath reflection signals as interference sources, generating invalid nulls, occupying the antenna array's degrees of freedom, and reducing the main beam gain.

[0004] As millimeter-wave communication applications expand to densely populated urban areas, complex industrial parks, and multi-vehicle cooperative road sections, the number and types of interference sources around the communication link continue to increase. The time-domain, frequency-domain, and spatial-domain parameters of the interference change dynamically, and existing single-dimensional interference suppression technologies can no longer meet the communication reliability requirements in complex scenarios. Summary of the Invention

[0005] This invention proposes a millimeter-wave communication interference suppression method to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a millimeter-wave communication interference suppression method, comprising the following steps: The system collects real-time channel status information across the entire frequency band of the target millimeter-wave communication link, including known pilot signal parameters at the transmitting end, the original received signal sequence after synchronization at the receiving end, and electromagnetic environment scan data with 360-degree omnidirectional coverage and time resolution consistent with the communication signal time slot length. Multi-dimensional feature extraction in time domain, frequency domain, and spatial domain is performed on the original received signal sequence. All interference source attributes are calibrated by matching the pre-trained interference feature library, and suppression interference, spoofing interference, adjacent channel interference, and spurious interference are distinguished. Based on the calibrated interference attributes, a three-dimensional interference suppression threshold in space, time, and frequency is generated. The filtering operation is performed synchronously on the time-domain time slot, frequency-domain sub-band, and spatial-domain receiving channel where the interference is located in the original received signal to complete the initial interference suppression. Adjusting the weighting coefficients of each element of the millimeter-wave phased array antenna generates a high-gain main lobe beam pointing to the communication transmitter and a deep null beam pointing to each calibrated interference source. Secondary suppression of spatial interference is achieved through adaptive beamforming. The received signal after secondary suppression is subjected to residual interference cancellation processing. The residual weak interference components are separated sequentially by a serial interference cancellation algorithm, and the target communication signal is accurately reconstructed based on channel state information. The reconstructed communication signal is verified in multiple dimensions, including bit error rate, signal-to-noise ratio, and link transmission rate. If the verification results meet the preset communication quality threshold, valid communication data is directly output; otherwise, the iterative processing from channel state information acquisition to signal reconstruction is re-executed.

[0007] Furthermore, all interference source attributes are calibrated, including: The spatial azimuth confidence of the interference source is calculated using the following formula. The calculation result is used to filter high-confidence interference location results and eliminate falsely identified interference targets. The preset confidence threshold is configured according to the complexity of the electromagnetic environment in which the communication link is located: ; in Indicates the first The azimuth angle of the interference source to be determined. Indicates the first The azimuth confidence level of each interference source to be determined. This indicates the total number of receiving channels for the phased array antenna. Indicates the first Preset weighting coefficients for each receiving channel, Indicates the first Each receiving channel is at the corresponding azimuth angle The spatial spectrum value of the received signal at the azimuth angle, where i is the summation variable for all azimuth angles of the interference sources to be determined. Indicates the first Each receiving channel is at the corresponding azimuth angle The spatial spectrum value of the received signal.

[0008] Furthermore, after labeling all interference source attributes, the process also includes: storing all feature parameters of the newly identified interference types into the interference feature library, while performing weighted iterative updates on the feature parameters of existing interference types. The feature library update process adopts a combination of offline and online methods. Online updates only store feature parameters, while offline updates periodically complete feature clustering and redundancy cleanup.

[0009] Furthermore, a three-dimensional spatiotemporal interference suppression threshold is generated based on the calibrated interference attributes, including: The interference suppression dynamic threshold for each sub-band is calculated using the following formula, taking into account the noise floor of the current channel and the real-time power intensity of the interference: ; in Indicates the first Each sub-band corresponds to a dynamic threshold for interference suppression. The preset weighting coefficients represent the noise floor. Indicates the first The measured Gaussian white noise power corresponding to each sub-band The preset weighting coefficients represent the interference power. Indicates the first The measured interference power corresponding to each subband, and the weighting coefficient of the noise floor. and the weighting coefficient of interference power The dynamic threshold is pre-configured according to the specific application scenario of the millimeter-wave communication link, and the configured threshold is adjusted in real time according to the change of interference power.

[0010] Furthermore, multi-dimensional feature extraction is performed on the original received signal sequence in the time domain, frequency domain, and spatial domain, including: The corresponding frequency domain signal sequence is obtained by performing a fast Fourier transform on the original received signal sequence; Power spectrum calculations are performed on frequency domain signal sequences to obtain the frequency domain distribution characteristics of the interference, including the center frequency, bandwidth, and power spectral density parameters of the interference. The time-domain start and end times and duty cycle characteristics of the interference are obtained by performing sliding window energy detection on the original received signal sequence. For the received signals of each receiving channel, a multi-signal classification algorithm is used to calculate the angle of arrival, thereby obtaining the spatial azimuth and elevation angle distribution characteristics of the interference.

[0011] Furthermore, adaptive beamforming operations are performed, including: The pointing parameters of the main lobe beam are generated based on the known location of the communication transmitter. The pointing parameters of the null beam are generated based on the azimuth and elevation angles of all real interference sources obtained from the calibration. The weighting coefficients of each element of the phased array antenna are solved based on the linear constraint minimum variance criterion, and the beam parameters are configured in real time.

[0012] Furthermore, residual interference cancellation processing is performed on the received signal after secondary suppression, including: Cross-correlation detection of interference components is performed on the received signal after secondary suppression to identify weak residual interference with power below a preset suppression threshold; According to the order of residual interference power from high to low, each residual interference component is separated in turn. After each separation, the signal-to-noise ratio of the remaining received signal is updated until the power of all residual interference is lower than the allowable interference threshold of the communication link.

[0013] Furthermore, before performing multi-dimensional feature extraction in the time, frequency, and spatial domains on the original received signal sequence, the following steps are also included: Calculate the channel's frequency response and delay spread parameters based on pilot signal parameters; A combination of first-order and second-order phase-locked loops is used to compensate for the local oscillator frequency offset and sampling clock offset at the receiving end.

[0014] Furthermore, the electromagnetic environment scanning data acquisition operation is performed, including: The step-scan frequency method is used to cover all operating frequency bands of the target millimeter-wave communication link and adjacent protection frequency bands. The sampling rate of the collected electromagnetic environment scan data is more than twice the highest operating frequency of the communication signal.

[0015] Furthermore, the interference feature library update operation is performed, including: Redundancy is eliminated from the interference feature parameters stored in the feature library. For interference of the same type, the three sets of feature parameter samples with the greatest feature differences are retained. The feature parameters include the modulation method, duty cycle, center frequency, bandwidth and spatial transmission characteristics of the interference. The feature library is stored in a distributed storage mode.

[0016] Compared with existing technologies, the beneficial effects of this invention are: This invention employs a three-dimensional joint interference suppression process involving space, time, and frequency, sequentially completing initial interference filtering in the time and frequency domains, secondary suppression through spatial beamforming, and cancellation of residual interference. It can cover all types of interference in different domains and has the ability to suppress burst interference in the time domain, adjacent channel interference in the frequency domain, and co-channel interference in the spatial domain. Compared with single-dimensional interference suppression techniques, it can handle more types of interference and has stronger anti-interference capabilities for links.

[0017] This invention introduces an azimuth confidence screening mechanism during interference identification, which can eliminate falsely identified interference targets, avoid generating invalid nulls that occupy antenna array degrees of freedom, and ensure the gain level of the main beam. The interference suppression threshold of this invention adopts a dynamic adjustment mechanism, adjusting the filtering threshold of each sub-band based on the current channel noise floor and real-time interference power. This avoids the interference leakage or damage to effective signals caused by fixed thresholds, resulting in higher accuracy in interference identification and filtering.

[0018] This invention features an incrementally updatable interference feature library. Newly identified interference type parameters not yet included in the library can be stored, and feature clustering and redundancy cleanup are performed periodically. This eliminates the need to pre-define feature parameters for all interference types, adapting to dynamic electromagnetic environments where new interferences constantly emerge. The feature library employs a distributed storage architecture, supporting feature sharing among multiple communication nodes, thus meeting the joint interference suppression requirements in multi-node collaborative millimeter-wave communication scenarios.

[0019] The technical solution of this invention does not require the addition of an extra hardware acquisition module and can be directly adapted to the hardware architecture of existing millimeter-wave phased array communication equipment. It can be deployed without modifying existing hardware and is compatible with civilian consumer-grade millimeter-wave equipment, high-reliability industrial millimeter-wave communication equipment, and vehicle-to-everything (V2X) millimeter-wave communication equipment, thus having a wider range of applicable scenarios. Attached Figure Description

[0020] Figure 1 This is a schematic block diagram of the overall control process for millimeter-wave communication interference suppression proposed in this invention. Figure 2 This is a schematic block diagram of the signal multi-source synchronous acquisition and multi-dimensional feature calibration and extraction process proposed in this invention; Figure 3 This is the spatiotemporal frequency three-dimensional dynamic threshold generation and primary suppression control diagram proposed in this invention; Figure 4 This is a flowchart of the azimuth confidence screening and phased array adaptive beamforming proposed in this invention; Figure 5 This is a loop diagram for residual interference serial elimination and communication quality multi-dimensional verification proposed in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Reference Figures 1 to 5A method for suppressing interference in millimeter-wave communication includes the following steps: The system acquires real-time channel state information across the entire frequency band of the target millimeter-wave communication link. It collects known pilot signal parameters from the transmitter, the original received signal sequence after synchronization at the receiver, and electromagnetic environment scan data covering 360 degrees around the link. The time resolution of the collected electromagnetic environment scan data is consistent with the time slot length of the communication signal. For example, the time synchronization error can be less than one sampling point to ensure that the collected interference data is synchronized with the time of the received signal. Based on the pilot sequence, the system completes the basic characteristic calibration of the communication link, eliminating the negative impact of inherent offset of the receiver hardware and large-scale fading of the channel on the signal acquisition accuracy. Multi-dimensional feature extraction in the time domain, frequency domain, and spatial domain is performed on the acquired raw received signal sequence. The type identification, quantity statistics, and real-time transmission parameter calculation of the interference sources are completed in sequence. The attributes of all interference sources are labeled by matching the pre-trained interference feature library, and the intentional suppression interference, spoofing interference, and unintentional adjacent channel interference and spurious interference types are distinguished. Based on the calibrated interference attributes, corresponding spatial-temporal-frequency three-dimensional interference suppression thresholds are generated. Filtering operations are performed synchronously on the time-domain time slot, frequency-domain sub-band, and spatial-domain receiving channel where the interference is located in the original received signal to complete the initial interference suppression process, initially reduce the proportion of interference power in the received signal, and ensure that the signal-to-noise ratio of the received signal after initial suppression meets the minimum requirements of subsequent beamforming processing. By adjusting the weighting coefficients of each element of the millimeter-wave phased array antenna, a high-gain main lobe beam pointing to the communication transmitter and a deep null beam pointing to each identified interference source are generated. Secondary suppression of spatial interference is achieved through adaptive beamforming, further reducing the interference penetration power in the spatial direction and ensuring that the gain in the main lobe direction meets the transmission distance requirements of the communication link. The received signal after secondary suppression is subjected to residual interference cancellation processing. The residual weak interference components are separated sequentially using a serial interference cancellation algorithm. The target communication signal is accurately reconstructed based on channel state information, thereby improving the purity of the communication signal and reducing the bit error rate in the signal demodulation process. The reconstructed communication signal undergoes multi-dimensional verification of bit error rate, signal-to-noise ratio, and link transmission rate. If the verification result meets the preset communication quality threshold, valid communication data is directly output. If not, the iterative processing from the channel state information acquisition step to the signal reconstruction step is repeated until the verification result meets the requirements.

[0023] The aforementioned channel state information is a set of parameters characterizing the transmission characteristics of a wireless communication link, including channel amplitude, phase, fading, and transmission loss. It serves as the foundation for signal processing, interference suppression, and signal reconstruction. Pilot signals are known reference signals transmitted by the transmitter, used by the receiver to perform channel estimation, signal synchronization, and hardware error calibration. The original received signal sequence is a baseband signal data stream directly acquired by the receiver antenna without noise reduction or anti-interference processing. Electromagnetic environment scan data is electromagnetic signal data obtained from omnidirectional detection of the space surrounding the communication link, used to identify external interference sources in the environment. A time slot is the smallest time unit divided in the wireless communication frame structure, used to distinguish between communication and interference signals in different time periods. Time resolution is the smallest resolvable time interval of the data acquisition system, used to ensure time axis alignment and synchronization of multiple data streams. Inherent hardware offset refers to fixed hardware errors such as DC offset, amplitude offset, and phase offset inherent in the receiver's RF and baseband circuits. Large-scale channel fading is a large-scale power attenuation phenomenon caused by factors such as transmission distance and obstruction during wireless signal propagation. Time-domain, frequency-domain, and spatial-domain features are signal characteristics extracted from three dimensions: signal time waveform, spectral distribution, and incident spatial angle, respectively. They are the core basis for interference identification. The interference feature database is a pre-stored database of typical features of various types of interference, which achieves automatic identification of interference types and attribute labeling through feature matching.

[0024] Suppression jamming is a type of deliberate interference that uses high-power signals to suppress target communication signals, significantly reducing the received signal-to-noise ratio. Spoofing jamming is also a type of deliberate interference that mimics the characteristics of legitimate communication signals, misleading the receiver's demodulation and disrupting normal communication. Adjacent-channel interference is unintentional interference caused by signal spectrum leakage from adjacent communication frequency bands. Spurious interference is also unintentional interference generated by radio frequency spurious radiation from communication equipment and electromagnetic radiation from surrounding electronic devices. The space-time-frequency three-dimensional interference suppression threshold is a power judgment threshold set in three dimensions: space, time, and frequency. Signals exceeding the threshold are judged as interference and filtered out. A millimeter-wave phased array antenna is an array antenna composed of multiple antenna elements, with beam pointing and gain adjusted electronically. Element weighting coefficients are amplitude and phase weights assigned to each antenna element, used to control beam pointing, gain, and null positions. The main lobe beam is the region with the highest gain in the antenna beam, aligned with the legitimate transmitter to ensure effective communication transmission. The null beam is an artificially created deep attenuation region in the antenna beam, aligned with the interference source to suppress the reception of interference signals. Adaptive beamforming is a spatial anti-interference technique that dynamically adjusts antenna weighting coefficients based on the real-time electromagnetic environment to adaptively generate main lobes and nulls. The serial interference cancellation algorithm is a signal processing algorithm that sequentially separates and cancels residual weak interference. It sorts the residual interference components from highest to lowest power, demodulates and reconstructs each component, and then subtracts the reconstructed interference component from the current received signal. After each component stripping, the signal-to-noise ratio of the received signal is recalculated until all residual interference power meets the target. Signal reconstruction relies on channel state information to repair the target signal damaged by interference, restoring a clean and effective communication signal. The bit error rate (BER) is the ratio of the number of erroneous symbols to the total number of symbols after demodulation at the receiver, and is a core indicator for evaluating communication quality. Iterative processing involves repeatedly performing the entire process of signal acquisition, processing, suppression, and reconstruction when communication quality does not meet requirements, until the target is achieved.

[0025] In this invention, the attributes of all interference sources are calibrated, including: The spatial azimuth confidence of the interference source is calculated using the following formula. The calculation result is used to filter high-confidence interference location results, eliminate falsely identified interference targets, and reduce the amount of unnecessary calculation in the subsequent beamforming process. The preset confidence threshold is configured according to the complexity of the electromagnetic environment in which the communication link is located. The more complex the electromagnetic environment, the higher the corresponding confidence threshold is set, further reducing the probability of false interference misjudgment. ; in Indicates the first The azimuth angle of the interference source to be determined. Indicates the first The azimuth confidence level of each interference source to be determined. This indicates the total number of receiving channels for the phased array antenna. Indicates the first The preset weighting coefficients for each receiving channel can be configured based on the channel noise level or receiving gain, and are used to adjust the weight of each channel in the spatial spectrum calculation. Indicates the first Each receiving channel is at the corresponding azimuth angle The spatial spectrum value of the received signal at the azimuth angle, where i is the summation variable for all azimuth angles of the interference sources to be determined. Let be the azimuth angle of the i-th interference source to be determined. Indicates the first Each receiving channel is at the corresponding azimuth angle The received signal spatial spectrum value is used to calculate the confidence level value. If the confidence level value is within the range of 0 to 1, the interference source with a confidence level greater than the preset confidence level threshold is determined to be a real interference source, and the rest are determined to be false interference sources and are eliminated.

[0026] In this invention, after calibrating the attributes of all interference sources, the method further includes: storing the full set of feature parameters of the newly identified interference type into the interference feature library, while performing weighted iterative updates on the feature parameters of existing interference types to continuously optimize the accuracy of feature matching and reduce the identification error of new interference. The feature library update process adopts a combination of offline and online methods. Online updates only store feature parameters, while offline updates periodically complete feature clustering and redundancy cleanup.

[0027] In this invention, a three-dimensional spatiotemporal interference suppression threshold is generated based on calibrated interference attributes, including: The dynamic threshold for interference suppression in each sub-band is calculated using the following formula. The calculation process incorporates the current channel noise floor and the real-time power intensity of the interference to achieve adaptive dynamic adjustment of the threshold. The threshold serves as the power threshold for interference detection; when the measured power within a sub-band exceeds the threshold, it is determined to be an interference component and filtered out. For sub-bands with high interference power, the threshold increases synchronously with the interference power to avoid misclassifying noise and interference as valid signals. For interference-free sub-bands, the threshold decreases with the noise floor to avoid mistakenly filtering out valid signals, reduce signal loss, and ensure the integrity of the valid signal. ; in Indicates the first Each sub-band corresponds to a dynamic threshold for interference suppression. The preset weighting coefficients represent the noise floor. Indicates the first The measured Gaussian white noise power corresponding to each sub-band (e.g., noise power measured during periods of no signal and no interference). The preset weighting coefficients represent the interference power. Indicates the first The measured interference power corresponding to each subband, and the weighting coefficient of the noise floor. and the weighting coefficient of interference power Pre-configured according to the specific application scenario of the millimeter-wave communication link, for example, the signal-to-noise ratio can be appropriately increased in low signal-to-noise ratio scenarios. To enhance interference suppression capabilities, the signal-to-noise ratio can be appropriately increased in high signal-to-noise ratio scenarios. To reduce the risk of false signal filtering, the configured dynamic threshold is adjusted in real time according to the change of interference power, ensuring the integrity of interference filtering while minimizing damage to effective communication signals.

[0028] In this invention, multi-dimensional feature extraction in the time domain, frequency domain, and spatial domain of the original received signal sequence can include: performing a Fast Fourier Transform on the original received signal sequence to obtain the corresponding frequency domain signal sequence; calculating the power spectrum of the frequency domain signal sequence to obtain the frequency domain distribution characteristics of the interference, including the center frequency, bandwidth, and power spectral density parameters of the interference, which are used for subsequent interference type identification and frequency domain subband suppression; performing sliding window energy detection on the original received signal sequence to obtain the time domain start and end times and duty cycle characteristics of the interference, where the sliding window energy detection identifies interference periods with significantly higher energy than the noise floor by setting an energy threshold, thereby determining the start and end times and duty cycle of the interference; and using a multi-signal classification algorithm to calculate the angle of arrival for the received signals of each receiving channel to obtain the spatial azimuth and elevation angle distribution characteristics of the interference. The multiple signal classification algorithm is a high-resolution angle-of-arrival estimation algorithm that can obtain the spatial angle information of the interference source, providing a basis for subsequent beamforming. The algorithm first constructs the covariance matrix of the signals of each receiving channel, performs eigenvalue decomposition on the covariance matrix to divide the signal subspace and noise subspace, constructs the spatial spectrum function using the orthogonality of the two subspaces, and searches for the spatial spectrum peak by traversing the angle interval. The angle corresponding to the peak is the azimuth and elevation angle of the interference source.

[0029] In this invention, the adaptive beamforming operation can include: generating pointing parameters for the main lobe beam based on the known location of the communication transmitter; generating pointing parameters for the null beam based on the azimuth and elevation angles of all real interference sources obtained from calibration; and solving for the weighting coefficients of each element of the phased array antenna based on the linear constraint minimum variance criterion. This criterion can form nulls in the interference direction while ensuring the main lobe gain, thus completing the real-time configuration of beam parameters. This ensures that the main lobe gain meets the transmission distance requirements of the communication link while the null depth meets the interference suppression requirements. The null depth is dynamically adjusted according to the actual power of the interference; the higher the interference power, the deeper the null depth is set to ensure that high-power interference can be effectively suppressed while avoiding unnecessary attenuation of the main lobe communication signal. Specifically, the linear constraint minimum variance criterion constructs a constraint matrix with the direction of the legitimate communication transmitter as the main constraint and the directions of each interference source as null constraints, and calculates the optimal weighting coefficients for each element with the minimum output power of the antenna array as the optimization objective.

[0030] In this invention, residual interference cancellation processing is performed on the received signal after secondary suppression, which may include: performing cross-correlation detection on the received signal after secondary suppression, identifying weak residual interference with power below a preset suppression threshold by comparing it with a known interference feature template; separating each residual interference component in descending order of power, updating the signal-to-noise ratio of the remaining received signal after each separation, prioritizing the separation of high-power interference can reduce the residual interference effect in subsequent steps and improve the cancellation efficiency, until the power of all residual interference is lower than the allowable interference threshold of the communication link. During the residual interference separation process, demodulation and reconstruction are performed in conjunction with the known modulation method of the interference, and the reconstructed interference component is subtracted from the received signal to eliminate the quadrature and in-phase components of the residual interference.

[0031] In this invention, before performing multi-dimensional feature extraction in the time, frequency, and spatial domains on the original received signal sequence, the method further includes: calculating the channel frequency response and delay spread parameters based on the pilot signal parameters to provide a channel state basis for subsequent signal compensation and interference feature extraction; using a combination of first-order and second-order phase-locked loops to compensate for the local oscillator frequency offset and sampling clock offset at the receiver, eliminating the negative impact of inherent hardware deviations at the receiver on the subsequent interference identification process, and improving the accuracy of interference feature extraction. The first-order phase-locked loop is used to quickly track large frequency offsets, and the second-order phase-locked loop is used to lock small frequency offsets with high precision, ensuring the compensation speed under large frequency offsets while taking into account the compensation accuracy under small frequency offsets.

[0032] In this invention, performing electromagnetic environment scanning data acquisition operations may include: using a step-scanning frequency method to cover the entire operating frequency band of the target millimeter-wave communication link and the adjacent guard band; the sampling rate of the acquired electromagnetic environment scanning data is more than twice the highest operating frequency of the communication signal to ensure that the acquired interference signals have no spectral aliasing and to avoid the problem of interference being missed; the frequency scanning step size is set to one-tenth of the sub-band bandwidth of the communication signal to ensure the identification accuracy of narrowband interference and to avoid interference positioning errors caused by excessive step size.

[0033] In this invention, the operation of updating the interference feature library may include: removing redundancy from the interference feature parameters stored in the feature library; removing duplicate or highly similar samples by calculating feature similarity; retaining the three sets of feature parameter samples with the greatest feature differences for the same type of interference to cover the typical feature range of that type of interference. The feature parameters include the modulation method, duty cycle, center frequency, bandwidth, and spatial transmission characteristics of the interference. This reduces the computational load of the feature matching process and improves the response speed of interference identification. The feature library is stored in a distributed storage manner, which supports feature library sharing among multiple communication nodes and improves the collaborative efficiency of multi-node joint interference suppression.

[0034] The following two examples further illustrate the specific implementation of this system: Example 1 is applied to a millimeter-wave backhaul link scenario for a production private network in an industrial park. The two ends of the backhaul link are deployed on the high-level support structure of the core computer room in the park and the industrial base station support structure at the edge of the park, respectively. The link transmission distance is adapted to the conventional coverage range of industrial-grade millimeter-wave communication equipment. The communication operating frequency band is 24GHz to 26GHz. The surrounding area has wireless control signals from industrial automation equipment, adjacent channel signals from cross-workshop equipment interaction, and illegally installed suppression interference sources. The time-domain and frequency-domain parameters of various interferences change dynamically with the production rhythm.

[0035] Before officially starting the interference suppression process, configure the electromagnetic environment scan coverage range to be 23GHz to 27GHz, fully covering the entire operating frequency band and the guard bands 1GHz above and below. Set the scan step size to one-tenth of the communication subband bandwidth and the scan sampling rate to three times the highest operating frequency of the communication signal to ensure that the collected interference signal has no spectral aliasing. The time resolution of the scan data is completely consistent with the time slot length of the communication signal to ensure that the collected interference data is time-synchronized with the original received signal at the receiving end.

[0036] After the parameter configuration is completed, the frequency response and delay spread parameters of the channel are calculated based on the pilot signal parameters agreed upon by the transmitter. A combination of first-order and second-order phase-locked loops is used to compensate for the local oscillator frequency offset and sampling clock offset of the receiver, correct the inherent offset of the receiver hardware and the large-scale fading error of the channel, eliminate the negative impact of hardware deviation on the subsequent interference identification process, and improve the accuracy of subsequent feature extraction.

[0037] After link calibration is completed, multi-dimensional feature extraction is performed on the original received signal sequence after synchronization at the receiver. First, a Fast Fourier Transform is performed on the original received signal sequence to obtain the corresponding frequency domain signal sequence. Welch power spectrum estimation is then performed on the frequency domain signal sequence to obtain three types of frequency domain distribution features of the interference: center frequency, bandwidth, and power spectral density. Next, sliding window energy detection is performed on the original received signal sequence to obtain the time domain start and end times and duty cycle features of the interference. Finally, the angle of arrival is estimated using a multi-signal classification algorithm on the received signals of all phased array receiving channels to obtain the spatial azimuth and elevation angle distribution features of the interference. After all features are extracted, the spatial azimuth confidence of each interference source to be determined is calculated. A corresponding confidence threshold is set based on the electromagnetic environment complexity of the industrial park. Finally, three groups of real interference sources with confidence scores higher than the threshold are selected, corresponding to industrial wireless control signals, cross-workshop adjacent channel interference, and suppression interference, respectively. The attributes of the three types of interference are then labeled by matching them with a pre-stored interference feature library.

[0038] After completing the interference attribute calibration, the first step is to generate a three-dimensional dynamic threshold for interference suppression in space, time, and frequency by combining the noise floor of the current channel with the real-time interference power of each sub-band. The threshold is raised for the frequency sub-band containing the interference and lowered for the non-interfering sub-band. Filtering operations are simultaneously performed on the time-domain slots, frequency sub-bands, and spatial receiving channels containing the interference in the original received signal to complete initial interference suppression. The signal-to-noise ratio of the received signal after initial suppression meets the input requirements for subsequent beamforming processing. Then, the main lobe beam pointing parameters are generated based on the known position of the communication transmitter, and null beam pointing parameters are generated based on the azimuth and elevation angles of the three real interference sources obtained from calibration. The weighting coefficients of each element of the phased array antenna are solved based on the linear constraint minimum variance criterion to generate a high-gain main lobe beam pointing towards the transmitter. Simultaneously, null beams pointing towards the three interference sources at corresponding depths are generated. The null depth is dynamically adjusted with the interference power; the higher the interference power, the deeper the null depth. Secondary spatial interference suppression is achieved through adaptive beamforming, reducing the interference penetration power in the spatial direction.

[0039] After secondary suppression, residual interference cancellation is performed on the received signal. First, cross-correlation detection of interference components is performed on the received signal after secondary suppression to identify weak residual interference with power below the initial suppression threshold. Each residual interference component is separated sequentially in descending order of power. After each separation, the signal-to-noise ratio of the remaining received signal is updated until the power of all residual interference is below the allowable interference threshold of the communication link. During the separation process, demodulation and reconstruction are performed in conjunction with the known modulation scheme of the interference to eliminate the quadrature and in-phase components of the residual interference. Then, the target communication signal is accurately reconstructed based on channel state information. The reconstructed communication signal undergoes multi-dimensional verification of bit error rate, signal-to-noise ratio, and link transmission rate. Valid return data is output after the verification results meet the preset communication quality threshold. The suppression-type interference identified in this study is a new type that has not been included in the interference feature library. All of its feature parameters are stored in the interference feature library. During the online update phase, only the feature parameters are written. During the offline update phase when the link is idle, feature clustering and redundancy cleanup are completed. For the same type of interference, only the three sets of feature parameter samples with the largest feature differences are retained to reduce the computational load of subsequent feature matching. The updated feature library is synchronized to all networked millimeter-wave communication nodes in the park to meet the collaborative requirements of multi-node joint interference suppression.

[0040] This embodiment employs a three-dimensional joint space-time-frequency interference suppression process, capable of simultaneously handling multi-domain mixed interference in industrial scenarios. It covers various types of interference, including adjacent channel interference, industrial control signal interference, and suppression interference. Compared to traditional single-dimensional suppression techniques, it can handle a wider range of interference types and provides stronger link anti-interference capabilities. The azimuth confidence screening mechanism avoids misidentifying multipath reflections from surrounding metal buildings and production equipment as interference sources, preventing the generation of invalid nulls that occupy array degrees of freedom, and maintaining stable main beam gain. The dynamic threshold adjustment mechanism adapts to the dynamic fluctuations in interference power with production rhythm, preventing interference leakage or damage to effective signals. This results in higher accuracy in interference suppression and meets the reliability requirements of production network backhaul links in complex electromagnetic environments within industrial parks.

[0041] Example 2 is applied to a vehicle-road cooperative millimeter-wave direct communication scenario on an urban main road. The roadside communication unit is deployed on the support structure of public facilities next to the road, and the vehicle-mounted communication unit is deployed on the top of moving vehicles. The communication operating frequency band is 28GHz to 29GHz. There are interferences from vehicle-to-vehicle communication on the same frequency in adjacent lanes, wireless transmission interference from road high-definition monitoring equipment, and millimeter-wave traffic radar interference deployed by surrounding commercial buildings. The airspace azimuth angle of various interferences changes continuously and dynamically with the movement of vehicles.

[0042] Before officially starting the interference suppression process, the electromagnetic environment scan coverage range is configured to be 27.5GHz to 29.5GHz, fully covering the entire operating frequency band and the guard bands above and below it by 0.5GHz. The scan step size is set to one-tenth of the communication subband bandwidth, and the scan sampling rate is set to more than twice the highest operating frequency of the communication signal to ensure that the collected interference signal has no spectral aliasing. The time resolution of the scan data is completely consistent with the time slot length of the communication signal to ensure that the collected interference data is time-synchronized with the original received signal of the vehicle unit.

[0043] After completing the parameter configuration, the frequency response and delay spread parameters of the channel are calculated based on the pilot signal parameters pre-agreed by the roadside communication unit. A combination of first-order and second-order phase-locked loops is used to compensate for the Doppler frequency offset and sampling clock offset caused by the high-speed movement of the vehicle unit, correct the inherent offset of the receiver hardware and the large-scale fading error of the channel, eliminate the negative impact of hardware deviation and mobility errors on the subsequent interference identification process, and improve the accuracy of subsequent feature extraction.

[0044] After link calibration, multi-dimensional feature extraction is performed on the original received signal sequence after synchronization of the vehicle-mounted unit. First, a Fast Fourier Transform is performed on the original received signal sequence to obtain the corresponding frequency domain signal sequence. Welch power spectrum estimation is performed on the frequency domain signal sequence to obtain three types of frequency domain distribution features of the interference: center frequency, bandwidth, and power spectral density. Then, sliding window energy detection is performed on the original received signal sequence to obtain the time domain start and end times and duty cycle features of the interference. Finally, the angle of arrival is estimated by a multi-signal classification algorithm on the received signals of all receiving channels of the vehicle-mounted phased array to obtain the spatial azimuth and elevation angle distribution features of the interference. After all features are extracted, the spatial azimuth confidence of each interference source to be determined is calculated. A corresponding confidence threshold is set based on the electromagnetic environment complexity of the urban main road. Finally, three groups of real interference sources with confidence scores higher than the threshold are selected, corresponding to vehicle-to-vehicle communication co-frequency interference from adjacent vehicles, wireless transmission interference from road monitoring equipment, and millimeter-wave radar interference from commercial buildings, respectively. The attributes of the three types of interference are labeled by matching them with a pre-stored interference feature library.

[0045] After completing the interference attribute calibration, the first step is to generate a three-dimensional dynamic threshold for interference suppression in space, time, and frequency by combining the noise floor of the current channel with the real-time interference power of each sub-band. The threshold is raised for the frequency sub-band containing the interference and lowered for the non-interfering sub-band. Filtering operations are simultaneously performed on the time-domain slots, frequency sub-bands, and spatial receiving channels containing the interference in the original received signal to complete initial interference suppression. The signal-to-noise ratio of the received signal after initial suppression meets the input requirements for subsequent beamforming processing. Subsequently, main lobe beam pointing parameters are generated based on the real-time positioning data of the roadside communication unit. Null beam pointing parameters are generated based on the azimuth and elevation angles of the three real interference sources obtained from the calibration. The weighting coefficients of each element of the vehicle-mounted phased array antenna are solved based on the linear constraint minimum variance criterion to generate a high-gain main lobe beam pointing towards the roadside communication unit. Simultaneously, null beams pointing towards the three interference sources at corresponding depths are generated. The null depth is dynamically adjusted with the interference power; the higher the interference power, the deeper the null depth. Secondary spatial interference suppression is achieved through adaptive beamforming, reducing the interference penetration power in the spatial direction.

[0046] After secondary suppression, residual interference cancellation is performed on the received signal. First, cross-correlation detection of interference components is performed on the received signal after secondary suppression to identify weak residual interference with power below the initial suppression threshold. Each residual interference component is separated sequentially in descending order of power. After each separation, the signal-to-noise ratio of the remaining received signal is updated until the power of all residual interference is below the allowable interference threshold of the communication link. During the separation process, demodulation and reconstruction are performed in conjunction with the known modulation scheme of the interference to eliminate the quadrature and in-phase components of the residual interference. Then, the target communication signal is accurately reconstructed based on channel state information. The reconstructed communication signal undergoes multi-dimensional verification of bit error rate, signal-to-noise ratio, and link transmission rate. After the verification results meet the preset communication quality threshold, traffic warning data issued by the roadside unit is output. The millimeter-wave radar interference identified in this study already exists in the interference feature library. Weighted iterative updates are performed on its feature parameters to optimize the accuracy of subsequent feature matching. Feature clustering and redundancy cleanup are completed during the offline update phase when the link is idle. Only the three sets of feature parameter samples with the largest feature differences for the same type of interference are retained to reduce the computational load of subsequent feature matching. The updated feature library is synchronized to all surrounding networked vehicle-mounted units and roadside units to meet the collaborative requirements of multi-node joint interference suppression.

[0047] This embodiment employs a three-dimensional joint space-time-frequency interference suppression process, capable of simultaneously handling dynamic multi-domain hybrid interference in vehicle-to-infrastructure (V2I) scenarios. It covers various types of interference, including co-channel interference, wireless transmission interference, and radar interference. Compared to traditional single-dimensional suppression techniques, it handles a wider range of interference types and offers stronger link anti-interference capabilities. The azimuth confidence screening mechanism avoids misidentifying multipath reflections from surrounding road infrastructure and adjacent vehicles as interference sources, preventing the generation of invalid nulls that occupy array degrees of freedom, and maintaining stable main beam gain. The incrementally updatable feature library adapts to newly emerging interference types, and the distributed storage architecture supports multi-node sharing of feature parameters, meeting the communication reliability requirements of highly dynamic V2I scenarios.

[0048] Reference Figure 2This diagram details the decoupled parallel logic of high-precision synchronous signal acquisition and multi-dimensional feature extraction. The acquisition end uses a step-sweep method to obtain omnidirectional electromagnetic environment data without spectral aliasing, maintaining strict synchronization with the original received signal sequence at the time slot length level. During the calibration phase, the system utilizes a combined first- and second-order phase-locked loop to precisely compensate for the local oscillator frequency offset and sampling clock offset, counteracting inherent hardware deviations and large-scale fading. The calibrated sequence is then split into three dimensions for feature mining: in the time domain, a sliding window energy detection method locks in the start and end times and duty cycle; in the frequency domain, a fast Fourier transform and power spectrum estimation are used to obtain the center frequency and bandwidth; and in the spatial domain, a multi-signal classification algorithm is used to estimate the angle of arrival. Finally, these multi-dimensional features are matched with a real-time incrementally updated distributed feature library to complete interference source classification and attribute labeling.

[0049] Reference Figure 3 This figure illustrates the adaptive generation of the three-dimensional interference suppression threshold and the primary filtering elimination mechanism in the space-time-frequency domain. The system synchronously measures the Gaussian white noise power and real-time interference power within each sub-band. Based on the specific application scenario of the communication link, it assigns corresponding noise floor weighting coefficients and interference weighting coefficients, adaptively calculating the dynamic interference suppression threshold for each sub-band. This threshold can flexibly and dynamically adjust in response to real-time changes in interference power, automatically lowering the threshold in interference-free sub-bands and automatically raising it in strongly interfering sub-bands. Subsequently, the primary suppression operation maps this dynamic threshold to the three physical dimensions of space, time, and frequency, accurately locating the time-domain time slot, frequency-domain sub-band, and spatial-domain receiving channel where the interference exists in the original signal, and simultaneously performing filtering. This process, while minimizing damage to the effective communication signal, initially reduces the interference power, ensuring that the signal-to-noise ratio of the primary suppressed signal meets the minimum requirements for beamforming.

[0050] Reference Figure 4 This figure illustrates the phased array adaptive beamforming control process based on spatial spectrum confidence screening. After spatial azimuth estimation at the receiver, the system calculates the azimuth confidence of each interference source to be identified by combining the total number of antenna receiving channels, preset weighting coefficients for each channel, and the spatial spectrum value of the received signal at the azimuth angle. By comparing this confidence value with a preset threshold configured based on electromagnetic environment complexity, falsely identified targets are directly eliminated, thereby significantly reducing the amount of unnecessary computation in subsequent beamforming algorithms. For the selected real interference sources, the system generates main lobe pointing parameters in conjunction with the known position of the communication transmitter. Based on the linear constraint minimum variance criterion, the weighting coefficients of each array element are solved and configured in real time, enabling the millimeter-wave phased array antenna to generate a high-gain main lobe and a deep null beam that dynamically deepens with the interference power, further blocking interference penetration in the spatial direction.

[0051] Reference Figure 5This diagram details the iterative closed-loop control logic for residual interference cancellation and multi-dimensional communication quality verification. The received signal, after beamforming secondary suppression, first enters the cross-correlation detection module, which identifies weak residual interference components with power below the initial suppression threshold. The system sorts the residual interference in descending order of power and uses a serial interference cancellation algorithm to demodulate and reconstruct the quadrature and in-phase components of the interference, which are then sequentially stripped. After each separation, the signal-to-noise ratio (SNR) of the current received signal is dynamically updated until all residual interference power is below the tolerable threshold. The reconstructed clean signal undergoes multi-dimensional verification of bit error rate, SNR, and link transmission rate. If the verification results meet the standards, valid communication data is directly output; if not, the system is forced to return to the initial channel state information acquisition stage, forming a rolling adaptive adjustment closed loop.

[0052] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for suppressing interference in millimeter-wave communication, characterized in that, Includes the following steps: The system collects real-time channel status information across the entire frequency band of the target millimeter-wave communication link, including known pilot signal parameters at the transmitting end, the original received signal sequence after synchronization at the receiving end, and electromagnetic environment scan data with 360-degree omnidirectional coverage and time resolution consistent with the communication signal time slot length. Multi-dimensional feature extraction in time domain, frequency domain, and spatial domain is performed on the original received signal sequence. All interference source attributes are calibrated by matching the pre-trained interference feature library, and suppression interference, spoofing interference, adjacent channel interference, and spurious interference are distinguished. Based on the calibrated interference attributes, a three-dimensional interference suppression threshold in space, time, and frequency is generated. The filtering operation is performed synchronously on the time-domain time slot, frequency-domain sub-band, and spatial-domain receiving channel where the interference is located in the original received signal to complete the initial interference suppression. Adjusting the weighting coefficients of each element of the millimeter-wave phased array antenna generates a high-gain main lobe beam pointing to the communication transmitter and a deep null beam pointing to each calibrated interference source. Secondary suppression of spatial interference is achieved through adaptive beamforming. The received signal after secondary suppression is subjected to residual interference cancellation processing. The residual weak interference components are separated sequentially by a serial interference cancellation algorithm, and the target communication signal is accurately reconstructed based on channel state information. The reconstructed communication signal is verified in multiple dimensions, including bit error rate, signal-to-noise ratio, and link transmission rate. If the verification results meet the preset communication quality threshold, valid communication data is directly output; otherwise, the iterative processing from channel state information acquisition to signal reconstruction is re-executed.

2. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Define all interference source attributes, including: The spatial azimuth confidence of the interference source is calculated using the following formula. The calculation result is used to filter high-confidence interference location results and eliminate falsely identified interference targets. The preset confidence threshold is configured according to the complexity of the electromagnetic environment in which the communication link is located: ； in Indicates the first The azimuth angle of the interference source to be determined. Indicates the first The azimuth confidence level of each interference source to be determined. This indicates the total number of receiving channels for the phased array antenna. Indicates the first Preset weighting coefficients for each receiving channel, Indicates the first Each receiving channel is at the corresponding azimuth angle The spatial spectrum value of the received signal at the azimuth angle, where i is the summation variable for all azimuth angles of the interference sources to be determined. Indicates the first Each receiving channel is at the corresponding azimuth angle The spatial spectrum value of the received signal.

3. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, After defining the attributes of all interference sources, the following is also included: All feature parameters of the newly identified interference types are stored in the interference feature library. At the same time, the feature parameters of existing interference types are updated iteratively with weights. The feature library update process adopts a combination of offline and online methods. Online updates only store feature parameters, while offline updates periodically perform feature clustering and redundancy cleanup.

4. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Based on the calibrated interference properties, a three-dimensional spatiotemporal interference suppression threshold is generated, including: The interference suppression dynamic threshold for each sub-band is calculated using the following formula, taking into account the noise floor of the current channel and the real-time power intensity of the interference: ； in Indicates the first Each sub-band corresponds to a dynamic threshold for interference suppression. The preset weighting coefficients represent the noise floor. Indicates the first The measured Gaussian white noise power corresponding to each sub-band The preset weighting coefficients represent the interference power. Indicates the first The measured interference power corresponding to each subband, and the weighting coefficient of the noise floor. and the weighting coefficient of interference power The dynamic threshold is pre-configured according to the specific application scenario of the millimeter-wave communication link, and the configured threshold is adjusted in real time according to the change of interference power.

5. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Multi-dimensional feature extraction is performed on the original received signal sequence in the time domain, frequency domain, and spatial domain, including: The corresponding frequency domain signal sequence is obtained by performing a fast Fourier transform on the original received signal sequence; Power spectrum calculations are performed on frequency domain signal sequences to obtain the frequency domain distribution characteristics of the interference, including the center frequency, bandwidth, and power spectral density parameters of the interference. The time-domain start and end times and duty cycle characteristics of the interference are obtained by performing sliding window energy detection on the original received signal sequence. For the received signals of each receiving channel, a multi-signal classification algorithm is used to calculate the angle of arrival, thereby obtaining the spatial azimuth and elevation angle distribution characteristics of the interference.

6. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Perform adaptive beamforming operations, including: The pointing parameters of the main lobe beam are generated based on the known location of the communication transmitter. The pointing parameters of the null beam are generated based on the azimuth and elevation angles of all real interference sources obtained from the calibration. The weighting coefficients of each element of the phased array antenna are solved based on the linear constraint minimum variance criterion, and the beam parameters are configured in real time.

7. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Residual interference cancellation processing is performed on the received signal after secondary suppression, including: Cross-correlation detection of interference components is performed on the received signal after secondary suppression to identify weak residual interference with power below a preset suppression threshold; According to the order of residual interference power from high to low, each residual interference component is separated in turn. After each separation, the signal-to-noise ratio of the remaining received signal is updated until the power of all residual interference is lower than the allowable interference threshold of the communication link.

8. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Before performing multi-dimensional feature extraction in the time, frequency, and spatial domains on the original received signal sequence, the following steps are also included: Calculate the channel's frequency response and delay spread parameters based on pilot signal parameters; A combination of first-order and second-order phase-locked loops is used to compensate for the local oscillator frequency offset and sampling clock offset at the receiving end.

9. The millimeter-wave communication interference suppression method according to claim 1, characterized in that, Perform electromagnetic environment scanning data acquisition operations, including: The step-scan frequency method is used to cover all operating frequency bands of the target millimeter-wave communication link and adjacent protection frequency bands. The sampling rate of the collected electromagnetic environment scan data is more than twice the highest operating frequency of the communication signal.

10. A millimeter-wave communication interference suppression method according to claim 3, characterized in that, Perform the interference feature library update operation, including: Redundancy is eliminated from the interference feature parameters stored in the feature library. For interference of the same type, the three sets of feature parameter samples with the greatest feature differences are retained. The feature parameters include the modulation method, duty cycle, center frequency, bandwidth and spatial transmission characteristics of the interference. The feature library is stored in a distributed storage mode.

Images